diff options
Diffstat (limited to 'llvm')
276 files changed, 10754 insertions, 5695 deletions
diff --git a/llvm/docs/AMDGPUDwarfExtensionsForHeterogeneousDebugging.rst b/llvm/docs/AMDGPUDwarfExtensionsForHeterogeneousDebugging.rst index ba670d3..f472b862 100644 --- a/llvm/docs/AMDGPUDwarfExtensionsForHeterogeneousDebugging.rst +++ b/llvm/docs/AMDGPUDwarfExtensionsForHeterogeneousDebugging.rst @@ -37,13 +37,13 @@ includes contributions to open source projects such as LLVM [:ref:`LLVM The LLVM compiler has upstream support for commercially available AMD GPU hardware (AMDGPU) [:ref:`AMDGPU-LLVM <amdgpu-dwarf-AMDGPU-LLVM>`]. The open -source ROCgdb [:ref:`AMD-ROCgdb <amdgpu-dwarf-AMD-ROCgdb>`] GDB based debugger +source ROCgdb [:ref:`AMD-ROCgdb <amdgpu-dwarf-AMD-ROCgdb>`] GDB-based debugger also has support for AMDGPU which is being upstreamed. Support for AMDGPU is also being added by third parties to the GCC [:ref:`GCC <amdgpu-dwarf-GCC>`] compiler and the Perforce TotalView HPC Debugger [:ref:`Perforce-TotalView <amdgpu-dwarf-Perforce-TotalView>`]. -To support debugging heterogeneous programs several features that are not +To support debugging heterogeneous programs, several features that are not provided by current DWARF Version 5 [:ref:`DWARF <amdgpu-dwarf-DWARF>`] have been identified. The :ref:`amdgpu-dwarf-extensions` section gives an overview of the extensions devised to address the missing features. The extensions seek to @@ -107,7 +107,7 @@ for each in terms of heterogeneous debugging. DWARF Version 5 does not allow location descriptions to be entries on the DWARF expression stack. They can only be the final result of the evaluation of a DWARF expression. However, by allowing a location description to be a first-class -entry on the DWARF expression stack it becomes possible to compose expressions +entry on the DWARF expression stack, it becomes possible to compose expressions containing both values and location descriptions naturally. It allows objects to be located in any kind of memory address space, in registers, be implicit values, be undefined, or a composite of any of these. @@ -123,20 +123,20 @@ non-default address spaces and generalizing the power of composite location descriptions to any kind of location description. For those familiar with the definition of location descriptions in DWARF Version -5, the definitions in these extensions are presented differently, but does in +5, the definitions in these extensions are presented differently, but do in fact define the same concept with the same fundamental semantics. However, it does so in a way that allows the concept to extend to support address spaces, bit addressing, the ability for composite location descriptions to be composed of any kind of location description, and the ability to support objects located at multiple places. Collectively these changes expand the set of architectures -that can be supported and improves support for optimized code. +that can be supported and improve support for optimized code. Several approaches were considered, and the one presented, together with the extensions it enables, appears to be the simplest and cleanest one that offers the greatest improvement of DWARF's ability to support debugging optimized GPU and non-GPU code. Examining the GDB debugger and LLVM compiler, it appears only to require modest changes as they both already have to support general use of -location descriptions. It is anticipated that will also be the case for other +location descriptions. It is anticipated that this will also be the case for other debuggers and compilers. GDB has been modified to evaluate DWARF Version 5 expressions with location @@ -156,7 +156,7 @@ DWARF Expression Stack* [:ref:`AMDGPU-DWARF-LOC 2.2 Generalize CFI to Allow Any Location Description Kind --------------------------------------------------------- -CFI describes restoring callee saved registers that are spilled. Currently CFI +CFI describes restoring callee saved registers that are spilled. Currently, CFI only allows a location description that is a register, memory address, or implicit location description. AMDGPU optimized code may spill scalar registers into portions of vector registers. This requires extending CFI to allow any @@ -223,7 +223,7 @@ infinite precision offsets to allow it to correctly track a series of positive and negative offsets that may transiently overflow or underflow, but end up in range. This is simple for the arithmetic operations as they are defined in terms of two's complement arithmetic on a base type of a fixed size. Therefore, the -offset operation define that integer overflow is ill-formed. This is in contrast +offset operation defines that integer overflow is ill-formed. This is in contrast to the ``DW_OP_plus``, ``DW_OP_plus_uconst``, and ``DW_OP_minus`` arithmetic operations which define that it causes wrap-around. @@ -359,7 +359,7 @@ address space at a fixed address. The ``DW_OP_LLVM_form_aspace_address`` (see :ref:`amdgpu-dwarf-memory-location-description-operations`) operation is defined -to create a memory location description from an address and address space. If +to create a memory location description from an address and address space. It can be used to specify the location of a variable that is allocated in a specific address space. This allows the size of addresses in an address space to be larger than the generic type. It also allows a consumer great implementation @@ -372,7 +372,7 @@ In contrast, if the ``DW_OP_LLVM_form_aspace_address`` operation had been defined to produce a value, and an implicit conversion to a memory location description was defined, then it would be limited to the size of the generic type (which matches the size of the default address space). An implementation -would likely have to use *reserved ranges* of value to represent different +would likely have to use *reserved ranges* of values to represent different address spaces. Such a value would likely not match any address value in the actual hardware. That would require the consumer to have special treatment for such values. @@ -528,7 +528,7 @@ active. To describe the conceptual location of non-active lanes requires an attribute that has an expression that computes the source location PC for each lane. -For efficiency, the expression calculates the source location the wavefront as a +For efficiency, the expression calculates the source location of the wavefront as a whole. This can be done using the ``DW_OP_LLVM_select_bit_piece`` (see :ref:`amdgpu-dwarf-operation-to-create-vector-composite-location-descriptions`) operation. @@ -564,7 +564,7 @@ information entry to indicate that there is additional target architecture specific information in the debugging information entries of that compilation unit. This allows a consumer to know what extensions are present in the debugger information entries as is possible with the augmentation string of other -sections. See . +sections. The format that should be used for an augmentation string is also recommended. This allows a consumer to parse the string when it contains information from @@ -581,7 +581,7 @@ See :ref:`amdgpu-dwarf-full-and-partial-compilation-unit-entries`, AMDGPU supports programming languages that include online compilation where the source text may be created at runtime. For example, the OpenCL and HIP language -runtimes support online compilation. To support is, a way to embed the source +runtimes support online compilation. To support this, a way to embed the source text in the debug information is provided. See :ref:`amdgpu-dwarf-line-number-information`. @@ -589,16 +589,16 @@ See :ref:`amdgpu-dwarf-line-number-information`. 2.17 Allow MD5 Checksums to be Optionally Present ------------------------------------------------- -In DWARF Version 5 the file timestamp and file size can be optional, but if the -MD5 checksum is present it must be valid for all files. This is a problem if +In DWARF Version 5, the file timestamp and file size can be optional, but if the +MD5 checksum is present, it must be valid for all files. This is a problem if using link time optimization to combine compilation units where some have MD5 -checksums and some do not. Therefore, sSupport to allow MD5 checksums to be -optionally present in the line table is added. +checksums, and others do not. Therefore, the line table is extended to allow MD5 +checksums to be optional. See :ref:`amdgpu-dwarf-line-number-information`. -2.18 Add the HIP Programing Language ------------------------------------- +2.18 Add the HIP Programming Language +------------------------------------- The HIP programming language [:ref:`HIP <amdgpu-dwarf-HIP>`], which is supported by the AMDGPU, is added. @@ -617,7 +617,7 @@ hardware to allow a single instruction to execute multiple iterations using vector registers. Note that although this is similar to SIMT execution, the way a client debugger -uses the information is fundamentally different. In SIMT execution the debugger +uses the information is fundamentally different. In SIMT execution, the debugger needs to present the concurrent execution as distinct source language threads that the user can list and switch focus between. With iteration concurrency optimizations, such as software pipelining and vectorized SIMD, the debugger @@ -648,7 +648,7 @@ language loop iterations are executing concurrently. See It is common in SIMD vectorization for the compiler to generate code that promotes portions of an array into vector registers. For example, if the hardware has vector registers with 8 elements, and 8 wide SIMD instructions, the -compiler may vectorize a loop so that is executes 8 iterations concurrently for +compiler may vectorize a loop so that it executes 8 iterations concurrently for each vectorized loop iteration. On the first iteration of the generated vectorized loop, iterations 0 to 7 of @@ -691,7 +691,7 @@ Inside the loop body, the machine code loads ``src[i]`` and ``dst[i]`` into registers, adds them, and stores the result back into ``dst[i]``. Considering the location of ``dst`` and ``src`` in the loop body, the elements -``dst[i]`` and ``src[i]`` would be located in registers, all other elements are +``dst[i]`` and ``src[i]`` would be located in registers; all other elements are located in memory. Let register ``R0`` contain the base address of ``dst``, register ``R1`` contain ``i``, and register ``R2`` contain the registerized ``dst[i]`` element. We can describe the location of ``dst`` as a memory location @@ -722,7 +722,7 @@ with a register location overlaid at a runtime offset involving ``i``: ---------------------------------------------- AMDGPU supports languages, such as OpenCL, that define source language memory -spaces. Support is added to define language specific memory spaces so they can +spaces. Support is added to define language-specific memory spaces so they can be used in a consistent way by consumers. See :ref:`amdgpu-dwarf-memory-spaces`. A new attribute ``DW_AT_LLVM_memory_space`` is added to support using memory @@ -738,9 +738,9 @@ accommodates only 32 unique operations. In practice, the lack of a central registry and a desire for backwards compatibility means vendor extensions are never retired, even when standard versions are accepted into DWARF proper. This has produced a situation where the effective encoding space available for new -vendor extensions is miniscule today. +vendor extensions is minuscule today. -To expand this encoding space a new DWARF operation ``DW_OP_LLVM_user`` is +To expand this encoding space, a new DWARF operation ``DW_OP_LLVM_user`` is added which acts as a "prefix" for vendor extensions. It is followed by a ULEB128 encoded vendor extension opcode, which is then followed by the operands of the corresponding vendor extension operation. @@ -776,7 +776,7 @@ A. Changes Relative to DWARF Version 5 .. note:: Notes are included to describe how the changes are to be applied to the - DWARF Version 5 standard. They also describe rational and issues that may + DWARF Version 5 standard. They also describe rationale and issues that may need further consideration. A.2 General Description @@ -898,7 +898,7 @@ elements that can be specified are: *A current lane* - The 0 based SIMT lane identifier to be used in evaluating a user presented + The 0-based SIMT lane identifier to be used in evaluating a user presented expression. This applies to source languages that are implemented for a target architecture using a SIMT execution model. These implementations map source language threads of execution to lanes of the target architecture threads. @@ -917,7 +917,7 @@ elements that can be specified are: *A current iteration* - The 0 based source language iteration instance to be used in evaluating a user + The 0-based source language iteration instance to be used in evaluating a user presented expression. This applies to target architectures that support optimizations that result in executing multiple source language loop iterations concurrently. @@ -1845,7 +1845,7 @@ There are these special value operations currently defined: interpreted as a value of T. If a conversion is wanted it can be done explicitly using a ``DW_OP_convert`` operation. - GDB has a per register hook that allows a target specific conversion on a + GDB has a per register hook that allows a target-specific conversion on a register by register basis. It defaults to truncation of bigger registers. Removing use of the target hook does not cause any test failures in common architectures. If the compiler for a target architecture did want some @@ -1855,7 +1855,7 @@ There are these special value operations currently defined: If T is a larger type than the register size, then the default GDB register hook reads bytes from the next register (or reads out of bounds for the last register!). Removing use of the target hook does not cause - any test failures in common architectures (except an illegal hand written + any test failures in common architectures (except an illegal hand-written assembly test). If a target architecture requires this behavior, these extensions allow a composite location description to be used to combine multiple registers. @@ -2283,7 +2283,7 @@ bit offset equal to V scaled by 8 (the byte size). The implicit conversion could also be defined as target architecture specific. For example, GDB checks if V is an integral type. If it is not it gives an error. Otherwise, GDB zero-extends V to 64 bits. If the GDB target defines a - hook function, then it is called. The target specific hook function can modify + hook function, then it is called. The target-specific hook function can modify the 64-bit value, possibly sign extending based on the original value type. Finally, GDB treats the 64-bit value V as a memory location address. diff --git a/llvm/docs/AMDGPUUsage.rst b/llvm/docs/AMDGPUUsage.rst index a4d110f..402fd05 100644 --- a/llvm/docs/AMDGPUUsage.rst +++ b/llvm/docs/AMDGPUUsage.rst @@ -4172,7 +4172,7 @@ non-AMD key names should be prefixed by "*vendor-name*.". "Image", or "Pipe". This may be more restrictive than indicated by "AccQual" to reflect what the - kernel actual does. If not + kernel actually does. If not present then the runtime must assume what is implied by "AccQual" and "IsConst". Values @@ -5436,8 +5436,8 @@ The fields used by CP for code objects before V3 also match those specified in ``COMPUTE_PGM_RSRC1.PRIORITY``. 13:12 2 bits FLOAT_ROUND_MODE_32 Wavefront starts execution with specified rounding - mode for single (32 - bit) floating point + mode for single (32-bit) + floating point precision floating point operations. @@ -5769,7 +5769,7 @@ The fields used by CP for code objects before V3 also match those specified in Wavefront starts execution with memory violation - exceptions exceptions + exceptions enabled which are generated when a memory violation has occurred for this wavefront from @@ -6005,7 +6005,7 @@ The fields used by CP for code objects before V3 also match those specified in FLOAT_DENORM_MODE_FLUSH_NONE 3 No Flush ====================================== ===== ==================================== - Denormal flushing is sign respecting. i.e. the behavior expected by + Denormal flushing is sign respecting, i.e., the behavior expected by ``"denormal-fp-math"="preserve-sign"``. The behavior is undefined with ``"denormal-fp-math"="positive-zero"`` @@ -16831,7 +16831,7 @@ For GFX125x: * Some memory operations contain a ``nv`` bit, for "non-volatile", which indicates memory that is not expected to change during a kernel's execution. This information is propagated to the cache lines for that address - (refered to as ``$nv``). + (referred to as ``$nv``). * When ``nv=0`` reads hit dirty ``$nv=1`` data in cache, the hardware will writeback the data to the next level in the hierarchy and then subsequently read @@ -18970,7 +18970,7 @@ On entry to a function: #. All other registers are unspecified. #. Any necessary ``s_waitcnt`` has been performed to ensure memory is available to the function. -#. Use pass-by-reference (byref) in stead of pass-by-value (byval) for struct +#. Use pass-by-reference (byref) instead of pass-by-value (byval) for struct arguments in C ABI. Callee is responsible for allocating stack memory and copying the value of the struct if modified. Note that the backend still supports byval for struct arguments. @@ -20214,7 +20214,7 @@ from the value of the ``-mcpu`` option that is passed to the assembler. .amdgpu_hsa_kernel (name) +++++++++++++++++++++++++ -This directives specifies that the symbol with given name is a kernel entry +This directive specifies that the symbol with given name is a kernel entry point (label) and the object should contain corresponding symbol of type STT_AMDGPU_HSA_KERNEL. diff --git a/llvm/docs/CMakeLists.txt b/llvm/docs/CMakeLists.txt index b4522e3..fc37c6d 100644 --- a/llvm/docs/CMakeLists.txt +++ b/llvm/docs/CMakeLists.txt @@ -136,17 +136,23 @@ if( NOT uses_ocaml LESS 0 AND LLVM_ENABLE_OCAMLDOC ) list(APPEND odoc_files -load ${odoc_file}) endforeach() - add_custom_target(ocaml_doc - COMMAND ${CMAKE_COMMAND} -E remove_directory ${CMAKE_CURRENT_BINARY_DIR}/ocamldoc/html - COMMAND ${CMAKE_COMMAND} -E make_directory ${CMAKE_CURRENT_BINARY_DIR}/ocamldoc/html - COMMAND ${OCAMLFIND} ocamldoc -d ${CMAKE_CURRENT_BINARY_DIR}/ocamldoc/html - -sort -colorize-code -html ${odoc_files} - COMMAND ${CMAKE_COMMAND} -E copy ${CMAKE_CURRENT_SOURCE_DIR}/_ocamldoc/style.css - ${CMAKE_CURRENT_BINARY_DIR}/ocamldoc/html) + set(OCAML_DOC_ADD_TO_ALL "") + if(LLVM_BUILD_DOCS) + set(OCAML_DOC_ADD_TO_ALL ALL) + endif() + + add_custom_target(ocaml_doc ${OCAML_DOC_ADD_TO_ALL} + COMMAND ${CMAKE_COMMAND} -E remove_directory ${CMAKE_CURRENT_BINARY_DIR}/ocamldoc/html + COMMAND ${CMAKE_COMMAND} -E make_directory ${CMAKE_CURRENT_BINARY_DIR}/ocamldoc/html + COMMAND ${OCAMLFIND} ocamldoc -d ${CMAKE_CURRENT_BINARY_DIR}/ocamldoc/html + -sort -colorize-code -html ${odoc_files} + COMMAND ${CMAKE_COMMAND} -E copy ${CMAKE_CURRENT_SOURCE_DIR}/_ocamldoc/style.css + ${CMAKE_CURRENT_BINARY_DIR}/ocamldoc/html) add_dependencies(ocaml_doc ${doc_targets}) - if (NOT LLVM_INSTALL_TOOLCHAIN_ONLY) + + if (NOT LLVM_INSTALL_TOOLCHAIN_ONLY AND LLVM_BUILD_DOCS) # ./ suffix is needed to copy the contents of html directory without # appending html/ into LLVM_INSTALL_OCAMLDOC_HTML_DIR. install(DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}/ocamldoc/html/. diff --git a/llvm/docs/CallGraphSection.md b/llvm/docs/CallGraphSection.md index 8b18727..84d6061 100644 --- a/llvm/docs/CallGraphSection.md +++ b/llvm/docs/CallGraphSection.md @@ -1,10 +1,10 @@ -# .callgraph Section Layout +# .llvm.callgraph Section Layout -The `.callgraph` section is used to store call graph information for each function. The section contains a series of records, with each record corresponding to a single function. +The `.llvm.callgraph` section is used to store call graph information for each function. The section contains a series of records, with each record corresponding to a single function. ## Per Function Record Layout -Each record in the `.callgraph` section has the following binary layout: +Each record in the `.llvm.callgraph` section has the following binary layout: | Field | Type | Size (bits) | Description | | -------------------------------------- | ------------- | ----------- | ------------------------------------------------------------------------------------------------------- | diff --git a/llvm/docs/GettingStartedVS.rst b/llvm/docs/GettingStartedVS.rst index bc5746d..e65fd8f 100644 --- a/llvm/docs/GettingStartedVS.rst +++ b/llvm/docs/GettingStartedVS.rst @@ -126,6 +126,15 @@ These instructions were tested with Visual Studio 2019 and Python 3.9.6: cmake -S llvm\llvm -B build -DLLVM_ENABLE_PROJECTS=clang -DLLVM_TARGETS_TO_BUILD=X86 -Thost=x64 exit + .. note:: + By default, the Visual Studio project files generated by CMake use the + 32-bit toolset. If you are developing on a 64-bit version of Windows and + want to use the 64-bit toolset, pass the ``-Thost=x64`` flag when + generating the Visual Studio solution. This requires CMake 3.8.0 or later. + + For Windows on Arm the equivalent is ``-Thost=ARM64``, but this the default + for those hosts, so you do not have to use this option. + ``LLVM_ENABLE_PROJECTS`` specifies any additional LLVM projects you want to build while ``LLVM_TARGETS_TO_BUILD`` selects the compiler targets. If ``LLVM_TARGETS_TO_BUILD`` is omitted by default all targets are built @@ -149,10 +158,6 @@ These instructions were tested with Visual Studio 2019 and Python 3.9.6: * CMake generates project files for all build types. To select a specific build type, use the Configuration manager from the VS IDE or the ``/property:Configuration`` command-line option when using MSBuild. - * By default, the Visual Studio project files generated by CMake use the - 32-bit toolset. If you are developing on a 64-bit version of Windows and - want to use the 64-bit toolset, pass the ``-Thost=x64`` flag when - generating the Visual Studio solution. This requires CMake 3.8.0 or later. 13. Start Visual Studio and select configuration: diff --git a/llvm/docs/LangRef.rst b/llvm/docs/LangRef.rst index 4884e2d..0c54f57 100644 --- a/llvm/docs/LangRef.rst +++ b/llvm/docs/LangRef.rst @@ -7517,12 +7517,12 @@ sections that the user does not want removed after linking. '``unpredictable``' Metadata ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ -``unpredictable`` metadata may be attached to any branch or switch -instruction. It can be used to express the unpredictability of control -flow. Similar to the ``llvm.expect`` intrinsic, it may be used to alter -optimizations related to compare and branch instructions. The metadata -is treated as a boolean value; if it exists, it signals that the branch -or switch that it is attached to is completely unpredictable. +``unpredictable`` metadata may be attached to any branch, select, or switch +instruction. It can be used to express the unpredictability of control flow. +Similar to the ``llvm.expect`` intrinsic, it may be used to alter optimizations +related to compare and branch instructions. The metadata is treated as a +boolean value; if it exists, it signals that the branch, select, or switch that +it is attached to is completely unpredictable. .. _md_dereferenceable: diff --git a/llvm/docs/ReleaseNotes.md b/llvm/docs/ReleaseNotes.md index 640516a..c352cd6 100644 --- a/llvm/docs/ReleaseNotes.md +++ b/llvm/docs/ReleaseNotes.md @@ -139,6 +139,8 @@ Changes to the Windows Target Changes to the X86 Backend -------------------------- +* `-mcpu=wildcatlake` is now supported. + Changes to the OCaml bindings ----------------------------- @@ -174,6 +176,9 @@ Changes to LLDB * LLDB can now set breakpoints, show backtraces, and display variables when debugging Wasm with supported runtimes (WAMR and V8). +* LLDB no longer stops processes by default when receiving SIGWINCH signals + (window resize events) on Linux. This is the default on other Unix platforms. + You can re-enable it using `process handle --notify=true --stop=true SIGWINCH`. * The `show-progress` setting, which became a NOOP with the introduction of the statusline, now defaults to off and controls using OSC escape codes to show a native progress bar in supporting terminals like Ghostty and ConEmu. diff --git a/llvm/include/llvm/ADT/StringSwitch.h b/llvm/include/llvm/ADT/StringSwitch.h index a96535c..26d5682 100644 --- a/llvm/include/llvm/ADT/StringSwitch.h +++ b/llvm/include/llvm/ADT/StringSwitch.h @@ -14,6 +14,7 @@ #define LLVM_ADT_STRINGSWITCH_H #include "llvm/ADT/StringRef.h" +#include "llvm/Support/Compiler.h" #include "llvm/Support/ErrorHandling.h" #include <cassert> #include <cstring> @@ -38,7 +39,7 @@ namespace llvm { /// .Case("green", Green) /// .Case("blue", Blue) /// .Case("indigo", Indigo) -/// .Cases("violet", "purple", Violet) +/// .Cases({"violet", "purple"}, Violet) /// .Default(UnknownColor); /// \endcode template<typename T, typename R = T> @@ -54,21 +55,18 @@ public: explicit StringSwitch(StringRef S) : Str(S), Result() { } + StringSwitch(StringSwitch &&) = default; + // StringSwitch is not copyable. StringSwitch(const StringSwitch &) = delete; // StringSwitch is not assignable due to 'Str' being 'const'. void operator=(const StringSwitch &) = delete; - void operator=(StringSwitch &&other) = delete; - - StringSwitch(StringSwitch &&other) - : Str(other.Str), Result(std::move(other.Result)) { } - - ~StringSwitch() = default; + void operator=(StringSwitch &&) = delete; // Case-sensitive case matchers StringSwitch &Case(StringLiteral S, T Value) { - CaseImpl(Value, S); + CaseImpl(S, Value); return *this; } @@ -88,63 +86,68 @@ public: StringSwitch &Cases(std::initializer_list<StringLiteral> CaseStrings, T Value) { - return CasesImpl(Value, CaseStrings); + return CasesImpl(CaseStrings, Value); } StringSwitch &Cases(StringLiteral S0, StringLiteral S1, T Value) { - return CasesImpl(Value, {S0, S1}); + return CasesImpl({S0, S1}, Value); } StringSwitch &Cases(StringLiteral S0, StringLiteral S1, StringLiteral S2, T Value) { - return CasesImpl(Value, {S0, S1, S2}); + return CasesImpl({S0, S1, S2}, Value); } StringSwitch &Cases(StringLiteral S0, StringLiteral S1, StringLiteral S2, StringLiteral S3, T Value) { - return CasesImpl(Value, {S0, S1, S2, S3}); + return CasesImpl({S0, S1, S2, S3}, Value); } StringSwitch &Cases(StringLiteral S0, StringLiteral S1, StringLiteral S2, StringLiteral S3, StringLiteral S4, T Value) { - return CasesImpl(Value, {S0, S1, S2, S3, S4}); + return CasesImpl({S0, S1, S2, S3, S4}, Value); } + [[deprecated("Pass cases in std::initializer_list instead")]] StringSwitch &Cases(StringLiteral S0, StringLiteral S1, StringLiteral S2, StringLiteral S3, StringLiteral S4, StringLiteral S5, T Value) { - return CasesImpl(Value, {S0, S1, S2, S3, S4, S5}); + return CasesImpl({S0, S1, S2, S3, S4, S5}, Value); } + [[deprecated("Pass cases in std::initializer_list instead")]] StringSwitch &Cases(StringLiteral S0, StringLiteral S1, StringLiteral S2, StringLiteral S3, StringLiteral S4, StringLiteral S5, StringLiteral S6, T Value) { - return CasesImpl(Value, {S0, S1, S2, S3, S4, S5, S6}); + return CasesImpl({S0, S1, S2, S3, S4, S5, S6}, Value); } + [[deprecated("Pass cases in std::initializer_list instead")]] StringSwitch &Cases(StringLiteral S0, StringLiteral S1, StringLiteral S2, StringLiteral S3, StringLiteral S4, StringLiteral S5, StringLiteral S6, StringLiteral S7, T Value) { - return CasesImpl(Value, {S0, S1, S2, S3, S4, S5, S6, S7}); + return CasesImpl({S0, S1, S2, S3, S4, S5, S6, S7}, Value); } + [[deprecated("Pass cases in std::initializer_list instead")]] StringSwitch &Cases(StringLiteral S0, StringLiteral S1, StringLiteral S2, StringLiteral S3, StringLiteral S4, StringLiteral S5, StringLiteral S6, StringLiteral S7, StringLiteral S8, T Value) { - return CasesImpl(Value, {S0, S1, S2, S3, S4, S5, S6, S7, S8}); + return CasesImpl({S0, S1, S2, S3, S4, S5, S6, S7, S8}, Value); } + [[deprecated("Pass cases in std::initializer_list instead")]] StringSwitch &Cases(StringLiteral S0, StringLiteral S1, StringLiteral S2, StringLiteral S3, StringLiteral S4, StringLiteral S5, StringLiteral S6, StringLiteral S7, StringLiteral S8, StringLiteral S9, T Value) { - return CasesImpl(Value, {S0, S1, S2, S3, S4, S5, S6, S7, S8, S9}); + return CasesImpl({S0, S1, S2, S3, S4, S5, S6, S7, S8, S9}, Value); } // Case-insensitive case matchers. StringSwitch &CaseLower(StringLiteral S, T Value) { - CaseLowerImpl(Value, S); + CaseLowerImpl(S, Value); return *this; } @@ -164,26 +167,26 @@ public: StringSwitch &CasesLower(std::initializer_list<StringLiteral> CaseStrings, T Value) { - return CasesLowerImpl(Value, CaseStrings); + return CasesLowerImpl(CaseStrings, Value); } StringSwitch &CasesLower(StringLiteral S0, StringLiteral S1, T Value) { - return CasesLowerImpl(Value, {S0, S1}); + return CasesLowerImpl({S0, S1}, Value); } StringSwitch &CasesLower(StringLiteral S0, StringLiteral S1, StringLiteral S2, T Value) { - return CasesLowerImpl(Value, {S0, S1, S2}); + return CasesLowerImpl({S0, S1, S2}, Value); } StringSwitch &CasesLower(StringLiteral S0, StringLiteral S1, StringLiteral S2, StringLiteral S3, T Value) { - return CasesLowerImpl(Value, {S0, S1, S2, S3}); + return CasesLowerImpl({S0, S1, S2, S3}, Value); } StringSwitch &CasesLower(StringLiteral S0, StringLiteral S1, StringLiteral S2, StringLiteral S3, StringLiteral S4, T Value) { - return CasesLowerImpl(Value, {S0, S1, S2, S3, S4}); + return CasesLowerImpl({S0, S1, S2, S3, S4}, Value); } [[nodiscard]] R Default(T Value) { @@ -204,7 +207,7 @@ public: private: // Returns true when `Str` matches the `S` argument, and stores the result. - bool CaseImpl(T &Value, StringLiteral S) { + bool CaseImpl(StringLiteral S, T &Value) { if (!Result && Str == S) { Result = std::move(Value); return true; @@ -214,7 +217,7 @@ private: // Returns true when `Str` matches the `S` argument (case-insensitive), and // stores the result. - bool CaseLowerImpl(T &Value, StringLiteral S) { + bool CaseLowerImpl(StringLiteral S, T &Value) { if (!Result && Str.equals_insensitive(S)) { Result = std::move(Value); return true; @@ -222,20 +225,20 @@ private: return false; } - StringSwitch &CasesImpl(T &Value, - std::initializer_list<StringLiteral> Cases) { + StringSwitch &CasesImpl(std::initializer_list<StringLiteral> Cases, + T &Value) { // Stop matching after the string is found. for (StringLiteral S : Cases) - if (CaseImpl(Value, S)) + if (CaseImpl(S, Value)) break; return *this; } - StringSwitch &CasesLowerImpl(T &Value, - std::initializer_list<StringLiteral> Cases) { + StringSwitch &CasesLowerImpl(std::initializer_list<StringLiteral> Cases, + T &Value) { // Stop matching after the string is found. for (StringLiteral S : Cases) - if (CaseLowerImpl(Value, S)) + if (CaseLowerImpl(S, Value)) break; return *this; } diff --git a/llvm/include/llvm/Analysis/IR2Vec.h b/llvm/include/llvm/Analysis/IR2Vec.h index 6bc51fe..5ad6288 100644 --- a/llvm/include/llvm/Analysis/IR2Vec.h +++ b/llvm/include/llvm/Analysis/IR2Vec.h @@ -575,7 +575,7 @@ public: /// cached embeddings should be invalidated to ensure /// correctness/recomputation. This is a no-op for SymbolicEmbedder but /// removes all the cached entries in FlowAwareEmbedder. - virtual void invalidateEmbeddings() { return; } + virtual void invalidateEmbeddings() {} }; /// Class for computing the Symbolic embeddings of IR2Vec. diff --git a/llvm/include/llvm/Analysis/StaticDataProfileInfo.h b/llvm/include/llvm/Analysis/StaticDataProfileInfo.h index f06e7ce..ac03137 100644 --- a/llvm/include/llvm/Analysis/StaticDataProfileInfo.h +++ b/llvm/include/llvm/Analysis/StaticDataProfileInfo.h @@ -32,8 +32,11 @@ bool IsAnnotationOK(const GlobalVariable &GV); /// profile information and provides methods to operate on them. class StaticDataProfileInfo { public: - /// Accummulate the profile count of a constant that will be lowered to static - /// data sections. + /// A constant is tracked only if the following conditions are met. + /// 1) It has local (i.e., private or internal) linkage. + // 2) Its data kind is one of {.rodata, .data, .bss, .data.rel.ro}. + // 3) It's eligible for section prefix annotation. See `AnnotationKind` + // above for ineligible reasons. DenseMap<const Constant *, uint64_t> ConstantProfileCounts; /// Keeps track of the constants that are seen at least once without profile @@ -44,8 +47,31 @@ public: LLVM_ABI std::optional<uint64_t> getConstantProfileCount(const Constant *C) const; + /// Use signed enums for enum value comparison, and make 'LukewarmOrUnknown' + /// as 0 so any accidentally uninitialized value will default to unknown. + enum class StaticDataHotness : int8_t { + Cold = -1, + LukewarmOrUnknown = 0, + Hot = 1, + }; + + /// Return the hotness of the constant \p C based on its profile count \p + /// Count. + LLVM_ABI StaticDataHotness getConstantHotnessUsingProfileCount( + const Constant *C, const ProfileSummaryInfo *PSI, uint64_t Count) const; + + /// Return the hotness based on section prefix \p SectionPrefix. + LLVM_ABI StaticDataHotness getSectionHotnessUsingDataAccessProfile( + std::optional<StringRef> SectionPrefix) const; + + /// Return the string representation of the hotness enum \p Hotness. + LLVM_ABI StringRef hotnessToStr(StaticDataHotness Hotness) const; + + bool EnableDataAccessProf = false; + public: - StaticDataProfileInfo() = default; + StaticDataProfileInfo(bool EnableDataAccessProf) + : EnableDataAccessProf(EnableDataAccessProf) {} /// If \p Count is not nullopt, add it to the profile count of the constant \p /// C in a saturating way, and clamp the count to \p getInstrMaxCountValue if @@ -54,14 +80,10 @@ public: LLVM_ABI void addConstantProfileCount(const Constant *C, std::optional<uint64_t> Count); - /// Return a section prefix for the constant \p C based on its profile count. - /// - If a constant doesn't have a counter, return an empty string. - /// - Otherwise, - /// - If it has a hot count, return "hot". - /// - If it is seen by unprofiled function, return an empty string. - /// - If it has a cold count, return "unlikely". - /// - Otherwise (e.g. it's used by lukewarm functions), return an empty - /// string. + /// Given a constant \p C, returns a section prefix. + /// If \p C is a global variable, the section prefix is the bigger one + /// between its existing section prefix and its use profile count. Otherwise, + /// the section prefix is based on its use profile count. LLVM_ABI StringRef getConstantSectionPrefix( const Constant *C, const ProfileSummaryInfo *PSI) const; }; diff --git a/llvm/include/llvm/Analysis/TargetLibraryInfo.h b/llvm/include/llvm/Analysis/TargetLibraryInfo.h index 26963ed..3f39b47 100644 --- a/llvm/include/llvm/Analysis/TargetLibraryInfo.h +++ b/llvm/include/llvm/Analysis/TargetLibraryInfo.h @@ -373,12 +373,10 @@ public: /// Disables all builtins. /// /// This can be used for options like -fno-builtin. - void disableAllFunctions() LLVM_ATTRIBUTE_UNUSED { - OverrideAsUnavailable.set(); - } + [[maybe_unused]] void disableAllFunctions() { OverrideAsUnavailable.set(); } /// Forces a function to be marked as unavailable. - void setUnavailable(LibFunc F) LLVM_ATTRIBUTE_UNUSED { + [[maybe_unused]] void setUnavailable(LibFunc F) { assert(F < OverrideAsUnavailable.size() && "out-of-bounds LibFunc"); OverrideAsUnavailable.set(F); } diff --git a/llvm/include/llvm/BinaryFormat/ELFRelocs/AArch64.def b/llvm/include/llvm/BinaryFormat/ELFRelocs/AArch64.def index 8dcc292..1cfcdbf 100644 --- a/llvm/include/llvm/BinaryFormat/ELFRelocs/AArch64.def +++ b/llvm/include/llvm/BinaryFormat/ELFRelocs/AArch64.def @@ -62,6 +62,7 @@ ELF_RELOC(R_AARCH64_LD64_GOTPAGE_LO15, 0x139) ELF_RELOC(R_AARCH64_PLT32, 0x13a) ELF_RELOC(R_AARCH64_GOTPCREL32, 0x13b) ELF_RELOC(R_AARCH64_PATCHINST, 0x13c) +ELF_RELOC(R_AARCH64_FUNCINIT64, 0x13d) // General dynamic TLS relocations ELF_RELOC(R_AARCH64_TLSGD_ADR_PREL21, 0x200) ELF_RELOC(R_AARCH64_TLSGD_ADR_PAGE21, 0x201) diff --git a/llvm/include/llvm/CAS/CASID.h b/llvm/include/llvm/CAS/CASID.h index 8820994..f508ed3 100644 --- a/llvm/include/llvm/CAS/CASID.h +++ b/llvm/include/llvm/CAS/CASID.h @@ -95,8 +95,7 @@ public: } friend hash_code hash_value(const CASID &ID) { - ArrayRef<uint8_t> Hash = ID.getHash(); - return hash_combine_range(Hash.begin(), Hash.end()); + return hash_combine_range(ID.getHash()); } const CASContext &getContext() const { diff --git a/llvm/include/llvm/CodeGen/AsmPrinter.h b/llvm/include/llvm/CodeGen/AsmPrinter.h index 19ca444..9ace255 100644 --- a/llvm/include/llvm/CodeGen/AsmPrinter.h +++ b/llvm/include/llvm/CodeGen/AsmPrinter.h @@ -16,6 +16,7 @@ #define LLVM_CODEGEN_ASMPRINTER_H #include "llvm/ADT/DenseMap.h" +#include "llvm/ADT/IntrusiveRefCntPtr.h" #include "llvm/ADT/MapVector.h" #include "llvm/ADT/SmallSet.h" #include "llvm/ADT/SmallVector.h" @@ -87,6 +88,10 @@ namespace remarks { class RemarkStreamer; } +namespace vfs { +class FileSystem; +} + /// This class is intended to be used as a driving class for all asm writers. class LLVM_ABI AsmPrinter : public MachineFunctionPass { public: @@ -105,6 +110,9 @@ public: /// generating (such as the current section etc). std::unique_ptr<MCStreamer> OutStreamer; + /// The VFS to resolve asm include directives. + IntrusiveRefCntPtr<vfs::FileSystem> VFS; + /// The current machine function. MachineFunction *MF = nullptr; diff --git a/llvm/include/llvm/CodeGen/LiveIntervals.h b/llvm/include/llvm/CodeGen/LiveIntervals.h index 1050b3d..c252f9d 100644 --- a/llvm/include/llvm/CodeGen/LiveIntervals.h +++ b/llvm/include/llvm/CodeGen/LiveIntervals.h @@ -229,8 +229,8 @@ public: /// doing something wrong if you call pruneValue directly on a /// LiveInterval. Indeed, you are supposed to call pruneValue on the main /// LiveRange and all the LiveRanges of the subranges if any. - LLVM_ATTRIBUTE_UNUSED void pruneValue(LiveInterval &, SlotIndex, - SmallVectorImpl<SlotIndex> *) { + [[maybe_unused]] void pruneValue(LiveInterval &, SlotIndex, + SmallVectorImpl<SlotIndex> *) { llvm_unreachable( "Use pruneValue on the main LiveRange and on each subrange"); } diff --git a/llvm/include/llvm/CodeGen/LiveRangeCalc.h b/llvm/include/llvm/CodeGen/LiveRangeCalc.h index e9b62fb..67f5b69 100644 --- a/llvm/include/llvm/CodeGen/LiveRangeCalc.h +++ b/llvm/include/llvm/CodeGen/LiveRangeCalc.h @@ -259,7 +259,7 @@ public: /// jointly dominated by the blocks corresponding to the slot indices /// in @p Defs. This function is mainly for use in self-verification /// checks. - LLVM_ABI LLVM_ATTRIBUTE_UNUSED static bool + [[maybe_unused]] LLVM_ABI static bool isJointlyDominated(const MachineBasicBlock *MBB, ArrayRef<SlotIndex> Defs, const SlotIndexes &Indexes); }; diff --git a/llvm/include/llvm/CodeGen/SelectionDAGNodes.h b/llvm/include/llvm/CodeGen/SelectionDAGNodes.h index 1169116..69713d0 100644 --- a/llvm/include/llvm/CodeGen/SelectionDAGNodes.h +++ b/llvm/include/llvm/CodeGen/SelectionDAGNodes.h @@ -1950,7 +1950,7 @@ LLVM_ABI bool isOnesOrOnesSplat(SDValue N, bool AllowUndefs = false); /// Return true if the value is a constant 0 integer or a splatted vector of a /// constant 0 integer (with no undefs). -/// Does not permit build vector implicit truncation. +/// Build vector implicit truncation is allowed. LLVM_ABI bool isZeroOrZeroSplat(SDValue N, bool AllowUndefs = false); /// Return true if \p V is either a integer or FP constant. diff --git a/llvm/include/llvm/DebugInfo/GSYM/DwarfTransformer.h b/llvm/include/llvm/DebugInfo/GSYM/DwarfTransformer.h index 77ce052..2c59a52 100644 --- a/llvm/include/llvm/DebugInfo/GSYM/DwarfTransformer.h +++ b/llvm/include/llvm/DebugInfo/GSYM/DwarfTransformer.h @@ -43,8 +43,14 @@ public: /// /// \param LDCS Flag to indicate whether we should load the call site /// information from DWARF `DW_TAG_call_site` entries - DwarfTransformer(DWARFContext &D, GsymCreator &G, bool LDCS = false) - : DICtx(D), Gsym(G), LoadDwarfCallSites(LDCS) {} + /// + /// \param MachO Flag to indicate if the object file is mach-o (Apple's + /// executable format). Apple has some compile unit attributes that look like + /// split DWARF, but they aren't and they can cause warnins to be emitted + /// about missing DWO files. + DwarfTransformer(DWARFContext &D, GsymCreator &G, bool LDCS = false, + bool MachO = false) + : DICtx(D), Gsym(G), LoadDwarfCallSites(LDCS), IsMachO(MachO) {} /// Extract the DWARF from the supplied object file and convert it into the /// Gsym format in the GsymCreator object that is passed in. Returns an @@ -97,6 +103,7 @@ private: DWARFContext &DICtx; GsymCreator &Gsym; bool LoadDwarfCallSites; + bool IsMachO; friend class DwarfTransformerTest; }; diff --git a/llvm/include/llvm/ExecutionEngine/Orc/Shared/AllocationActions.h b/llvm/include/llvm/ExecutionEngine/Orc/Shared/AllocationActions.h index 596cc18..b0197f0 100644 --- a/llvm/include/llvm/ExecutionEngine/Orc/Shared/AllocationActions.h +++ b/llvm/include/llvm/ExecutionEngine/Orc/Shared/AllocationActions.h @@ -13,7 +13,6 @@ #ifndef LLVM_EXECUTIONENGINE_ORC_SHARED_ALLOCATIONACTIONS_H #define LLVM_EXECUTIONENGINE_ORC_SHARED_ALLOCATIONACTIONS_H -#include "llvm/ADT/FunctionExtras.h" #include "llvm/ExecutionEngine/Orc/Shared/ExecutorAddress.h" #include "llvm/ExecutionEngine/Orc/Shared/WrapperFunctionUtils.h" #include "llvm/Support/Compiler.h" @@ -54,9 +53,6 @@ inline size_t numDeallocActions(const AllocActions &AAs) { AAs, [](const AllocActionCallPair &P) { return !!P.Dealloc; }); } -using OnRunFinalizeActionsCompleteFn = - unique_function<void(Expected<std::vector<WrapperFunctionCall>>)>; - /// Run finalize actions. /// /// If any finalize action fails then the corresponding dealloc actions will be @@ -67,16 +63,13 @@ using OnRunFinalizeActionsCompleteFn = /// be returned. The dealloc actions should be run by calling /// runDeallocationActions. If this function succeeds then the AA argument will /// be cleared before the function returns. -LLVM_ABI void runFinalizeActions(AllocActions &AAs, - OnRunFinalizeActionsCompleteFn OnComplete); - -using OnRunDeallocActionsComeleteFn = unique_function<void(Error)>; +LLVM_ABI Expected<std::vector<WrapperFunctionCall>> +runFinalizeActions(AllocActions &AAs); /// Run deallocation actions. /// Dealloc actions will be run in reverse order (from last element of DAs to /// first). -LLVM_ABI void runDeallocActions(ArrayRef<WrapperFunctionCall> DAs, - OnRunDeallocActionsComeleteFn OnComplete); +LLVM_ABI Error runDeallocActions(ArrayRef<WrapperFunctionCall> DAs); using SPSAllocActionCallPair = SPSTuple<SPSWrapperFunctionCall, SPSWrapperFunctionCall>; diff --git a/llvm/include/llvm/IR/CFG.h b/llvm/include/llvm/IR/CFG.h index 7c7e988..96d3b2f 100644 --- a/llvm/include/llvm/IR/CFG.h +++ b/llvm/include/llvm/IR/CFG.h @@ -42,9 +42,9 @@ template <class Ptr, class USE_iterator> // Predecessor Iterator class PredIterator { public: using iterator_category = std::forward_iterator_tag; - using value_type = Ptr; + using value_type = Ptr *; using difference_type = std::ptrdiff_t; - using pointer = Ptr *; + using pointer = Ptr **; using reference = Ptr *; protected: @@ -141,7 +141,8 @@ class SuccIterator std::random_access_iterator_tag, BlockT, int, BlockT *, BlockT *> { public: - using difference_type = int; + using value_type = BlockT *; + using difference_type = std::ptrdiff_t; using pointer = BlockT *; using reference = BlockT *; diff --git a/llvm/include/llvm/Object/ELFTypes.h b/llvm/include/llvm/Object/ELFTypes.h index 5a26e2f..e9a417d 100644 --- a/llvm/include/llvm/Object/ELFTypes.h +++ b/llvm/include/llvm/Object/ELFTypes.h @@ -833,6 +833,7 @@ struct BBAddrMap { bool MultiBBRange : 1; bool OmitBBEntries : 1; bool CallsiteEndOffsets : 1; + bool BBHash : 1; bool hasPGOAnalysis() const { return FuncEntryCount || BBFreq || BrProb; } @@ -845,7 +846,8 @@ struct BBAddrMap { (static_cast<uint8_t>(BrProb) << 2) | (static_cast<uint8_t>(MultiBBRange) << 3) | (static_cast<uint8_t>(OmitBBEntries) << 4) | - (static_cast<uint8_t>(CallsiteEndOffsets) << 5); + (static_cast<uint8_t>(CallsiteEndOffsets) << 5) | + (static_cast<uint8_t>(BBHash) << 6); } // Decodes from minimum bit width representation and validates no @@ -854,7 +856,8 @@ struct BBAddrMap { Features Feat{ static_cast<bool>(Val & (1 << 0)), static_cast<bool>(Val & (1 << 1)), static_cast<bool>(Val & (1 << 2)), static_cast<bool>(Val & (1 << 3)), - static_cast<bool>(Val & (1 << 4)), static_cast<bool>(Val & (1 << 5))}; + static_cast<bool>(Val & (1 << 4)), static_cast<bool>(Val & (1 << 5)), + static_cast<bool>(Val & (1 << 6))}; if (Feat.encode() != Val) return createStringError( std::error_code(), "invalid encoding for BBAddrMap::Features: 0x%x", @@ -864,10 +867,10 @@ struct BBAddrMap { bool operator==(const Features &Other) const { return std::tie(FuncEntryCount, BBFreq, BrProb, MultiBBRange, - OmitBBEntries, CallsiteEndOffsets) == + OmitBBEntries, CallsiteEndOffsets, BBHash) == std::tie(Other.FuncEntryCount, Other.BBFreq, Other.BrProb, Other.MultiBBRange, Other.OmitBBEntries, - Other.CallsiteEndOffsets); + Other.CallsiteEndOffsets, Other.BBHash); } }; @@ -920,17 +923,19 @@ struct BBAddrMap { false}; // Metdata for this basic block. // Offsets of end of call instructions, relative to the basic block start. SmallVector<uint32_t, 1> CallsiteEndOffsets; + uint64_t Hash = 0; // Hash for this basic block. BBEntry(uint32_t ID, uint32_t Offset, uint32_t Size, Metadata MD, - SmallVector<uint32_t, 1> CallsiteEndOffsets) + SmallVector<uint32_t, 1> CallsiteEndOffsets, uint64_t Hash) : ID(ID), Offset(Offset), Size(Size), MD(MD), - CallsiteEndOffsets(std::move(CallsiteEndOffsets)) {} + CallsiteEndOffsets(std::move(CallsiteEndOffsets)), Hash(Hash) {} UniqueBBID getID() const { return {ID, 0}; } bool operator==(const BBEntry &Other) const { return ID == Other.ID && Offset == Other.Offset && Size == Other.Size && - MD == Other.MD && CallsiteEndOffsets == Other.CallsiteEndOffsets; + MD == Other.MD && CallsiteEndOffsets == Other.CallsiteEndOffsets && + Hash == Other.Hash; } bool hasReturn() const { return MD.HasReturn; } diff --git a/llvm/include/llvm/ObjectYAML/ELFYAML.h b/llvm/include/llvm/ObjectYAML/ELFYAML.h index c90591d..a7c7c7c 100644 --- a/llvm/include/llvm/ObjectYAML/ELFYAML.h +++ b/llvm/include/llvm/ObjectYAML/ELFYAML.h @@ -163,6 +163,7 @@ struct BBAddrMapEntry { llvm::yaml::Hex64 Size; llvm::yaml::Hex64 Metadata; std::optional<std::vector<llvm::yaml::Hex64>> CallsiteEndOffsets; + std::optional<llvm::yaml::Hex64> Hash; }; uint8_t Version; llvm::yaml::Hex8 Feature; diff --git a/llvm/include/llvm/ProfileData/InstrProfCorrelator.h b/llvm/include/llvm/ProfileData/InstrProfCorrelator.h index d460eb1..1617ae7 100644 --- a/llvm/include/llvm/ProfileData/InstrProfCorrelator.h +++ b/llvm/include/llvm/ProfileData/InstrProfCorrelator.h @@ -13,6 +13,7 @@ #define LLVM_PROFILEDATA_INSTRPROFCORRELATOR_H #include "llvm/ADT/DenseSet.h" +#include "llvm/DebugInfo/DWARF/DWARFContext.h" #include "llvm/Debuginfod/BuildIDFetcher.h" #include "llvm/Object/BuildID.h" #include "llvm/ProfileData/InstrProf.h" @@ -24,7 +25,6 @@ #include <vector> namespace llvm { -class DWARFContext; class DWARFDie; namespace object { class ObjectFile; diff --git a/llvm/include/llvm/Support/Caching.h b/llvm/include/llvm/Support/Caching.h index 7fd9bef..cebf071 100644 --- a/llvm/include/llvm/Support/Caching.h +++ b/llvm/include/llvm/Support/Caching.h @@ -17,11 +17,10 @@ #include "llvm/Support/Compiler.h" #include "llvm/Support/Error.h" +#include "llvm/Support/MemoryBuffer.h" namespace llvm { -class MemoryBuffer; - /// This class wraps an output stream for a file. Most clients should just be /// able to return an instance of this base class from the stream callback, but /// if a client needs to perform some action after the stream is written to, diff --git a/llvm/include/llvm/Support/DebugLog.h b/llvm/include/llvm/Support/DebugLog.h index 7025ca14..fd67d7a 100644 --- a/llvm/include/llvm/Support/DebugLog.h +++ b/llvm/include/llvm/Support/DebugLog.h @@ -221,12 +221,10 @@ constexpr ::llvm::StringRef strip_quotes(const char *Str) { #define LDBG_GET_DEBUG_TYPE_STR() LDBG_GET_DEBUG_TYPE_STR_(DEBUG_TYPE) /// Helper to call isCurrentDebugType with a StringRef. -static LLVM_ATTRIBUTE_UNUSED bool ldbgIsCurrentDebugType(StringRef Type, - int Level) { +[[maybe_unused]] static bool ldbgIsCurrentDebugType(StringRef Type, int Level) { return ::llvm::isCurrentDebugType(Type.str().c_str(), Level); } -static LLVM_ATTRIBUTE_UNUSED bool ldbgIsCurrentDebugType(int Level, - StringRef Type) { +[[maybe_unused]] static bool ldbgIsCurrentDebugType(int Level, StringRef Type) { return ::llvm::isCurrentDebugType(Type.str().c_str(), Level); } @@ -302,7 +300,7 @@ public: }; /// Remove the path prefix from the file name. -static LLVM_ATTRIBUTE_UNUSED constexpr const char * +[[maybe_unused]] static constexpr const char * getShortFileName(const char *path) { const char *filename = path; for (const char *p = path; *p != '\0'; ++p) { @@ -315,7 +313,7 @@ getShortFileName(const char *path) { /// Compute the prefix for the debug log in the form of: /// "[DebugType] File:Line " /// Where the File is the file name without the path prefix. -static LLVM_ATTRIBUTE_UNUSED std::string +[[maybe_unused]] static std::string computePrefix(StringRef DebugType, const char *File, int Line, int Level) { std::string Prefix; raw_string_ostream OsPrefix(Prefix); @@ -326,7 +324,7 @@ computePrefix(StringRef DebugType, const char *File, int Line, int Level) { return OsPrefix.str(); } /// Overload allowing to swap the order of the DebugType and Level arguments. -static LLVM_ATTRIBUTE_UNUSED std::string +[[maybe_unused]] static std::string computePrefix(int Level, const char *File, int Line, StringRef DebugType) { return computePrefix(DebugType, File, Line, Level); } diff --git a/llvm/include/llvm/Support/SourceMgr.h b/llvm/include/llvm/Support/SourceMgr.h index 5637b64..8320006 100644 --- a/llvm/include/llvm/Support/SourceMgr.h +++ b/llvm/include/llvm/Support/SourceMgr.h @@ -15,6 +15,7 @@ #ifndef LLVM_SUPPORT_SOURCEMGR_H #define LLVM_SUPPORT_SOURCEMGR_H +#include "llvm/ADT/IntrusiveRefCntPtr.h" #include "llvm/ADT/SmallVector.h" #include "llvm/Support/Compiler.h" #include "llvm/Support/MemoryBuffer.h" @@ -23,6 +24,10 @@ namespace llvm { +namespace vfs { +class FileSystem; +} // end namespace vfs + class raw_ostream; class SMDiagnostic; class SMFixIt; @@ -91,15 +96,25 @@ private: DiagHandlerTy DiagHandler = nullptr; void *DiagContext = nullptr; + // Optional file system for finding include files. + IntrusiveRefCntPtr<vfs::FileSystem> FS; + bool isValidBufferID(unsigned i) const { return i && i <= Buffers.size(); } public: - SourceMgr() = default; + /// Create new source manager without support for include files. + SourceMgr(); + /// Create new source manager with the capability of finding include files + /// via the provided file system. + explicit SourceMgr(IntrusiveRefCntPtr<vfs::FileSystem> FS); SourceMgr(const SourceMgr &) = delete; SourceMgr &operator=(const SourceMgr &) = delete; - SourceMgr(SourceMgr &&) = default; - SourceMgr &operator=(SourceMgr &&) = default; - ~SourceMgr() = default; + SourceMgr(SourceMgr &&); + SourceMgr &operator=(SourceMgr &&); + ~SourceMgr(); + + IntrusiveRefCntPtr<vfs::FileSystem> getVirtualFileSystem() const; + void setVirtualFileSystem(IntrusiveRefCntPtr<vfs::FileSystem> FS); /// Return the include directories of this source manager. ArrayRef<std::string> getIncludeDirs() const { return IncludeDirectories; } diff --git a/llvm/include/llvm/TableGen/CodeGenHelpers.h b/llvm/include/llvm/TableGen/CodeGenHelpers.h index 5b823db..e22c6d4 100644 --- a/llvm/include/llvm/TableGen/CodeGenHelpers.h +++ b/llvm/include/llvm/TableGen/CodeGenHelpers.h @@ -34,20 +34,37 @@ private: raw_ostream &OS; }; +// Simple RAII helper for emitting header include guard (ifndef-define-endif). +class IncludeGuardEmitter { +public: + IncludeGuardEmitter(raw_ostream &OS, StringRef Name) + : Name(Name.str()), OS(OS) { + OS << "#ifndef " << Name << "\n" + << "#define " << Name << "\n\n"; + } + ~IncludeGuardEmitter() { OS << "\n#endif // " << Name << "\n"; } + +private: + std::string Name; + raw_ostream &OS; +}; + // Simple RAII helper for emitting namespace scope. Name can be a single -// namespace (empty for anonymous namespace) or nested namespace. +// namespace or nested namespace. If the name is empty, will not generate any +// namespace scope. class NamespaceEmitter { public: - NamespaceEmitter(raw_ostream &OS, StringRef Name) - : Name(trim(Name).str()), OS(OS) { - OS << "namespace " << this->Name << " {\n"; + NamespaceEmitter(raw_ostream &OS, StringRef NameUntrimmed) + : Name(trim(NameUntrimmed).str()), OS(OS) { + if (!Name.empty()) + OS << "namespace " << Name << " {\n"; } ~NamespaceEmitter() { close(); } // Explicit function to close the namespace scopes. void close() { - if (!Closed) + if (!Closed && !Name.empty()) OS << "} // namespace " << Name << "\n"; Closed = true; } diff --git a/llvm/include/llvm/TargetParser/X86TargetParser.def b/llvm/include/llvm/TargetParser/X86TargetParser.def index e62aa6d..254587b 100644 --- a/llvm/include/llvm/TargetParser/X86TargetParser.def +++ b/llvm/include/llvm/TargetParser/X86TargetParser.def @@ -115,6 +115,7 @@ X86_CPU_SUBTYPE_ALIAS(INTEL_COREI7_ALDERLAKE, "meteorlake") X86_CPU_SUBTYPE_ALIAS(INTEL_COREI7_SAPPHIRERAPIDS, "emeraldrapids") X86_CPU_SUBTYPE_ALIAS(INTEL_COREI7_ARROWLAKE_S,"lunarlake") X86_CPU_SUBTYPE_ALIAS(INTEL_COREI7_ALDERLAKE, "gracemont") +X86_CPU_SUBTYPE_ALIAS(INTEL_COREI7_PANTHERLAKE, "wildcatlake") #undef X86_CPU_SUBTYPE_ALIAS #undef X86_CPU_SUBTYPE diff --git a/llvm/include/llvm/TargetParser/X86TargetParser.h b/llvm/include/llvm/TargetParser/X86TargetParser.h index f6aeaad..e4c43cd 100644 --- a/llvm/include/llvm/TargetParser/X86TargetParser.h +++ b/llvm/include/llvm/TargetParser/X86TargetParser.h @@ -116,6 +116,7 @@ enum CPUKind { CK_ArrowlakeS, CK_Lunarlake, CK_Pantherlake, + CK_Wildcatlake, CK_Sierraforest, CK_Grandridge, CK_Graniterapids, diff --git a/llvm/lib/Analysis/InstructionSimplify.cpp b/llvm/lib/Analysis/InstructionSimplify.cpp index 4e38626..e08ef60 100644 --- a/llvm/lib/Analysis/InstructionSimplify.cpp +++ b/llvm/lib/Analysis/InstructionSimplify.cpp @@ -6644,7 +6644,7 @@ Value *llvm::simplifyBinaryIntrinsic(Intrinsic::ID IID, Type *ReturnType, "Invalid mask width"); // If index-width (mask size) is less than pointer-size then mask is // 1-extended. - if (match(Op1, m_PtrToInt(m_Specific(Op0)))) + if (match(Op1, m_PtrToIntOrAddr(m_Specific(Op0)))) return Op0; // NOTE: We may have attributes associated with the return value of the diff --git a/llvm/lib/Analysis/MLInlineAdvisor.cpp b/llvm/lib/Analysis/MLInlineAdvisor.cpp index f90717d..1d1a5560 100644 --- a/llvm/lib/Analysis/MLInlineAdvisor.cpp +++ b/llvm/lib/Analysis/MLInlineAdvisor.cpp @@ -61,6 +61,9 @@ static cl::opt<SkipMLPolicyCriteria> SkipPolicy( static cl::opt<std::string> ModelSelector("ml-inliner-model-selector", cl::Hidden, cl::init("")); +static cl::opt<bool> StopImmediatelyForTest("ml-inliner-stop-immediately", + cl::Hidden); + #if defined(LLVM_HAVE_TF_AOT_INLINERSIZEMODEL) // codegen-ed file #include "InlinerSizeModel.h" // NOLINT @@ -214,6 +217,7 @@ MLInlineAdvisor::MLInlineAdvisor( return; } ModelRunner->switchContext(""); + ForceStop = StopImmediatelyForTest; } unsigned MLInlineAdvisor::getInitialFunctionLevel(const Function &F) const { @@ -379,9 +383,17 @@ std::unique_ptr<InlineAdvice> MLInlineAdvisor::getAdviceImpl(CallBase &CB) { auto &ORE = FAM.getResult<OptimizationRemarkEmitterAnalysis>(Caller); if (SkipPolicy == SkipMLPolicyCriteria::IfCallerIsNotCold) { - if (!PSI.isFunctionEntryCold(&Caller)) - return std::make_unique<InlineAdvice>(this, CB, ORE, - GetDefaultAdvice(CB)); + if (!PSI.isFunctionEntryCold(&Caller)) { + // Return a MLInlineAdvice, despite delegating to the default advice, + // because we need to keep track of the internal state. This is different + // from the other instances where we return a "default" InlineAdvice, + // which happen at points we won't come back to the MLAdvisor for + // decisions requiring that state. + return ForceStop ? std::make_unique<InlineAdvice>(this, CB, ORE, + GetDefaultAdvice(CB)) + : std::make_unique<MLInlineAdvice>(this, CB, ORE, + GetDefaultAdvice(CB)); + } } auto MandatoryKind = InlineAdvisor::getMandatoryKind(CB, FAM, ORE); // If this is a "never inline" case, there won't be any changes to internal diff --git a/llvm/lib/Analysis/MemorySSA.cpp b/llvm/lib/Analysis/MemorySSA.cpp index ab37338..0b2e3fc 100644 --- a/llvm/lib/Analysis/MemorySSA.cpp +++ b/llvm/lib/Analysis/MemorySSA.cpp @@ -393,7 +393,7 @@ static bool isUseTriviallyOptimizableToLiveOnEntry(AliasAnalysisType &AA, /// \param AA The AliasAnalysis we used for our search. /// \param AllowImpreciseClobber Always false, unless we do relaxed verify. -LLVM_ATTRIBUTE_UNUSED static void +[[maybe_unused]] static void checkClobberSanity(const MemoryAccess *Start, MemoryAccess *ClobberAt, const MemoryLocation &StartLoc, const MemorySSA &MSSA, const UpwardsMemoryQuery &Query, BatchAAResults &AA, diff --git a/llvm/lib/Analysis/ScalarEvolution.cpp b/llvm/lib/Analysis/ScalarEvolution.cpp index 3fab6b0..a64b93d 100644 --- a/llvm/lib/Analysis/ScalarEvolution.cpp +++ b/llvm/lib/Analysis/ScalarEvolution.cpp @@ -6417,8 +6417,18 @@ APInt ScalarEvolution::getConstantMultipleImpl(const SCEV *S, case scSequentialUMinExpr: return GetGCDMultiple(cast<SCEVNAryExpr>(S)); case scUnknown: { - // ask ValueTracking for known bits + // Ask ValueTracking for known bits. SCEVUnknown only become available at + // the point their underlying IR instruction has been defined. If CtxI was + // not provided, use: + // * the first instruction in the entry block if it is an argument + // * the instruction itself otherwise. const SCEVUnknown *U = cast<SCEVUnknown>(S); + if (!CtxI) { + if (isa<Argument>(U->getValue())) + CtxI = &*F.getEntryBlock().begin(); + else if (auto *I = dyn_cast<Instruction>(U->getValue())) + CtxI = I; + } unsigned Known = computeKnownBits(U->getValue(), getDataLayout(), &AC, CtxI, &DT) .countMinTrailingZeros(); @@ -15761,6 +15771,21 @@ void ScalarEvolution::LoopGuards::collectFromBlock( const SCEV *OneAlignedUp = GetNextSCEVDividesByDivisor(One, DividesBy); To = SE.getUMaxExpr(FromRewritten, OneAlignedUp); + } else { + if (LHS->getType()->isPointerTy()) { + LHS = SE.getLosslessPtrToIntExpr(LHS); + RHS = SE.getLosslessPtrToIntExpr(RHS); + if (isa<SCEVCouldNotCompute>(LHS) || isa<SCEVCouldNotCompute>(RHS)) + break; + } + auto AddSubRewrite = [&](const SCEV *A, const SCEV *B) { + const SCEV *Sub = SE.getMinusSCEV(A, B); + AddRewrite(Sub, Sub, + SE.getUMaxExpr(Sub, SE.getOne(From->getType()))); + }; + AddSubRewrite(LHS, RHS); + AddSubRewrite(RHS, LHS); + continue; } break; default: diff --git a/llvm/lib/Analysis/StaticDataProfileInfo.cpp b/llvm/lib/Analysis/StaticDataProfileInfo.cpp index 1f751ee..61d4935 100644 --- a/llvm/lib/Analysis/StaticDataProfileInfo.cpp +++ b/llvm/lib/Analysis/StaticDataProfileInfo.cpp @@ -1,10 +1,14 @@ #include "llvm/Analysis/StaticDataProfileInfo.h" #include "llvm/Analysis/ProfileSummaryInfo.h" #include "llvm/IR/Constant.h" +#include "llvm/IR/Constants.h" #include "llvm/IR/GlobalVariable.h" +#include "llvm/IR/Module.h" #include "llvm/InitializePasses.h" #include "llvm/ProfileData/InstrProf.h" +#define DEBUG_TYPE "static-data-profile-info" + using namespace llvm; namespace llvm { @@ -60,6 +64,47 @@ void StaticDataProfileInfo::addConstantProfileCount( OriginalCount = getInstrMaxCountValue(); } +StaticDataProfileInfo::StaticDataHotness +StaticDataProfileInfo::getConstantHotnessUsingProfileCount( + const Constant *C, const ProfileSummaryInfo *PSI, uint64_t Count) const { + // The accummulated counter shows the constant is hot. Return enum 'hot' + // whether this variable is seen by unprofiled functions or not. + if (PSI->isHotCount(Count)) + return StaticDataHotness::Hot; + // The constant is not hot, and seen by unprofiled functions. We don't want to + // assign it to unlikely sections, even if the counter says 'cold'. So return + // enum 'LukewarmOrUnknown'. + if (ConstantWithoutCounts.count(C)) + return StaticDataHotness::LukewarmOrUnknown; + // The accummulated counter shows the constant is cold so return enum 'cold'. + if (PSI->isColdCount(Count)) + return StaticDataHotness::Cold; + + return StaticDataHotness::LukewarmOrUnknown; +} + +StaticDataProfileInfo::StaticDataHotness +StaticDataProfileInfo::getSectionHotnessUsingDataAccessProfile( + std::optional<StringRef> MaybeSectionPrefix) const { + if (!MaybeSectionPrefix) + return StaticDataHotness::LukewarmOrUnknown; + StringRef Prefix = *MaybeSectionPrefix; + assert((Prefix == "hot" || Prefix == "unlikely") && + "Expect section_prefix to be one of hot or unlikely"); + return Prefix == "hot" ? StaticDataHotness::Hot : StaticDataHotness::Cold; +} + +StringRef StaticDataProfileInfo::hotnessToStr(StaticDataHotness Hotness) const { + switch (Hotness) { + case StaticDataHotness::Cold: + return "unlikely"; + case StaticDataHotness::Hot: + return "hot"; + default: + return ""; + } +} + std::optional<uint64_t> StaticDataProfileInfo::getConstantProfileCount(const Constant *C) const { auto I = ConstantProfileCounts.find(C); @@ -70,27 +115,67 @@ StaticDataProfileInfo::getConstantProfileCount(const Constant *C) const { StringRef StaticDataProfileInfo::getConstantSectionPrefix( const Constant *C, const ProfileSummaryInfo *PSI) const { - auto Count = getConstantProfileCount(C); + std::optional<uint64_t> Count = getConstantProfileCount(C); + +#ifndef NDEBUG + auto DbgPrintPrefix = [](StringRef Prefix) { + return Prefix.empty() ? "<empty>" : Prefix; + }; +#endif + + if (EnableDataAccessProf) { + // Module flag `HasDataAccessProf` is 1 -> empty section prefix means + // unknown hotness except for string literals. + if (const GlobalVariable *GV = dyn_cast<GlobalVariable>(C); + GV && llvm::memprof::IsAnnotationOK(*GV) && + !GV->getName().starts_with(".str")) { + auto HotnessFromDataAccessProf = + getSectionHotnessUsingDataAccessProfile(GV->getSectionPrefix()); + + if (!Count) { + StringRef Prefix = hotnessToStr(HotnessFromDataAccessProf); + LLVM_DEBUG(dbgs() << GV->getName() << " has section prefix " + << DbgPrintPrefix(Prefix) + << ", solely from data access profiles\n"); + return Prefix; + } + + // Both data access profiles and PGO counters are available. Use the + // hotter one. + auto HotnessFromPGO = getConstantHotnessUsingProfileCount(C, PSI, *Count); + StaticDataHotness GlobalVarHotness = StaticDataHotness::LukewarmOrUnknown; + if (HotnessFromDataAccessProf == StaticDataHotness::Hot || + HotnessFromPGO == StaticDataHotness::Hot) { + GlobalVarHotness = StaticDataHotness::Hot; + } else if (HotnessFromDataAccessProf == + StaticDataHotness::LukewarmOrUnknown || + HotnessFromPGO == StaticDataHotness::LukewarmOrUnknown) { + GlobalVarHotness = StaticDataHotness::LukewarmOrUnknown; + } else { + GlobalVarHotness = StaticDataHotness::Cold; + } + StringRef Prefix = hotnessToStr(GlobalVarHotness); + LLVM_DEBUG( + dbgs() << GV->getName() << " has section prefix " + << DbgPrintPrefix(Prefix) + << ", the max from data access profiles as " + << DbgPrintPrefix(hotnessToStr(HotnessFromDataAccessProf)) + << " and PGO counters as " + << DbgPrintPrefix(hotnessToStr(HotnessFromPGO)) << "\n"); + return Prefix; + } + } if (!Count) return ""; - // The accummulated counter shows the constant is hot. Return 'hot' whether - // this variable is seen by unprofiled functions or not. - if (PSI->isHotCount(*Count)) - return "hot"; - // The constant is not hot, and seen by unprofiled functions. We don't want to - // assign it to unlikely sections, even if the counter says 'cold'. So return - // an empty prefix before checking whether the counter is cold. - if (ConstantWithoutCounts.count(C)) - return ""; - // The accummulated counter shows the constant is cold. Return 'unlikely'. - if (PSI->isColdCount(*Count)) - return "unlikely"; - // The counter says lukewarm. Return an empty prefix. - return ""; + return hotnessToStr(getConstantHotnessUsingProfileCount(C, PSI, *Count)); } bool StaticDataProfileInfoWrapperPass::doInitialization(Module &M) { - Info.reset(new StaticDataProfileInfo()); + bool EnableDataAccessProf = false; + if (auto *MD = mdconst::extract_or_null<ConstantInt>( + M.getModuleFlag("EnableDataAccessProf"))) + EnableDataAccessProf = MD->getZExtValue(); + Info.reset(new StaticDataProfileInfo(EnableDataAccessProf)); return false; } diff --git a/llvm/lib/AsmParser/LLParser.cpp b/llvm/lib/AsmParser/LLParser.cpp index 380b192..cf63285 100644 --- a/llvm/lib/AsmParser/LLParser.cpp +++ b/llvm/lib/AsmParser/LLParser.cpp @@ -329,10 +329,6 @@ bool LLParser::validateEndOfModule(bool UpgradeDebugInfo) { for (const auto &[Name, Info] : make_early_inc_range(ForwardRefVals)) { if (StringRef(Name).starts_with("llvm.")) { Intrinsic::ID IID = Intrinsic::lookupIntrinsicID(Name); - if (IID == Intrinsic::not_intrinsic) - // Don't do anything for unknown intrinsics. - continue; - // Automatically create declarations for intrinsics. Intrinsics can only // be called directly, so the call function type directly determines the // declaration function type. @@ -346,11 +342,26 @@ bool LLParser::validateEndOfModule(bool UpgradeDebugInfo) { return error(Info.second, "intrinsic can only be used as callee"); SmallVector<Type *> OverloadTys; - if (!Intrinsic::getIntrinsicSignature(IID, CB->getFunctionType(), - OverloadTys)) - return error(Info.second, "invalid intrinsic signature"); - - U.set(Intrinsic::getOrInsertDeclaration(M, IID, OverloadTys)); + if (IID != Intrinsic::not_intrinsic && + Intrinsic::getIntrinsicSignature(IID, CB->getFunctionType(), + OverloadTys)) { + U.set(Intrinsic::getOrInsertDeclaration(M, IID, OverloadTys)); + } else { + // Try to upgrade the intrinsic. + Function *TmpF = Function::Create(CB->getFunctionType(), + Function::ExternalLinkage, Name, M); + Function *NewF = nullptr; + if (!UpgradeIntrinsicFunction(TmpF, NewF)) { + if (IID == Intrinsic::not_intrinsic) + return error(Info.second, "unknown intrinsic '" + Name + "'"); + return error(Info.second, "invalid intrinsic signature"); + } + + U.set(TmpF); + UpgradeIntrinsicCall(CB, NewF); + if (TmpF->use_empty()) + TmpF->eraseFromParent(); + } } Info.first->eraseFromParent(); @@ -1259,7 +1270,7 @@ bool LLParser::parseAliasOrIFunc(const std::string &Name, unsigned NameID, if (parseToken(lltok::StringConstant, "expected partition string")) return true; } else if (!IsAlias && Lex.getKind() == lltok::MetadataVar) { - if (parseGlobalObjectMetadataAttachment(*GI.get())) + if (parseGlobalObjectMetadataAttachment(*GI)) return true; } else { return tokError("unknown alias or ifunc property!"); @@ -5865,6 +5876,7 @@ bool LLParser::parseDICompileUnit(MDNode *&Result, bool IsDistinct) { REQUIRED(file, MDField, (/* AllowNull */ false)); \ OPTIONAL(language, DwarfLangField, ); \ OPTIONAL(sourceLanguageName, DwarfSourceLangNameField, ); \ + OPTIONAL(sourceLanguageVersion, MDUnsignedField, (0, UINT32_MAX)); \ OPTIONAL(producer, MDStringField, ); \ OPTIONAL(isOptimized, MDBoolField, ); \ OPTIONAL(flags, MDStringField, ); \ @@ -5894,10 +5906,15 @@ bool LLParser::parseDICompileUnit(MDNode *&Result, bool IsDistinct) { return error(Loc, "can only specify one of 'language' and " "'sourceLanguageName' on !DICompileUnit"); + if (sourceLanguageVersion.Seen && !sourceLanguageName.Seen) + return error(Loc, "'sourceLanguageVersion' requires an associated " + "'sourceLanguageName' on !DICompileUnit"); + Result = DICompileUnit::getDistinct( Context, language.Seen ? DISourceLanguageName(language.Val) - : DISourceLanguageName(sourceLanguageName.Val, 0), + : DISourceLanguageName(sourceLanguageName.Val, + sourceLanguageVersion.Val), file.Val, producer.Val, isOptimized.Val, flags.Val, runtimeVersion.Val, splitDebugFilename.Val, emissionKind.Val, enums.Val, retainedTypes.Val, globals.Val, imports.Val, macros.Val, dwoId.Val, splitDebugInlining.Val, diff --git a/llvm/lib/BinaryFormat/XCOFF.cpp b/llvm/lib/BinaryFormat/XCOFF.cpp index e0a4471..19d5b98 100644 --- a/llvm/lib/BinaryFormat/XCOFF.cpp +++ b/llvm/lib/BinaryFormat/XCOFF.cpp @@ -112,26 +112,26 @@ StringRef XCOFF::getNameForTracebackTableLanguageId( XCOFF::CFileCpuId XCOFF::getCpuID(StringRef CPUName) { StringRef CPU = PPC::normalizeCPUName(CPUName); return StringSwitch<XCOFF::CFileCpuId>(CPU) - .Cases("generic", "COM", XCOFF::TCPU_COM) + .Cases({"generic", "COM"}, XCOFF::TCPU_COM) .Case("601", XCOFF::TCPU_601) - .Cases("602", "603", "603e", "603ev", XCOFF::TCPU_603) - .Cases("604", "604e", XCOFF::TCPU_604) + .Cases({"602", "603", "603e", "603ev"}, XCOFF::TCPU_603) + .Cases({"604", "604e"}, XCOFF::TCPU_604) .Case("620", XCOFF::TCPU_620) .Case("970", XCOFF::TCPU_970) - .Cases("a2", "g3", "g4", "g5", "e500", XCOFF::TCPU_COM) - .Cases("pwr3", "pwr4", XCOFF::TCPU_COM) - .Cases("pwr5", "PWR5", XCOFF::TCPU_PWR5) - .Cases("pwr5x", "PWR5X", XCOFF::TCPU_PWR5X) - .Cases("pwr6", "PWR6", XCOFF::TCPU_PWR6) - .Cases("pwr6x", "PWR6E", XCOFF::TCPU_PWR6E) - .Cases("pwr7", "PWR7", XCOFF::TCPU_PWR7) - .Cases("pwr8", "PWR8", XCOFF::TCPU_PWR8) - .Cases("pwr9", "PWR9", XCOFF::TCPU_PWR9) - .Cases("pwr10", "PWR10", XCOFF::TCPU_PWR10) - .Cases("ppc", "PPC", "ppc32", "ppc64", XCOFF::TCPU_COM) + .Cases({"a2", "g3", "g4", "g5", "e500"}, XCOFF::TCPU_COM) + .Cases({"pwr3", "pwr4"}, XCOFF::TCPU_COM) + .Cases({"pwr5", "PWR5"}, XCOFF::TCPU_PWR5) + .Cases({"pwr5x", "PWR5X"}, XCOFF::TCPU_PWR5X) + .Cases({"pwr6", "PWR6"}, XCOFF::TCPU_PWR6) + .Cases({"pwr6x", "PWR6E"}, XCOFF::TCPU_PWR6E) + .Cases({"pwr7", "PWR7"}, XCOFF::TCPU_PWR7) + .Cases({"pwr8", "PWR8"}, XCOFF::TCPU_PWR8) + .Cases({"pwr9", "PWR9"}, XCOFF::TCPU_PWR9) + .Cases({"pwr10", "PWR10"}, XCOFF::TCPU_PWR10) + .Cases({"ppc", "PPC", "ppc32", "ppc64"}, XCOFF::TCPU_COM) .Case("ppc64le", XCOFF::TCPU_PWR8) .Case("future", XCOFF::TCPU_PWR10) - .Cases("any", "ANY", XCOFF::TCPU_ANY) + .Cases({"any", "ANY"}, XCOFF::TCPU_ANY) .Default(XCOFF::TCPU_INVALID); } diff --git a/llvm/lib/Bitcode/Reader/MetadataLoader.cpp b/llvm/lib/Bitcode/Reader/MetadataLoader.cpp index cdcf7a8..ed0443f 100644 --- a/llvm/lib/Bitcode/Reader/MetadataLoader.cpp +++ b/llvm/lib/Bitcode/Reader/MetadataLoader.cpp @@ -1860,7 +1860,7 @@ Error MetadataLoader::MetadataLoaderImpl::parseOneMetadata( break; } case bitc::METADATA_COMPILE_UNIT: { - if (Record.size() < 14 || Record.size() > 22) + if (Record.size() < 14 || Record.size() > 23) return error("Invalid record"); // Ignore Record[0], which indicates whether this compile unit is @@ -1869,11 +1869,13 @@ Error MetadataLoader::MetadataLoaderImpl::parseOneMetadata( const auto LangVersionMask = (uint64_t(1) << 63); const bool HasVersionedLanguage = Record[1] & LangVersionMask; + const uint32_t LanguageVersion = Record.size() > 22 ? Record[22] : 0; auto *CU = DICompileUnit::getDistinct( Context, HasVersionedLanguage - ? DISourceLanguageName(Record[1] & ~LangVersionMask, 0) + ? DISourceLanguageName(Record[1] & ~LangVersionMask, + LanguageVersion) : DISourceLanguageName(Record[1]), getMDOrNull(Record[2]), getMDString(Record[3]), Record[4], getMDString(Record[5]), Record[6], getMDString(Record[7]), Record[8], diff --git a/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp b/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp index 54e916e..8ff3aa9 100644 --- a/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp +++ b/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp @@ -2142,6 +2142,7 @@ void ModuleBitcodeWriter::writeDICompileUnit(const DICompileUnit *N, Record.push_back(N->getRangesBaseAddress()); Record.push_back(VE.getMetadataOrNullID(N->getRawSysRoot())); Record.push_back(VE.getMetadataOrNullID(N->getRawSDK())); + Record.push_back(Lang.hasVersionedName() ? Lang.getVersion() : 0); Stream.EmitRecord(bitc::METADATA_COMPILE_UNIT, Record, Abbrev); Record.clear(); diff --git a/llvm/lib/CAS/OnDiskTrieRawHashMap.cpp b/llvm/lib/CAS/OnDiskTrieRawHashMap.cpp index 323b21e..4e6f93e 100644 --- a/llvm/lib/CAS/OnDiskTrieRawHashMap.cpp +++ b/llvm/lib/CAS/OnDiskTrieRawHashMap.cpp @@ -1102,8 +1102,6 @@ void TrieRawHashMapHandle::print( if (auto Err = Printer.printRecords()) OS << "error: " << toString(std::move(Err)) << "\n"; - - return; } Error TrieRawHashMapHandle::validate( diff --git a/llvm/lib/CodeGen/AsmPrinter/AsmPrinter.cpp b/llvm/lib/CodeGen/AsmPrinter/AsmPrinter.cpp index 219bbc9..e2af0c5 100644 --- a/llvm/lib/CodeGen/AsmPrinter/AsmPrinter.cpp +++ b/llvm/lib/CodeGen/AsmPrinter/AsmPrinter.cpp @@ -119,6 +119,7 @@ #include "llvm/Support/MathExtras.h" #include "llvm/Support/Path.h" #include "llvm/Support/VCSRevision.h" +#include "llvm/Support/VirtualFileSystem.h" #include "llvm/Support/raw_ostream.h" #include "llvm/Target/TargetLoweringObjectFile.h" #include "llvm/Target/TargetMachine.h" @@ -476,6 +477,7 @@ void AsmPrinter::getAnalysisUsage(AnalysisUsage &AU) const { } bool AsmPrinter::doInitialization(Module &M) { + VFS = vfs::getRealFileSystem(); auto *MMIWP = getAnalysisIfAvailable<MachineModuleInfoWrapperPass>(); MMI = MMIWP ? &MMIWP->getMMI() : nullptr; HasSplitStack = false; @@ -1437,7 +1439,8 @@ getBBAddrMapFeature(const MachineFunction &MF, int NumMBBSectionRanges, BrProbEnabled, MF.hasBBSections() && NumMBBSectionRanges > 1, static_cast<bool>(BBAddrMapSkipEmitBBEntries), - HasCalls}; + HasCalls, + false}; } void AsmPrinter::emitBBAddrMapSection(const MachineFunction &MF) { @@ -1682,7 +1685,7 @@ static ConstantInt *extractNumericCGTypeId(const Function &F) { return nullptr; } -/// Emits .callgraph section. +/// Emits .llvm.callgraph section. void AsmPrinter::emitCallGraphSection(const MachineFunction &MF, FunctionCallGraphInfo &FuncCGInfo) { if (!MF.getTarget().Options.EmitCallGraphSection) diff --git a/llvm/lib/CodeGen/AsmPrinter/AsmPrinterInlineAsm.cpp b/llvm/lib/CodeGen/AsmPrinter/AsmPrinterInlineAsm.cpp index c364ffc..8dd8b9da 100644 --- a/llvm/lib/CodeGen/AsmPrinter/AsmPrinterInlineAsm.cpp +++ b/llvm/lib/CodeGen/AsmPrinter/AsmPrinterInlineAsm.cpp @@ -36,6 +36,7 @@ #include "llvm/Support/ErrorHandling.h" #include "llvm/Support/MemoryBuffer.h" #include "llvm/Support/SourceMgr.h" +#include "llvm/Support/VirtualFileSystem.h" #include "llvm/Support/raw_ostream.h" #include "llvm/Target/TargetMachine.h" using namespace llvm; @@ -98,6 +99,7 @@ void AsmPrinter::emitInlineAsm(StringRef Str, const MCSubtargetInfo &STI, unsigned BufNum = addInlineAsmDiagBuffer(Str, LocMDNode); SourceMgr &SrcMgr = *MMI->getContext().getInlineSourceManager(); SrcMgr.setIncludeDirs(MCOptions.IASSearchPaths); + SrcMgr.setVirtualFileSystem(VFS); std::unique_ptr<MCAsmParser> Parser( createMCAsmParser(SrcMgr, OutContext, *OutStreamer, *MAI, BufNum)); diff --git a/llvm/lib/CodeGen/AtomicExpandPass.cpp b/llvm/lib/CodeGen/AtomicExpandPass.cpp index 4931403..53f1cfe2 100644 --- a/llvm/lib/CodeGen/AtomicExpandPass.cpp +++ b/llvm/lib/CodeGen/AtomicExpandPass.cpp @@ -770,7 +770,7 @@ struct PartwordMaskValues { Value *Inv_Mask = nullptr; }; -LLVM_ATTRIBUTE_UNUSED +[[maybe_unused]] raw_ostream &operator<<(raw_ostream &O, const PartwordMaskValues &PMV) { auto PrintObj = [&O](auto *V) { if (V) diff --git a/llvm/lib/CodeGen/BranchRelaxation.cpp b/llvm/lib/CodeGen/BranchRelaxation.cpp index 2d50167..fae952e 100644 --- a/llvm/lib/CodeGen/BranchRelaxation.cpp +++ b/llvm/lib/CodeGen/BranchRelaxation.cpp @@ -491,6 +491,20 @@ bool BranchRelaxation::fixupConditionalBranch(MachineInstr &MI) { return true; } if (FBB) { + // If we get here with a MBB which ends like this: + // + // bb.1: + // successors: %bb.2; + // ... + // BNE $x1, $x0, %bb.2 + // PseudoBR %bb.2 + // + // Just remove conditional branch. + if (TBB == FBB) { + removeBranch(MBB); + insertUncondBranch(MBB, TBB); + return true; + } // We need to split the basic block here to obtain two long-range // unconditional branches. NewBB = createNewBlockAfter(*MBB); diff --git a/llvm/lib/CodeGen/CodeGenPrepare.cpp b/llvm/lib/CodeGen/CodeGenPrepare.cpp index 4320b1d..9e78ec9 100644 --- a/llvm/lib/CodeGen/CodeGenPrepare.cpp +++ b/llvm/lib/CodeGen/CodeGenPrepare.cpp @@ -819,7 +819,7 @@ void CodeGenPrepare::removeAllAssertingVHReferences(Value *V) { } // Verify BFI has been updated correctly by recomputing BFI and comparing them. -void LLVM_ATTRIBUTE_UNUSED CodeGenPrepare::verifyBFIUpdates(Function &F) { +[[maybe_unused]] void CodeGenPrepare::verifyBFIUpdates(Function &F) { DominatorTree NewDT(F); LoopInfo NewLI(NewDT); BranchProbabilityInfo NewBPI(F, NewLI, TLInfo); diff --git a/llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp b/llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp index 9b9a6a2..04d9309 100644 --- a/llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp +++ b/llvm/lib/CodeGen/GlobalISel/GISelValueTracking.cpp @@ -112,7 +112,7 @@ APInt GISelValueTracking::getKnownOnes(Register R) { return getKnownBits(R).One; } -LLVM_ATTRIBUTE_UNUSED static void +[[maybe_unused]] static void dumpResult(const MachineInstr &MI, const KnownBits &Known, unsigned Depth) { dbgs() << "[" << Depth << "] Compute known bits: " << MI << "[" << Depth << "] Computed for: " << MI << "[" << Depth << "] Known: 0x" diff --git a/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp b/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp index cffaf7c..38ec83f 100644 --- a/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp +++ b/llvm/lib/CodeGen/GlobalISel/LegalizerHelper.cpp @@ -3292,8 +3292,7 @@ LegalizerHelper::widenScalar(MachineInstr &MI, unsigned TypeIdx, LLT WideTy) { if (TypeIdx != 2) return UnableToLegalize; Observer.changingInstr(MI); - // TODO: Probably should be zext - widenScalarSrc(MI, WideTy, 2, TargetOpcode::G_SEXT); + widenScalarSrc(MI, WideTy, 2, TargetOpcode::G_ZEXT); Observer.changedInstr(MI); return Legalized; } @@ -3325,8 +3324,7 @@ LegalizerHelper::widenScalar(MachineInstr &MI, unsigned TypeIdx, LLT WideTy) { if (TypeIdx == 2) { Observer.changingInstr(MI); - // TODO: Probably should be zext - widenScalarSrc(MI, WideTy, 3, TargetOpcode::G_SEXT); + widenScalarSrc(MI, WideTy, 3, TargetOpcode::G_ZEXT); Observer.changedInstr(MI); return Legalized; } diff --git a/llvm/lib/CodeGen/LiveIntervals.cpp b/llvm/lib/CodeGen/LiveIntervals.cpp index 0e38017..d2f2c3e 100644 --- a/llvm/lib/CodeGen/LiveIntervals.cpp +++ b/llvm/lib/CodeGen/LiveIntervals.cpp @@ -661,7 +661,10 @@ void LiveIntervals::extendToIndices(LiveRange &LR, void LiveIntervals::pruneValue(LiveRange &LR, SlotIndex Kill, SmallVectorImpl<SlotIndex> *EndPoints) { LiveQueryResult LRQ = LR.Query(Kill); - VNInfo *VNI = LRQ.valueOutOrDead(); + // LR may have liveness reachable from early clobber slot, which may be + // only live-in instead of live-out of the instruction. + // For example, LR =[1r, 3r), Kill = 3e, we have to prune [3e, 3r) of LR. + VNInfo *VNI = LRQ.valueOutOrDead() ? LRQ.valueOutOrDead() : LRQ.valueIn(); if (!VNI) return; diff --git a/llvm/lib/CodeGen/MachineCopyPropagation.cpp b/llvm/lib/CodeGen/MachineCopyPropagation.cpp index e359831..ea08365 100644 --- a/llvm/lib/CodeGen/MachineCopyPropagation.cpp +++ b/llvm/lib/CodeGen/MachineCopyPropagation.cpp @@ -1257,7 +1257,7 @@ void MachineCopyPropagation::BackwardCopyPropagateBlock( Tracker.clear(); } -static void LLVM_ATTRIBUTE_UNUSED printSpillReloadChain( +[[maybe_unused]] static void printSpillReloadChain( DenseMap<MachineInstr *, SmallVector<MachineInstr *>> &SpillChain, DenseMap<MachineInstr *, SmallVector<MachineInstr *>> &ReloadChain, MachineInstr *Leader) { diff --git a/llvm/lib/CodeGen/ScheduleDAGInstrs.cpp b/llvm/lib/CodeGen/ScheduleDAGInstrs.cpp index 3268c26..9662511 100644 --- a/llvm/lib/CodeGen/ScheduleDAGInstrs.cpp +++ b/llvm/lib/CodeGen/ScheduleDAGInstrs.cpp @@ -1551,7 +1551,7 @@ LLVM_DUMP_METHOD void ILPValue::dump() const { dbgs() << *this << '\n'; } -LLVM_ATTRIBUTE_UNUSED +[[maybe_unused]] raw_ostream &llvm::operator<<(raw_ostream &OS, const ILPValue &Val) { Val.print(OS); return OS; diff --git a/llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp b/llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp index 787a81a..c97300d 100644 --- a/llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp +++ b/llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp @@ -658,13 +658,13 @@ namespace { bool InexpensiveOnly = false, std::optional<EVT> OutVT = std::nullopt); SDValue BuildDivEstimate(SDValue N, SDValue Op, SDNodeFlags Flags); - SDValue buildRsqrtEstimate(SDValue Op, SDNodeFlags Flags); - SDValue buildSqrtEstimate(SDValue Op, SDNodeFlags Flags); - SDValue buildSqrtEstimateImpl(SDValue Op, SDNodeFlags Flags, bool Recip); + SDValue buildRsqrtEstimate(SDValue Op); + SDValue buildSqrtEstimate(SDValue Op); + SDValue buildSqrtEstimateImpl(SDValue Op, bool Recip); SDValue buildSqrtNROneConst(SDValue Arg, SDValue Est, unsigned Iterations, - SDNodeFlags Flags, bool Reciprocal); + bool Reciprocal); SDValue buildSqrtNRTwoConst(SDValue Arg, SDValue Est, unsigned Iterations, - SDNodeFlags Flags, bool Reciprocal); + bool Reciprocal); SDValue MatchBSwapHWordLow(SDNode *N, SDValue N0, SDValue N1, bool DemandHighBits = true); SDValue MatchBSwapHWord(SDNode *N, SDValue N0, SDValue N1); @@ -17759,7 +17759,6 @@ SDValue DAGCombiner::visitFADD(SDNode *N) { bool N1CFP = DAG.isConstantFPBuildVectorOrConstantFP(N1); EVT VT = N->getValueType(0); SDLoc DL(N); - const TargetOptions &Options = DAG.getTarget().Options; SDNodeFlags Flags = N->getFlags(); SelectionDAG::FlagInserter FlagsInserter(DAG, N); @@ -17825,7 +17824,7 @@ SDValue DAGCombiner::visitFADD(SDNode *N) { bool AllowNewConst = (Level < AfterLegalizeDAG); // If nnan is enabled, fold lots of things. - if ((Options.NoNaNsFPMath || Flags.hasNoNaNs()) && AllowNewConst) { + if (Flags.hasNoNaNs() && AllowNewConst) { // If allowed, fold (fadd (fneg x), x) -> 0.0 if (N0.getOpcode() == ISD::FNEG && N0.getOperand(0) == N1) return DAG.getConstantFP(0.0, DL, VT); @@ -17974,7 +17973,6 @@ SDValue DAGCombiner::visitFSUB(SDNode *N) { ConstantFPSDNode *N1CFP = isConstOrConstSplatFP(N1, true); EVT VT = N->getValueType(0); SDLoc DL(N); - const TargetOptions &Options = DAG.getTarget().Options; const SDNodeFlags Flags = N->getFlags(); SelectionDAG::FlagInserter FlagsInserter(DAG, N); @@ -18002,7 +18000,7 @@ SDValue DAGCombiner::visitFSUB(SDNode *N) { if (N0 == N1) { // (fsub x, x) -> 0.0 - if (Options.NoNaNsFPMath || Flags.hasNoNaNs()) + if (Flags.hasNoNaNs()) return DAG.getConstantFP(0.0f, DL, VT); } @@ -18313,7 +18311,6 @@ template <class MatchContextClass> SDValue DAGCombiner::visitFMA(SDNode *N) { ConstantFPSDNode *N2CFP = dyn_cast<ConstantFPSDNode>(N2); EVT VT = N->getValueType(0); SDLoc DL(N); - const TargetOptions &Options = DAG.getTarget().Options; // FMA nodes have flags that propagate to the created nodes. SelectionDAG::FlagInserter FlagsInserter(DAG, N); MatchContextClass matcher(DAG, TLI, N); @@ -18339,8 +18336,7 @@ template <class MatchContextClass> SDValue DAGCombiner::visitFMA(SDNode *N) { return matcher.getNode(ISD::FMA, DL, VT, NegN0, NegN1, N2); } - if ((Options.NoNaNsFPMath && N->getFlags().hasNoInfs()) || - (N->getFlags().hasNoNaNs() && N->getFlags().hasNoInfs())) { + if (N->getFlags().hasNoNaNs() && N->getFlags().hasNoInfs()) { if (N->getFlags().hasNoSignedZeros() || (N2CFP && !N2CFP->isExactlyValue(-0.0))) { if (N0CFP && N0CFP->isZero()) @@ -18590,20 +18586,18 @@ SDValue DAGCombiner::visitFDIV(SDNode *N) { // If this FDIV is part of a reciprocal square root, it may be folded // into a target-specific square root estimate instruction. if (N1.getOpcode() == ISD::FSQRT) { - if (SDValue RV = buildRsqrtEstimate(N1.getOperand(0), Flags)) + if (SDValue RV = buildRsqrtEstimate(N1.getOperand(0))) return DAG.getNode(ISD::FMUL, DL, VT, N0, RV); } else if (N1.getOpcode() == ISD::FP_EXTEND && N1.getOperand(0).getOpcode() == ISD::FSQRT) { - if (SDValue RV = - buildRsqrtEstimate(N1.getOperand(0).getOperand(0), Flags)) { + if (SDValue RV = buildRsqrtEstimate(N1.getOperand(0).getOperand(0))) { RV = DAG.getNode(ISD::FP_EXTEND, SDLoc(N1), VT, RV); AddToWorklist(RV.getNode()); return DAG.getNode(ISD::FMUL, DL, VT, N0, RV); } } else if (N1.getOpcode() == ISD::FP_ROUND && N1.getOperand(0).getOpcode() == ISD::FSQRT) { - if (SDValue RV = - buildRsqrtEstimate(N1.getOperand(0).getOperand(0), Flags)) { + if (SDValue RV = buildRsqrtEstimate(N1.getOperand(0).getOperand(0))) { RV = DAG.getNode(ISD::FP_ROUND, SDLoc(N1), VT, RV, N1.getOperand(1)); AddToWorklist(RV.getNode()); return DAG.getNode(ISD::FMUL, DL, VT, N0, RV); @@ -18635,7 +18629,7 @@ SDValue DAGCombiner::visitFDIV(SDNode *N) { SDValue AA = DAG.getNode(ISD::FMUL, DL, VT, A, A); SDValue AAZ = DAG.getNode(ISD::FMUL, DL, VT, AA, Sqrt.getOperand(0)); - if (SDValue Rsqrt = buildRsqrtEstimate(AAZ, Flags)) + if (SDValue Rsqrt = buildRsqrtEstimate(AAZ)) return DAG.getNode(ISD::FMUL, DL, VT, N0, Rsqrt); // Estimate creation failed. Clean up speculatively created nodes. @@ -18645,7 +18639,7 @@ SDValue DAGCombiner::visitFDIV(SDNode *N) { // We found a FSQRT, so try to make this fold: // X / (Y * sqrt(Z)) -> X * (rsqrt(Z) / Y) - if (SDValue Rsqrt = buildRsqrtEstimate(Sqrt.getOperand(0), Flags)) { + if (SDValue Rsqrt = buildRsqrtEstimate(Sqrt.getOperand(0))) { SDValue Div = DAG.getNode(ISD::FDIV, SDLoc(N1), VT, Rsqrt, Y); AddToWorklist(Div.getNode()); return DAG.getNode(ISD::FMUL, DL, VT, N0, Div); @@ -18742,11 +18736,12 @@ SDValue DAGCombiner::visitFSQRT(SDNode *N) { return SDValue(); // FSQRT nodes have flags that propagate to the created nodes. + SelectionDAG::FlagInserter FlagInserter(DAG, Flags); // TODO: If this is N0/sqrt(N0), and we reach this node before trying to // transform the fdiv, we may produce a sub-optimal estimate sequence // because the reciprocal calculation may not have to filter out a // 0.0 input. - return buildSqrtEstimate(N0, Flags); + return buildSqrtEstimate(N0); } /// copysign(x, fp_extend(y)) -> copysign(x, y) @@ -29743,28 +29738,27 @@ SDValue DAGCombiner::BuildDivEstimate(SDValue N, SDValue Op, /// X_{i+1} = X_i (1.5 - A X_i^2 / 2) /// As a result, we precompute A/2 prior to the iteration loop. SDValue DAGCombiner::buildSqrtNROneConst(SDValue Arg, SDValue Est, - unsigned Iterations, - SDNodeFlags Flags, bool Reciprocal) { + unsigned Iterations, bool Reciprocal) { EVT VT = Arg.getValueType(); SDLoc DL(Arg); SDValue ThreeHalves = DAG.getConstantFP(1.5, DL, VT); // We now need 0.5 * Arg which we can write as (1.5 * Arg - Arg) so that // this entire sequence requires only one FP constant. - SDValue HalfArg = DAG.getNode(ISD::FMUL, DL, VT, ThreeHalves, Arg, Flags); - HalfArg = DAG.getNode(ISD::FSUB, DL, VT, HalfArg, Arg, Flags); + SDValue HalfArg = DAG.getNode(ISD::FMUL, DL, VT, ThreeHalves, Arg); + HalfArg = DAG.getNode(ISD::FSUB, DL, VT, HalfArg, Arg); // Newton iterations: Est = Est * (1.5 - HalfArg * Est * Est) for (unsigned i = 0; i < Iterations; ++i) { - SDValue NewEst = DAG.getNode(ISD::FMUL, DL, VT, Est, Est, Flags); - NewEst = DAG.getNode(ISD::FMUL, DL, VT, HalfArg, NewEst, Flags); - NewEst = DAG.getNode(ISD::FSUB, DL, VT, ThreeHalves, NewEst, Flags); - Est = DAG.getNode(ISD::FMUL, DL, VT, Est, NewEst, Flags); + SDValue NewEst = DAG.getNode(ISD::FMUL, DL, VT, Est, Est); + NewEst = DAG.getNode(ISD::FMUL, DL, VT, HalfArg, NewEst); + NewEst = DAG.getNode(ISD::FSUB, DL, VT, ThreeHalves, NewEst); + Est = DAG.getNode(ISD::FMUL, DL, VT, Est, NewEst); } // If non-reciprocal square root is requested, multiply the result by Arg. if (!Reciprocal) - Est = DAG.getNode(ISD::FMUL, DL, VT, Est, Arg, Flags); + Est = DAG.getNode(ISD::FMUL, DL, VT, Est, Arg); return Est; } @@ -29775,8 +29769,7 @@ SDValue DAGCombiner::buildSqrtNROneConst(SDValue Arg, SDValue Est, /// => /// X_{i+1} = (-0.5 * X_i) * (A * X_i * X_i + (-3.0)) SDValue DAGCombiner::buildSqrtNRTwoConst(SDValue Arg, SDValue Est, - unsigned Iterations, - SDNodeFlags Flags, bool Reciprocal) { + unsigned Iterations, bool Reciprocal) { EVT VT = Arg.getValueType(); SDLoc DL(Arg); SDValue MinusThree = DAG.getConstantFP(-3.0, DL, VT); @@ -29789,9 +29782,9 @@ SDValue DAGCombiner::buildSqrtNRTwoConst(SDValue Arg, SDValue Est, // Newton iterations for reciprocal square root: // E = (E * -0.5) * ((A * E) * E + -3.0) for (unsigned i = 0; i < Iterations; ++i) { - SDValue AE = DAG.getNode(ISD::FMUL, DL, VT, Arg, Est, Flags); - SDValue AEE = DAG.getNode(ISD::FMUL, DL, VT, AE, Est, Flags); - SDValue RHS = DAG.getNode(ISD::FADD, DL, VT, AEE, MinusThree, Flags); + SDValue AE = DAG.getNode(ISD::FMUL, DL, VT, Arg, Est); + SDValue AEE = DAG.getNode(ISD::FMUL, DL, VT, AE, Est); + SDValue RHS = DAG.getNode(ISD::FADD, DL, VT, AEE, MinusThree); // When calculating a square root at the last iteration build: // S = ((A * E) * -0.5) * ((A * E) * E + -3.0) @@ -29799,13 +29792,13 @@ SDValue DAGCombiner::buildSqrtNRTwoConst(SDValue Arg, SDValue Est, SDValue LHS; if (Reciprocal || (i + 1) < Iterations) { // RSQRT: LHS = (E * -0.5) - LHS = DAG.getNode(ISD::FMUL, DL, VT, Est, MinusHalf, Flags); + LHS = DAG.getNode(ISD::FMUL, DL, VT, Est, MinusHalf); } else { // SQRT: LHS = (A * E) * -0.5 - LHS = DAG.getNode(ISD::FMUL, DL, VT, AE, MinusHalf, Flags); + LHS = DAG.getNode(ISD::FMUL, DL, VT, AE, MinusHalf); } - Est = DAG.getNode(ISD::FMUL, DL, VT, LHS, RHS, Flags); + Est = DAG.getNode(ISD::FMUL, DL, VT, LHS, RHS); } return Est; @@ -29814,8 +29807,7 @@ SDValue DAGCombiner::buildSqrtNRTwoConst(SDValue Arg, SDValue Est, /// Build code to calculate either rsqrt(Op) or sqrt(Op). In the latter case /// Op*rsqrt(Op) is actually computed, so additional postprocessing is needed if /// Op can be zero. -SDValue DAGCombiner::buildSqrtEstimateImpl(SDValue Op, SDNodeFlags Flags, - bool Reciprocal) { +SDValue DAGCombiner::buildSqrtEstimateImpl(SDValue Op, bool Reciprocal) { if (LegalDAG) return SDValue(); @@ -29843,8 +29835,8 @@ SDValue DAGCombiner::buildSqrtEstimateImpl(SDValue Op, SDNodeFlags Flags, if (Iterations > 0) Est = UseOneConstNR - ? buildSqrtNROneConst(Op, Est, Iterations, Flags, Reciprocal) - : buildSqrtNRTwoConst(Op, Est, Iterations, Flags, Reciprocal); + ? buildSqrtNROneConst(Op, Est, Iterations, Reciprocal) + : buildSqrtNRTwoConst(Op, Est, Iterations, Reciprocal); if (!Reciprocal) { SDLoc DL(Op); // Try the target specific test first. @@ -29862,12 +29854,12 @@ SDValue DAGCombiner::buildSqrtEstimateImpl(SDValue Op, SDNodeFlags Flags, return SDValue(); } -SDValue DAGCombiner::buildRsqrtEstimate(SDValue Op, SDNodeFlags Flags) { - return buildSqrtEstimateImpl(Op, Flags, true); +SDValue DAGCombiner::buildRsqrtEstimate(SDValue Op) { + return buildSqrtEstimateImpl(Op, true); } -SDValue DAGCombiner::buildSqrtEstimate(SDValue Op, SDNodeFlags Flags) { - return buildSqrtEstimateImpl(Op, Flags, false); +SDValue DAGCombiner::buildSqrtEstimate(SDValue Op) { + return buildSqrtEstimateImpl(Op, false); } /// Return true if there is any possibility that the two addresses overlap. diff --git a/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp b/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp index 876066e9..90edaf3 100644 --- a/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp +++ b/llvm/lib/CodeGen/SelectionDAG/SelectionDAG.cpp @@ -6405,8 +6405,9 @@ static SDValue foldCONCAT_VECTORS(const SDLoc &DL, EVT VT, if (VT.isScalableVector()) return SDValue(); - // A CONCAT_VECTOR with all UNDEF/BUILD_VECTOR operands can be - // simplified to one big BUILD_VECTOR. + // A CONCAT_VECTOR of scalar sources, such as UNDEF, BUILD_VECTOR and + // single-element INSERT_VECTOR_ELT operands can be simplified to one big + // BUILD_VECTOR. // FIXME: Add support for SCALAR_TO_VECTOR as well. EVT SVT = VT.getScalarType(); SmallVector<SDValue, 16> Elts; @@ -6416,6 +6417,10 @@ static SDValue foldCONCAT_VECTORS(const SDLoc &DL, EVT VT, Elts.append(OpVT.getVectorNumElements(), DAG.getUNDEF(SVT)); else if (Op.getOpcode() == ISD::BUILD_VECTOR) Elts.append(Op->op_begin(), Op->op_end()); + else if (Op.getOpcode() == ISD::INSERT_VECTOR_ELT && + OpVT.getVectorNumElements() == 1 && + isNullConstant(Op.getOperand(2))) + Elts.push_back(Op.getOperand(1)); else return SDValue(); } diff --git a/llvm/lib/DebugInfo/DWARF/DWARFDebugFrame.cpp b/llvm/lib/DebugInfo/DWARF/DWARFDebugFrame.cpp index 6610eef..c61f757 100644 --- a/llvm/lib/DebugInfo/DWARF/DWARFDebugFrame.cpp +++ b/llvm/lib/DebugInfo/DWARF/DWARFDebugFrame.cpp @@ -181,8 +181,8 @@ DWARFDebugFrame::DWARFDebugFrame(Triple::ArchType Arch, DWARFDebugFrame::~DWARFDebugFrame() = default; -static void LLVM_ATTRIBUTE_UNUSED dumpDataAux(DataExtractor Data, - uint64_t Offset, int Length) { +[[maybe_unused]] static void dumpDataAux(DataExtractor Data, uint64_t Offset, + int Length) { errs() << "DUMP: "; for (int i = 0; i < Length; ++i) { uint8_t c = Data.getU8(&Offset); diff --git a/llvm/lib/DebugInfo/GSYM/DwarfTransformer.cpp b/llvm/lib/DebugInfo/GSYM/DwarfTransformer.cpp index 7a0256f..fa39603 100644 --- a/llvm/lib/DebugInfo/GSYM/DwarfTransformer.cpp +++ b/llvm/lib/DebugInfo/GSYM/DwarfTransformer.cpp @@ -338,9 +338,13 @@ static void convertFunctionLineTable(OutputAggregator &Out, CUInfo &CUI, if (FilePath.empty()) { // If we had a DW_AT_decl_file, but got no file then we need to emit a // warning. + const uint64_t DwarfFileIdx = dwarf::toUnsigned( + Die.findRecursively(dwarf::DW_AT_decl_file), UINT32_MAX); + // Check if there is no DW_AT_decl_line attribute, and don't report an + // error if it isn't there. + if (DwarfFileIdx == UINT32_MAX) + return; Out.Report("Invalid file index in DW_AT_decl_file", [&](raw_ostream &OS) { - const uint64_t DwarfFileIdx = dwarf::toUnsigned( - Die.findRecursively(dwarf::DW_AT_decl_file), UINT32_MAX); OS << "error: function DIE at " << HEX32(Die.getOffset()) << " has an invalid file index " << DwarfFileIdx << " in its DW_AT_decl_file attribute, unable to create a single " @@ -629,6 +633,10 @@ Error DwarfTransformer::convert(uint32_t NumThreads, OutputAggregator &Out) { size_t NumBefore = Gsym.getNumFunctionInfos(); auto getDie = [&](DWARFUnit &DwarfUnit) -> DWARFDie { DWARFDie ReturnDie = DwarfUnit.getUnitDIE(false); + // Apple uses DW_AT_GNU_dwo_id for things other than split DWARF. + if (IsMachO) + return ReturnDie; + if (DwarfUnit.getDWOId()) { DWARFUnit *DWOCU = DwarfUnit.getNonSkeletonUnitDIE(false).getDwarfUnit(); if (!DWOCU->isDWOUnit()) diff --git a/llvm/lib/ExecutionEngine/JITLink/JITLinkMemoryManager.cpp b/llvm/lib/ExecutionEngine/JITLink/JITLinkMemoryManager.cpp index 5b3c05e..6c7e27e 100644 --- a/llvm/lib/ExecutionEngine/JITLink/JITLinkMemoryManager.cpp +++ b/llvm/lib/ExecutionEngine/JITLink/JITLinkMemoryManager.cpp @@ -260,22 +260,17 @@ public: } // Run finalization actions. - using WrapperFunctionCall = orc::shared::WrapperFunctionCall; - runFinalizeActions( - G->allocActions(), - [this, OnFinalized = std::move(OnFinalized)]( - Expected<std::vector<WrapperFunctionCall>> DeallocActions) mutable { - completeFinalization(std::move(OnFinalized), - std::move(DeallocActions)); - }); - } + auto DeallocActions = runFinalizeActions(G->allocActions()); + if (!DeallocActions) { + OnFinalized(DeallocActions.takeError()); + return; + } - void abandon(OnAbandonedFunction OnAbandoned) override { - Error Err = Error::success(); - if (auto EC = sys::Memory::releaseMappedMemory(FinalizationSegments)) - Err = joinErrors(std::move(Err), errorCodeToError(EC)); - if (auto EC = sys::Memory::releaseMappedMemory(StandardSegments)) - Err = joinErrors(std::move(Err), errorCodeToError(EC)); + // Release the finalize segments slab. + if (auto EC = sys::Memory::releaseMappedMemory(FinalizationSegments)) { + OnFinalized(errorCodeToError(EC)); + return; + } #ifndef NDEBUG // Set 'G' to null to flag that we've been successfully finalized. @@ -284,22 +279,17 @@ public: G = nullptr; #endif - OnAbandoned(std::move(Err)); + // Continue with finalized allocation. + OnFinalized(MemMgr.createFinalizedAlloc(std::move(StandardSegments), + std::move(*DeallocActions))); } -private: - void completeFinalization( - OnFinalizedFunction OnFinalized, - Expected<std::vector<orc::shared::WrapperFunctionCall>> DeallocActions) { - - if (!DeallocActions) - return OnFinalized(DeallocActions.takeError()); - - // Release the finalize segments slab. - if (auto EC = sys::Memory::releaseMappedMemory(FinalizationSegments)) { - OnFinalized(errorCodeToError(EC)); - return; - } + void abandon(OnAbandonedFunction OnAbandoned) override { + Error Err = Error::success(); + if (auto EC = sys::Memory::releaseMappedMemory(FinalizationSegments)) + Err = joinErrors(std::move(Err), errorCodeToError(EC)); + if (auto EC = sys::Memory::releaseMappedMemory(StandardSegments)) + Err = joinErrors(std::move(Err), errorCodeToError(EC)); #ifndef NDEBUG // Set 'G' to null to flag that we've been successfully finalized. @@ -308,11 +298,10 @@ private: G = nullptr; #endif - // Continue with finalized allocation. - OnFinalized(MemMgr.createFinalizedAlloc(std::move(StandardSegments), - std::move(*DeallocActions))); + OnAbandoned(std::move(Err)); } +private: Error applyProtections() { for (auto &KV : BL.segments()) { const auto &AG = KV.first; diff --git a/llvm/lib/ExecutionEngine/Orc/MemoryMapper.cpp b/llvm/lib/ExecutionEngine/Orc/MemoryMapper.cpp index 7b327af..7e606c6a 100644 --- a/llvm/lib/ExecutionEngine/Orc/MemoryMapper.cpp +++ b/llvm/lib/ExecutionEngine/Orc/MemoryMapper.cpp @@ -91,19 +91,9 @@ void InProcessMemoryMapper::initialize(MemoryMapper::AllocInfo &AI, sys::Memory::InvalidateInstructionCache(Base.toPtr<void *>(), Size); } - std::vector<shared::WrapperFunctionCall> DeinitializeActions; - { - std::promise<MSVCPExpected<std::vector<shared::WrapperFunctionCall>>> P; - auto F = P.get_future(); - shared::runFinalizeActions( - AI.Actions, [&](Expected<std::vector<shared::WrapperFunctionCall>> R) { - P.set_value(std::move(R)); - }); - if (auto DeinitializeActionsOrErr = F.get()) - DeinitializeActions = std::move(*DeinitializeActionsOrErr); - else - return OnInitialized(DeinitializeActionsOrErr.takeError()); - } + auto DeinitializeActions = shared::runFinalizeActions(AI.Actions); + if (!DeinitializeActions) + return OnInitialized(DeinitializeActions.takeError()); { std::lock_guard<std::mutex> Lock(Mutex); @@ -111,7 +101,7 @@ void InProcessMemoryMapper::initialize(MemoryMapper::AllocInfo &AI, // This is the maximum range whose permission have been possibly modified auto &Alloc = Allocations[MinAddr]; Alloc.Size = MaxAddr - MinAddr; - Alloc.DeinitializationActions = std::move(DeinitializeActions); + Alloc.DeinitializationActions = std::move(*DeinitializeActions); Reservations[AI.MappingBase.toPtr<void *>()].Allocations.push_back(MinAddr); } @@ -128,10 +118,10 @@ void InProcessMemoryMapper::deinitialize( for (auto Base : llvm::reverse(Bases)) { - shared::runDeallocActions( - Allocations[Base].DeinitializationActions, [&](Error Err) { - AllErr = joinErrors(std::move(AllErr), std::move(Err)); - }); + if (Error Err = shared::runDeallocActions( + Allocations[Base].DeinitializationActions)) { + AllErr = joinErrors(std::move(AllErr), std::move(Err)); + } // Reset protections to read/write so the area can be reused if (auto EC = sys::Memory::protectMappedMemory( diff --git a/llvm/lib/ExecutionEngine/Orc/Shared/AllocationActions.cpp b/llvm/lib/ExecutionEngine/Orc/Shared/AllocationActions.cpp index 08ab0c6..91f2899 100644 --- a/llvm/lib/ExecutionEngine/Orc/Shared/AllocationActions.cpp +++ b/llvm/lib/ExecutionEngine/Orc/Shared/AllocationActions.cpp @@ -12,39 +12,31 @@ namespace llvm { namespace orc { namespace shared { -void runFinalizeActions(AllocActions &AAs, - OnRunFinalizeActionsCompleteFn OnComplete) { +Expected<std::vector<WrapperFunctionCall>> +runFinalizeActions(AllocActions &AAs) { std::vector<WrapperFunctionCall> DeallocActions; DeallocActions.reserve(numDeallocActions(AAs)); for (auto &AA : AAs) { if (AA.Finalize) - - if (auto Err = AA.Finalize.runWithSPSRetErrorMerged()) { - while (!DeallocActions.empty()) { - Err = joinErrors(std::move(Err), - DeallocActions.back().runWithSPSRetErrorMerged()); - DeallocActions.pop_back(); - } - return OnComplete(std::move(Err)); - } + if (auto Err = AA.Finalize.runWithSPSRetErrorMerged()) + return joinErrors(std::move(Err), runDeallocActions(DeallocActions)); if (AA.Dealloc) DeallocActions.push_back(std::move(AA.Dealloc)); } AAs.clear(); - OnComplete(std::move(DeallocActions)); + return DeallocActions; } -void runDeallocActions(ArrayRef<WrapperFunctionCall> DAs, - OnRunDeallocActionsComeleteFn OnComplete) { +Error runDeallocActions(ArrayRef<WrapperFunctionCall> DAs) { Error Err = Error::success(); while (!DAs.empty()) { Err = joinErrors(std::move(Err), DAs.back().runWithSPSRetErrorMerged()); DAs = DAs.drop_back(); } - OnComplete(std::move(Err)); + return Err; } } // namespace shared diff --git a/llvm/lib/ExecutionEngine/Orc/TargetProcess/ExecutorSharedMemoryMapperService.cpp b/llvm/lib/ExecutionEngine/Orc/TargetProcess/ExecutorSharedMemoryMapperService.cpp index 8c24b1f..4fbf232 100644 --- a/llvm/lib/ExecutionEngine/Orc/TargetProcess/ExecutorSharedMemoryMapperService.cpp +++ b/llvm/lib/ExecutionEngine/Orc/TargetProcess/ExecutorSharedMemoryMapperService.cpp @@ -9,10 +9,8 @@ #include "llvm/ExecutionEngine/Orc/TargetProcess/ExecutorSharedMemoryMapperService.h" #include "llvm/Config/llvm-config.h" // for LLVM_ON_UNIX #include "llvm/ExecutionEngine/Orc/Shared/OrcRTBridge.h" -#include "llvm/Support/MSVCErrorWorkarounds.h" #include "llvm/Support/Process.h" #include "llvm/Support/WindowsError.h" -#include <future> #include <sstream> #if defined(LLVM_ON_UNIX) @@ -183,24 +181,15 @@ Expected<ExecutorAddr> ExecutorSharedMemoryMapperService::initialize( } // Run finalization actions and get deinitlization action list. - std::vector<shared::WrapperFunctionCall> DeinitializeActions; - { - std::promise<MSVCPExpected<std::vector<shared::WrapperFunctionCall>>> P; - auto F = P.get_future(); - shared::runFinalizeActions( - FR.Actions, [&](Expected<std::vector<shared::WrapperFunctionCall>> R) { - P.set_value(std::move(R)); - }); - if (auto DeinitializeActionsOrErr = F.get()) - DeinitializeActions = std::move(*DeinitializeActionsOrErr); - else - return DeinitializeActionsOrErr.takeError(); + auto DeinitializeActions = shared::runFinalizeActions(FR.Actions); + if (!DeinitializeActions) { + return DeinitializeActions.takeError(); } { std::lock_guard<std::mutex> Lock(Mutex); Allocations[MinAddr].DeinitializationActions = - std::move(DeinitializeActions); + std::move(*DeinitializeActions); Reservations[Reservation.toPtr<void *>()].Allocations.push_back(MinAddr); } @@ -221,11 +210,10 @@ Error ExecutorSharedMemoryMapperService::deinitialize( std::lock_guard<std::mutex> Lock(Mutex); for (auto Base : llvm::reverse(Bases)) { - shared::runDeallocActions( - Allocations[Base].DeinitializationActions, [&](Error Err) { - if (Err) - AllErr = joinErrors(std::move(AllErr), std::move(Err)); - }); + if (Error Err = shared::runDeallocActions( + Allocations[Base].DeinitializationActions)) { + AllErr = joinErrors(std::move(AllErr), std::move(Err)); + } // Remove the allocation from the allocation list of its reservation for (auto &Reservation : Reservations) { diff --git a/llvm/lib/IR/AsmWriter.cpp b/llvm/lib/IR/AsmWriter.cpp index 2430d98..3908a78 100644 --- a/llvm/lib/IR/AsmWriter.cpp +++ b/llvm/lib/IR/AsmWriter.cpp @@ -2374,16 +2374,21 @@ static void writeDICompileUnit(raw_ostream &Out, const DICompileUnit *N, Out << "!DICompileUnit("; MDFieldPrinter Printer(Out, WriterCtx); - auto Lang = N->getSourceLanguage(); - if (Lang.hasVersionedName()) + DISourceLanguageName Lang = N->getSourceLanguage(); + + if (Lang.hasVersionedName()) { Printer.printDwarfEnum( "sourceLanguageName", static_cast<llvm::dwarf::SourceLanguageName>(Lang.getName()), dwarf::SourceLanguageNameString, /* ShouldSkipZero */ false); - else + + Printer.printInt("sourceLanguageVersion", Lang.getVersion(), + /*ShouldSkipZero=*/true); + } else { Printer.printDwarfEnum("language", Lang.getName(), dwarf::LanguageString, /* ShouldSkipZero */ false); + } Printer.printMetadata("file", N->getRawFile(), /* ShouldSkipNull */ false); Printer.printString("producer", N->getProducer()); diff --git a/llvm/lib/MC/MCObjectFileInfo.cpp b/llvm/lib/MC/MCObjectFileInfo.cpp index a755c22..aee3c3b 100644 --- a/llvm/lib/MC/MCObjectFileInfo.cpp +++ b/llvm/lib/MC/MCObjectFileInfo.cpp @@ -553,7 +553,8 @@ void MCObjectFileInfo::initELFMCObjectFileInfo(const Triple &T, bool Large) { SFrameSection = Ctx->getELFSection(".sframe", ELF::SHT_GNU_SFRAME, ELF::SHF_ALLOC); - CallGraphSection = Ctx->getELFSection(".callgraph", ELF::SHT_PROGBITS, 0); + CallGraphSection = + Ctx->getELFSection(".llvm.callgraph", ELF::SHT_PROGBITS, 0); StackSizesSection = Ctx->getELFSection(".stack_sizes", ELF::SHT_PROGBITS, 0); @@ -1171,8 +1172,8 @@ MCObjectFileInfo::getCallGraphSection(const MCSection &TextSec) const { } return Ctx->getELFSection( - ".callgraph", ELF::SHT_PROGBITS, Flags, 0, GroupName, true, - ElfSec.getUniqueID(), + ".llvm.callgraph", ELF::SHT_PROGBITS, Flags, 0, GroupName, + /*IsComdat=*/true, ElfSec.getUniqueID(), static_cast<const MCSymbolELF *>(TextSec.getBeginSymbol())); } diff --git a/llvm/lib/ObjCopy/ConfigManager.cpp b/llvm/lib/ObjCopy/ConfigManager.cpp index eef8a21..6b7b4f1 100644 --- a/llvm/lib/ObjCopy/ConfigManager.cpp +++ b/llvm/lib/ObjCopy/ConfigManager.cpp @@ -122,14 +122,14 @@ ConfigManager::getDXContainerConfig() const { if (!Common.AddGnuDebugLink.empty() || !Common.SplitDWO.empty() || !Common.AllocSectionsPrefix.empty() || Common.DiscardMode != DiscardType::None || !Common.AddSection.empty() || - !Common.DumpSection.empty() || !Common.KeepSection.empty() || - !Common.SectionsToRename.empty() || !Common.SetSectionAlignment.empty() || - !Common.SetSectionFlags.empty() || !Common.SetSectionType.empty() || - Common.ExtractDWO || Common.OnlyKeepDebug || Common.StripAllGNU || - Common.StripDWO || Common.StripDebug || Common.StripNonAlloc || - Common.StripSections || Common.StripUnneeded || - Common.DecompressDebugSections || Common.GapFill != 0 || - Common.PadTo != 0 || Common.ChangeSectionLMAValAll != 0 || + !Common.KeepSection.empty() || !Common.SectionsToRename.empty() || + !Common.SetSectionAlignment.empty() || !Common.SetSectionFlags.empty() || + !Common.SetSectionType.empty() || Common.ExtractDWO || + Common.OnlyKeepDebug || Common.StripAllGNU || Common.StripDWO || + Common.StripDebug || Common.StripNonAlloc || Common.StripSections || + Common.StripUnneeded || Common.DecompressDebugSections || + Common.GapFill != 0 || Common.PadTo != 0 || + Common.ChangeSectionLMAValAll != 0 || !Common.ChangeSectionAddress.empty()) { return createStringError(llvm::errc::invalid_argument, "option is not supported for DXContainer"); diff --git a/llvm/lib/ObjCopy/DXContainer/DXContainerObjcopy.cpp b/llvm/lib/ObjCopy/DXContainer/DXContainerObjcopy.cpp index d7f3c0d..95ab3d9 100644 --- a/llvm/lib/ObjCopy/DXContainer/DXContainerObjcopy.cpp +++ b/llvm/lib/ObjCopy/DXContainer/DXContainerObjcopy.cpp @@ -9,8 +9,10 @@ #include "llvm/ObjCopy/DXContainer/DXContainerObjcopy.h" #include "DXContainerReader.h" #include "DXContainerWriter.h" +#include "llvm/BinaryFormat/DXContainer.h" #include "llvm/ObjCopy/CommonConfig.h" #include "llvm/ObjCopy/DXContainer/DXContainerConfig.h" +#include "llvm/Support/FileOutputBuffer.h" #include "llvm/Support/raw_ostream.h" namespace llvm { @@ -42,7 +44,47 @@ static Error extractPartAsObject(StringRef PartName, StringRef OutFilename, "part '%s' not found", PartName.str().c_str()); } +static Error dumpPartToFile(StringRef PartName, StringRef Filename, + StringRef InputFilename, Object &Obj) { + auto PartIter = llvm::find_if( + Obj.Parts, [&PartName](const Part &P) { return P.Name == PartName; }); + if (PartIter == Obj.Parts.end()) + return createFileError(Filename, + std::make_error_code(std::errc::invalid_argument), + "part '%s' not found", PartName.str().c_str()); + ArrayRef<uint8_t> Contents = PartIter->Data; + // The DXContainer format is a bit odd because the part-specific headers are + // contained inside the part data itself. For parts that contain LLVM bitcode + // when we dump the part we want to skip the part-specific header so that we + // get a valid .bc file that we can inspect. All the data contained inside the + // program header is pulled out of the bitcode, so the header can be + // reconstructed if needed from the bitcode itself. More comprehensive + // documentation on the DXContainer format can be found at + // https://llvm.org/docs/DirectX/DXContainer.html. + + if (PartName == "DXIL" || PartName == "STAT") + Contents = Contents.drop_front(sizeof(llvm::dxbc::ProgramHeader)); + if (Contents.empty()) + return createFileError(Filename, object_error::parse_failed, + "part '%s' is empty", PartName.str().c_str()); + Expected<std::unique_ptr<FileOutputBuffer>> BufferOrErr = + FileOutputBuffer::create(Filename, Contents.size()); + if (!BufferOrErr) + return createFileError(Filename, BufferOrErr.takeError()); + std::unique_ptr<FileOutputBuffer> Buf = std::move(*BufferOrErr); + llvm::copy(Contents, Buf->getBufferStart()); + if (Error E = Buf->commit()) + return createFileError(Filename, std::move(E)); + return Error::success(); +} + static Error handleArgs(const CommonConfig &Config, Object &Obj) { + for (StringRef Flag : Config.DumpSection) { + auto [SecName, FileName] = Flag.split("="); + if (Error E = dumpPartToFile(SecName, FileName, Config.InputFilename, Obj)) + return E; + } + // Extract all sections before any modifications. for (StringRef Flag : Config.ExtractSection) { StringRef SectionName; diff --git a/llvm/lib/Object/ELF.cpp b/llvm/lib/Object/ELF.cpp index 53699ce0..f256e7b 100644 --- a/llvm/lib/Object/ELF.cpp +++ b/llvm/lib/Object/ELF.cpp @@ -837,7 +837,7 @@ decodeBBAddrMapImpl(const ELFFile<ELFT> &EF, Version = Data.getU8(Cur); if (!Cur) break; - if (Version < 2 || Version > 3) + if (Version < 2 || Version > 4) return createError("unsupported SHT_LLVM_BB_ADDR_MAP version: " + Twine(static_cast<int>(Version))); Feature = Data.getU8(Cur); // Feature byte @@ -852,6 +852,11 @@ decodeBBAddrMapImpl(const ELFFile<ELFT> &EF, "callsite offsets feature is enabled: version = " + Twine(static_cast<int>(Version)) + " feature = " + Twine(static_cast<int>(Feature))); + if (FeatEnable.BBHash && Version < 4) + return createError("version should be >= 4 for SHT_LLVM_BB_ADDR_MAP when " + "basic block hash feature is enabled: version = " + + Twine(static_cast<int>(Version)) + + " feature = " + Twine(static_cast<int>(Feature))); uint32_t NumBlocksInBBRange = 0; uint32_t NumBBRanges = 1; typename ELFFile<ELFT>::uintX_t RangeBaseAddress = 0; @@ -907,6 +912,7 @@ decodeBBAddrMapImpl(const ELFFile<ELFT> &EF, uint32_t Size = readULEB128As<uint32_t>(Data, Cur, ULEBSizeErr) + LastCallsiteEndOffset; uint32_t MD = readULEB128As<uint32_t>(Data, Cur, ULEBSizeErr); + uint64_t Hash = FeatEnable.BBHash ? Data.getU64(Cur) : 0; Expected<BBAddrMap::BBEntry::Metadata> MetadataOrErr = BBAddrMap::BBEntry::Metadata::decode(MD); if (!MetadataOrErr) { @@ -914,7 +920,7 @@ decodeBBAddrMapImpl(const ELFFile<ELFT> &EF, break; } BBEntries.push_back({ID, Offset + PrevBBEndOffset, Size, - *MetadataOrErr, CallsiteEndOffsets}); + *MetadataOrErr, CallsiteEndOffsets, Hash}); PrevBBEndOffset += Offset + Size; } TotalNumBlocks += BBEntries.size(); diff --git a/llvm/lib/ObjectYAML/ELFEmitter.cpp b/llvm/lib/ObjectYAML/ELFEmitter.cpp index faeeab3..8b75fbe 100644 --- a/llvm/lib/ObjectYAML/ELFEmitter.cpp +++ b/llvm/lib/ObjectYAML/ELFEmitter.cpp @@ -1465,7 +1465,7 @@ void ELFState<ELFT>::writeSectionContent( for (const auto &[Idx, E] : llvm::enumerate(*Section.Entries)) { // Write version and feature values. if (Section.Type == llvm::ELF::SHT_LLVM_BB_ADDR_MAP) { - if (E.Version > 3) + if (E.Version > 4) WithColor::warning() << "unsupported SHT_LLVM_BB_ADDR_MAP version: " << static_cast<int>(E.Version) << "; encoding using the most recent version"; @@ -1526,6 +1526,12 @@ void ELFState<ELFT>::writeSectionContent( } SHeader.sh_size += CBA.writeULEB128(BBE.Size); SHeader.sh_size += CBA.writeULEB128(BBE.Metadata); + if (FeatureOrErr->BBHash || BBE.Hash.has_value()) { + uint64_t Hash = + BBE.Hash.has_value() ? BBE.Hash.value() : llvm::yaml::Hex64(0); + CBA.write<uint64_t>(Hash, ELFT::Endianness); + SHeader.sh_size += 8; + } } } if (!PGOAnalyses) diff --git a/llvm/lib/ObjectYAML/ELFYAML.cpp b/llvm/lib/ObjectYAML/ELFYAML.cpp index d9cce1e..421d6603 100644 --- a/llvm/lib/ObjectYAML/ELFYAML.cpp +++ b/llvm/lib/ObjectYAML/ELFYAML.cpp @@ -1887,6 +1887,7 @@ void MappingTraits<ELFYAML::BBAddrMapEntry::BBEntry>::mapping( IO.mapRequired("Size", E.Size); IO.mapRequired("Metadata", E.Metadata); IO.mapOptional("CallsiteEndOffsets", E.CallsiteEndOffsets); + IO.mapOptional("Hash", E.Hash); } void MappingTraits<ELFYAML::PGOAnalysisMapEntry>::mapping( diff --git a/llvm/lib/ProfileData/InstrProf.cpp b/llvm/lib/ProfileData/InstrProf.cpp index 3c8e44a..0208735 100644 --- a/llvm/lib/ProfileData/InstrProf.cpp +++ b/llvm/lib/ProfileData/InstrProf.cpp @@ -302,7 +302,7 @@ void ProfOStream::patch(ArrayRef<PatchItem> P) { std::string getPGOFuncName(StringRef Name, GlobalValue::LinkageTypes Linkage, StringRef FileName, - uint64_t Version LLVM_ATTRIBUTE_UNUSED) { + [[maybe_unused]] uint64_t Version) { // Value names may be prefixed with a binary '1' to indicate // that the backend should not modify the symbols due to any platform // naming convention. Do not include that '1' in the PGO profile name. diff --git a/llvm/lib/Support/PrettyStackTrace.cpp b/llvm/lib/Support/PrettyStackTrace.cpp index 82b0e6a..eff9947 100644 --- a/llvm/lib/Support/PrettyStackTrace.cpp +++ b/llvm/lib/Support/PrettyStackTrace.cpp @@ -141,7 +141,7 @@ extern "C" const char *__crashreporter_info__ asm(".desc ___crashreporter_info__, 0x10"); #endif -static void setCrashLogMessage(const char *msg) LLVM_ATTRIBUTE_UNUSED; +[[maybe_unused]] static void setCrashLogMessage(const char *msg); static void setCrashLogMessage(const char *msg) { #ifdef HAVE_CRASHREPORTERCLIENT_H (void)CRSetCrashLogMessage(msg); diff --git a/llvm/lib/Support/SourceMgr.cpp b/llvm/lib/Support/SourceMgr.cpp index a43cf37a..f2bbaab 100644 --- a/llvm/lib/Support/SourceMgr.cpp +++ b/llvm/lib/Support/SourceMgr.cpp @@ -24,6 +24,7 @@ #include "llvm/Support/MemoryBuffer.h" #include "llvm/Support/Path.h" #include "llvm/Support/SMLoc.h" +#include "llvm/Support/VirtualFileSystem.h" #include "llvm/Support/WithColor.h" #include "llvm/Support/raw_ostream.h" #include <algorithm> @@ -38,6 +39,22 @@ using namespace llvm; static const size_t TabStop = 8; +// Out of line to avoid needing definition of vfs::FileSystem in header. +SourceMgr::SourceMgr() = default; +SourceMgr::SourceMgr(IntrusiveRefCntPtr<vfs::FileSystem> FS) + : FS(std::move(FS)) {} +SourceMgr::SourceMgr(SourceMgr &&) = default; +SourceMgr &SourceMgr::operator=(SourceMgr &&) = default; +SourceMgr::~SourceMgr() = default; + +IntrusiveRefCntPtr<vfs::FileSystem> SourceMgr::getVirtualFileSystem() const { + return FS; +} + +void SourceMgr::setVirtualFileSystem(IntrusiveRefCntPtr<vfs::FileSystem> FS) { + this->FS = std::move(FS); +} + unsigned SourceMgr::AddIncludeFile(const std::string &Filename, SMLoc IncludeLoc, std::string &IncludedFile) { @@ -52,8 +69,11 @@ unsigned SourceMgr::AddIncludeFile(const std::string &Filename, ErrorOr<std::unique_ptr<MemoryBuffer>> SourceMgr::OpenIncludeFile(const std::string &Filename, std::string &IncludedFile) { + if (!FS) + reportFatalInternalError("Opening include file from SourceMgr without VFS"); + ErrorOr<std::unique_ptr<MemoryBuffer>> NewBufOrErr = - MemoryBuffer::getFile(Filename); + FS->getBufferForFile(Filename); SmallString<64> Buffer(Filename); // If the file didn't exist directly, see if it's in an include path. @@ -61,7 +81,7 @@ SourceMgr::OpenIncludeFile(const std::string &Filename, ++i) { Buffer = IncludeDirectories[i]; sys::path::append(Buffer, Filename); - NewBufOrErr = MemoryBuffer::getFile(Buffer); + NewBufOrErr = FS->getBufferForFile(Buffer); } if (NewBufOrErr) diff --git a/llvm/lib/Support/TextEncoding.cpp b/llvm/lib/Support/TextEncoding.cpp index 804ff07..41f5187 100644 --- a/llvm/lib/Support/TextEncoding.cpp +++ b/llvm/lib/Support/TextEncoding.cpp @@ -54,9 +54,9 @@ static std::optional<TextEncoding> getKnownEncoding(StringRef Name) { return std::nullopt; } -LLVM_ATTRIBUTE_UNUSED static void -HandleOverflow(size_t &Capacity, char *&Output, size_t &OutputLength, - SmallVectorImpl<char> &Result) { +[[maybe_unused]] static void HandleOverflow(size_t &Capacity, char *&Output, + size_t &OutputLength, + SmallVectorImpl<char> &Result) { // No space left in output buffer. Double the size of the underlying // memory in the SmallVectorImpl, adjust pointer and length and continue // the conversion. diff --git a/llvm/lib/Support/UnicodeNameToCodepoint.cpp b/llvm/lib/Support/UnicodeNameToCodepoint.cpp index 8d66348..6f8e091 100644 --- a/llvm/lib/Support/UnicodeNameToCodepoint.cpp +++ b/llvm/lib/Support/UnicodeNameToCodepoint.cpp @@ -476,7 +476,7 @@ nearestMatchesForCodepointName(StringRef Pattern, std::size_t MaxMatchesCount) { std::min(NormalizedName.size(), UnicodeNameToCodepointLargestNameSize) + 1; - LLVM_ATTRIBUTE_UNUSED static std::size_t Rows = + [[maybe_unused]] static std::size_t Rows = UnicodeNameToCodepointLargestNameSize + 1; std::vector<char> Distances( diff --git a/llvm/lib/TableGen/Main.cpp b/llvm/lib/TableGen/Main.cpp index 42043f7..b1024a8 100644 --- a/llvm/lib/TableGen/Main.cpp +++ b/llvm/lib/TableGen/Main.cpp @@ -26,6 +26,7 @@ #include "llvm/Support/SMLoc.h" #include "llvm/Support/SourceMgr.h" #include "llvm/Support/ToolOutputFile.h" +#include "llvm/Support/VirtualFileSystem.h" #include "llvm/Support/raw_ostream.h" #include "llvm/TableGen/Error.h" #include "llvm/TableGen/Record.h" @@ -128,6 +129,7 @@ int llvm::TableGenMain(const char *argv0, // Record the location of the include directory so that the lexer can find // it later. SrcMgr.setIncludeDirs(IncludeDirs); + SrcMgr.setVirtualFileSystem(vfs::getRealFileSystem()); TGParser Parser(SrcMgr, MacroNames, Records, NoWarnOnUnusedTemplateArgs); diff --git a/llvm/lib/TableGen/Parser.cpp b/llvm/lib/TableGen/Parser.cpp index 2c3726a..db45054 100644 --- a/llvm/lib/TableGen/Parser.cpp +++ b/llvm/lib/TableGen/Parser.cpp @@ -9,6 +9,7 @@ #include "llvm/TableGen/Parser.h" #include "TGParser.h" #include "llvm/Support/MemoryBuffer.h" +#include "llvm/Support/VirtualFileSystem.h" #include "llvm/TableGen/Record.h" using namespace llvm; @@ -21,6 +22,7 @@ bool llvm::TableGenParseFile(SourceMgr &InputSrcMgr, RecordKeeper &Records) { SrcMgr = SourceMgr(); SrcMgr.takeSourceBuffersFrom(InputSrcMgr); SrcMgr.setIncludeDirs(InputSrcMgr.getIncludeDirs()); + SrcMgr.setVirtualFileSystem(InputSrcMgr.getVirtualFileSystem()); SrcMgr.setDiagHandler(InputSrcMgr.getDiagHandler(), InputSrcMgr.getDiagContext()); diff --git a/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp b/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp index 9926a4d..be2f2e4 100644 --- a/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp +++ b/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp @@ -16254,7 +16254,7 @@ SDValue AArch64TargetLowering::LowerDIV(SDValue Op, SelectionDAG &DAG) const { SplatVal > 1) { SDValue Pg = getPredicateForScalableVector(DAG, DL, VT); SDValue Res = - DAG.getNode(AArch64ISD::SRAD_MERGE_OP1, DL, VT, Pg, Op->getOperand(0), + DAG.getNode(AArch64ISD::ASRD_MERGE_OP1, DL, VT, Pg, Op->getOperand(0), DAG.getTargetConstant(Log2_64(SplatVal), DL, MVT::i32)); if (Negated) Res = DAG.getNode(ISD::SUB, DL, VT, DAG.getConstant(0, DL, VT), Res); @@ -22942,7 +22942,7 @@ static SDValue performIntrinsicCombine(SDNode *N, return DAG.getNode(ISD::USUBSAT, SDLoc(N), N->getValueType(0), N->getOperand(1), N->getOperand(2)); case Intrinsic::aarch64_sve_asrd: - return DAG.getNode(AArch64ISD::SRAD_MERGE_OP1, SDLoc(N), N->getValueType(0), + return DAG.getNode(AArch64ISD::ASRD_MERGE_OP1, SDLoc(N), N->getValueType(0), N->getOperand(1), N->getOperand(2), N->getOperand(3)); case Intrinsic::aarch64_sve_cmphs: if (!N->getOperand(2).getValueType().isFloatingPoint()) @@ -30047,7 +30047,7 @@ SDValue AArch64TargetLowering::LowerFixedLengthVectorIntDivideToSVE( SDValue Pg = getPredicateForFixedLengthVector(DAG, DL, VT); SDValue Res = - DAG.getNode(AArch64ISD::SRAD_MERGE_OP1, DL, ContainerVT, Pg, Op1, Op2); + DAG.getNode(AArch64ISD::ASRD_MERGE_OP1, DL, ContainerVT, Pg, Op1, Op2); if (Negated) Res = DAG.getNode(ISD::SUB, DL, ContainerVT, DAG.getConstant(0, DL, ContainerVT), Res); diff --git a/llvm/lib/Target/AArch64/AArch64InstrGISel.td b/llvm/lib/Target/AArch64/AArch64InstrGISel.td index 7322212..fe84193 100644 --- a/llvm/lib/Target/AArch64/AArch64InstrGISel.td +++ b/llvm/lib/Target/AArch64/AArch64InstrGISel.td @@ -233,6 +233,12 @@ def G_SDOT : AArch64GenericInstruction { let hasSideEffects = 0; } +def G_USDOT : AArch64GenericInstruction { + let OutOperandList = (outs type0:$dst); + let InOperandList = (ins type0:$src1, type0:$src2, type0:$src3); + let hasSideEffects = 0; +} + // Generic instruction for the BSP pseudo. It is expanded into BSP, which // expands into BSL/BIT/BIF after register allocation. def G_BSP : AArch64GenericInstruction { @@ -278,6 +284,7 @@ def : GINodeEquiv<G_UADDLV, AArch64uaddlv>; def : GINodeEquiv<G_UDOT, AArch64udot>; def : GINodeEquiv<G_SDOT, AArch64sdot>; +def : GINodeEquiv<G_USDOT, AArch64usdot>; def : GINodeEquiv<G_EXTRACT_VECTOR_ELT, vector_extract>; diff --git a/llvm/lib/Target/AArch64/AArch64MachineFunctionInfo.cpp b/llvm/lib/Target/AArch64/AArch64MachineFunctionInfo.cpp index b3c9656..343fd81 100644 --- a/llvm/lib/Target/AArch64/AArch64MachineFunctionInfo.cpp +++ b/llvm/lib/Target/AArch64/AArch64MachineFunctionInfo.cpp @@ -40,7 +40,11 @@ yaml::AArch64FunctionInfo::AArch64FunctionInfo( getSVEStackSize(MFI, &llvm::AArch64FunctionInfo::getStackSizePPR)), HasStackFrame(MFI.hasStackFrame() ? std::optional<bool>(MFI.hasStackFrame()) - : std::nullopt) {} + : std::nullopt), + HasStreamingModeChanges( + MFI.hasStreamingModeChanges() + ? std::optional<bool>(MFI.hasStreamingModeChanges()) + : std::nullopt) {} void yaml::AArch64FunctionInfo::mappingImpl(yaml::IO &YamlIO) { MappingTraits<AArch64FunctionInfo>::mapping(YamlIO, *this); @@ -55,6 +59,8 @@ void AArch64FunctionInfo::initializeBaseYamlFields( YamlMFI.StackSizePPR.value_or(0)); if (YamlMFI.HasStackFrame) setHasStackFrame(*YamlMFI.HasStackFrame); + if (YamlMFI.HasStreamingModeChanges) + setHasStreamingModeChanges(*YamlMFI.HasStreamingModeChanges); } static std::pair<bool, bool> GetSignReturnAddress(const Function &F) { diff --git a/llvm/lib/Target/AArch64/AArch64MachineFunctionInfo.h b/llvm/lib/Target/AArch64/AArch64MachineFunctionInfo.h index bd0a17d..d1832f4 100644 --- a/llvm/lib/Target/AArch64/AArch64MachineFunctionInfo.h +++ b/llvm/lib/Target/AArch64/AArch64MachineFunctionInfo.h @@ -645,6 +645,7 @@ struct AArch64FunctionInfo final : public yaml::MachineFunctionInfo { std::optional<uint64_t> StackSizeZPR; std::optional<uint64_t> StackSizePPR; std::optional<bool> HasStackFrame; + std::optional<bool> HasStreamingModeChanges; AArch64FunctionInfo() = default; AArch64FunctionInfo(const llvm::AArch64FunctionInfo &MFI); @@ -659,6 +660,7 @@ template <> struct MappingTraits<AArch64FunctionInfo> { YamlIO.mapOptional("stackSizeZPR", MFI.StackSizeZPR); YamlIO.mapOptional("stackSizePPR", MFI.StackSizePPR); YamlIO.mapOptional("hasStackFrame", MFI.HasStackFrame); + YamlIO.mapOptional("hasStreamingModeChanges", MFI.HasStreamingModeChanges); } }; diff --git a/llvm/lib/Target/AArch64/AArch64PostCoalescerPass.cpp b/llvm/lib/Target/AArch64/AArch64PostCoalescerPass.cpp index cdf2822..a90950d 100644 --- a/llvm/lib/Target/AArch64/AArch64PostCoalescerPass.cpp +++ b/llvm/lib/Target/AArch64/AArch64PostCoalescerPass.cpp @@ -75,6 +75,10 @@ bool AArch64PostCoalescer::runOnMachineFunction(MachineFunction &MF) { if (Src != Dst) MRI->replaceRegWith(Dst, Src); + if (MI.getOperand(1).isUndef()) + for (MachineOperand &MO : MRI->use_operands(Dst)) + MO.setIsUndef(); + // MI must be erased from the basic block before recalculating the live // interval. LIS->RemoveMachineInstrFromMaps(MI); diff --git a/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td b/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td index bc6b931..98a128e 100644 --- a/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td +++ b/llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td @@ -265,7 +265,7 @@ def SDT_AArch64Arith_Imm : SDTypeProfile<1, 3, [ SDTCVecEltisVT<1,i1>, SDTCisSameAs<0,2> ]>; -def AArch64asrd_m1 : SDNode<"AArch64ISD::SRAD_MERGE_OP1", SDT_AArch64Arith_Imm>; +def AArch64asrd_m1 : SDNode<"AArch64ISD::ASRD_MERGE_OP1", SDT_AArch64Arith_Imm>; def AArch64urshri_p_node : SDNode<"AArch64ISD::URSHR_I_PRED", SDT_AArch64Arith_Imm>; def AArch64urshri_p : PatFrags<(ops node:$op1, node:$op2, node:$op3), diff --git a/llvm/lib/Target/AArch64/AsmParser/AArch64AsmParser.cpp b/llvm/lib/Target/AArch64/AsmParser/AArch64AsmParser.cpp index 2c3870c..636d4f8a 100644 --- a/llvm/lib/Target/AArch64/AsmParser/AArch64AsmParser.cpp +++ b/llvm/lib/Target/AArch64/AsmParser/AArch64AsmParser.cpp @@ -8217,6 +8217,8 @@ bool AArch64AsmParser::parseDataExpr(const MCExpr *&Res) { Spec = AArch64::S_GOTPCREL; else if (Identifier == "plt") Spec = AArch64::S_PLT; + else if (Identifier == "funcinit") + Spec = AArch64::S_FUNCINIT; } if (Spec == AArch64::S_None) return Error(Loc, "invalid relocation specifier"); diff --git a/llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp b/llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp index 9e2d698..05a4313 100644 --- a/llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp +++ b/llvm/lib/Target/AArch64/GISel/AArch64LegalizerInfo.cpp @@ -1855,6 +1855,8 @@ bool AArch64LegalizerInfo::legalizeIntrinsic(LegalizerHelper &Helper, return LowerTriOp(AArch64::G_UDOT); case Intrinsic::aarch64_neon_sdot: return LowerTriOp(AArch64::G_SDOT); + case Intrinsic::aarch64_neon_usdot: + return LowerTriOp(AArch64::G_USDOT); case Intrinsic::aarch64_neon_sqxtn: return LowerUnaryOp(TargetOpcode::G_TRUNC_SSAT_S); case Intrinsic::aarch64_neon_sqxtun: diff --git a/llvm/lib/Target/AArch64/MCTargetDesc/AArch64ELFObjectWriter.cpp b/llvm/lib/Target/AArch64/MCTargetDesc/AArch64ELFObjectWriter.cpp index a388216..892b8da 100644 --- a/llvm/lib/Target/AArch64/MCTargetDesc/AArch64ELFObjectWriter.cpp +++ b/llvm/lib/Target/AArch64/MCTargetDesc/AArch64ELFObjectWriter.cpp @@ -232,6 +232,8 @@ unsigned AArch64ELFObjectWriter::getRelocType(const MCFixup &Fixup, } if (RefKind == AArch64::S_AUTH || RefKind == AArch64::S_AUTHADDR) return ELF::R_AARCH64_AUTH_ABS64; + if (RefKind == AArch64::S_FUNCINIT) + return ELF::R_AARCH64_FUNCINIT64; return ELF::R_AARCH64_ABS64; } case AArch64::fixup_aarch64_add_imm12: diff --git a/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCAsmInfo.cpp b/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCAsmInfo.cpp index 2b5cf34..bc090c6 100644 --- a/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCAsmInfo.cpp +++ b/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCAsmInfo.cpp @@ -40,6 +40,7 @@ const MCAsmInfo::AtSpecifier ELFAtSpecifiers[] = { {AArch64::S_GOT, "GOT"}, {AArch64::S_GOTPCREL, "GOTPCREL"}, {AArch64::S_PLT, "PLT"}, + {AArch64::S_FUNCINIT, "FUNCINIT"}, }; const MCAsmInfo::AtSpecifier MachOAtSpecifiers[] = { diff --git a/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCAsmInfo.h b/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCAsmInfo.h index 0dfa61b..f2acff5 100644 --- a/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCAsmInfo.h +++ b/llvm/lib/Target/AArch64/MCTargetDesc/AArch64MCAsmInfo.h @@ -164,6 +164,7 @@ enum { // ELF relocation specifiers in data directives: S_PLT = 0x400, S_GOTPCREL, + S_FUNCINIT, // Mach-O @ relocation specifiers: S_MACHO_GOT, diff --git a/llvm/lib/Target/AArch64/MachineSMEABIPass.cpp b/llvm/lib/Target/AArch64/MachineSMEABIPass.cpp index 4749748..434ea67 100644 --- a/llvm/lib/Target/AArch64/MachineSMEABIPass.cpp +++ b/llvm/lib/Target/AArch64/MachineSMEABIPass.cpp @@ -294,6 +294,12 @@ struct MachineSMEABI : public MachineFunctionPass { MachineBasicBlock::iterator MBBI, LiveRegs PhysLiveRegs); + /// Attempts to find an insertion point before \p Inst where the status flags + /// are not live. If \p Inst is `Block.Insts.end()` a point before the end of + /// the block is found. + std::pair<MachineBasicBlock::iterator, LiveRegs> + findStateChangeInsertionPoint(MachineBasicBlock &MBB, const BlockInfo &Block, + SmallVectorImpl<InstInfo>::const_iterator Inst); void emitStateChange(EmitContext &, MachineBasicBlock &MBB, MachineBasicBlock::iterator MBBI, ZAState From, ZAState To, LiveRegs PhysLiveRegs); @@ -337,6 +343,28 @@ private: MachineRegisterInfo *MRI = nullptr; }; +static LiveRegs getPhysLiveRegs(LiveRegUnits const &LiveUnits) { + LiveRegs PhysLiveRegs = LiveRegs::None; + if (!LiveUnits.available(AArch64::NZCV)) + PhysLiveRegs |= LiveRegs::NZCV; + // We have to track W0 and X0 separately as otherwise things can get + // confused if we attempt to preserve X0 but only W0 was defined. + if (!LiveUnits.available(AArch64::W0)) + PhysLiveRegs |= LiveRegs::W0; + if (!LiveUnits.available(AArch64::W0_HI)) + PhysLiveRegs |= LiveRegs::W0_HI; + return PhysLiveRegs; +} + +static void setPhysLiveRegs(LiveRegUnits &LiveUnits, LiveRegs PhysLiveRegs) { + if (PhysLiveRegs & LiveRegs::NZCV) + LiveUnits.addReg(AArch64::NZCV); + if (PhysLiveRegs & LiveRegs::W0) + LiveUnits.addReg(AArch64::W0); + if (PhysLiveRegs & LiveRegs::W0_HI) + LiveUnits.addReg(AArch64::W0_HI); +} + FunctionInfo MachineSMEABI::collectNeededZAStates(SMEAttrs SMEFnAttrs) { assert((SMEFnAttrs.hasAgnosticZAInterface() || SMEFnAttrs.hasZT0State() || SMEFnAttrs.hasZAState()) && @@ -362,26 +390,13 @@ FunctionInfo MachineSMEABI::collectNeededZAStates(SMEAttrs SMEFnAttrs) { LiveRegUnits LiveUnits(*TRI); LiveUnits.addLiveOuts(MBB); - auto GetPhysLiveRegs = [&] { - LiveRegs PhysLiveRegs = LiveRegs::None; - if (!LiveUnits.available(AArch64::NZCV)) - PhysLiveRegs |= LiveRegs::NZCV; - // We have to track W0 and X0 separately as otherwise things can get - // confused if we attempt to preserve X0 but only W0 was defined. - if (!LiveUnits.available(AArch64::W0)) - PhysLiveRegs |= LiveRegs::W0; - if (!LiveUnits.available(AArch64::W0_HI)) - PhysLiveRegs |= LiveRegs::W0_HI; - return PhysLiveRegs; - }; - - Block.PhysLiveRegsAtExit = GetPhysLiveRegs(); + Block.PhysLiveRegsAtExit = getPhysLiveRegs(LiveUnits); auto FirstTerminatorInsertPt = MBB.getFirstTerminator(); auto FirstNonPhiInsertPt = MBB.getFirstNonPHI(); for (MachineInstr &MI : reverse(MBB)) { MachineBasicBlock::iterator MBBI(MI); LiveUnits.stepBackward(MI); - LiveRegs PhysLiveRegs = GetPhysLiveRegs(); + LiveRegs PhysLiveRegs = getPhysLiveRegs(LiveUnits); // The SMEStateAllocPseudo marker is added to a function if the save // buffer was allocated in SelectionDAG. It marks the end of the // allocation -- which is a safe point for this pass to insert any TPIDR2 @@ -476,6 +491,49 @@ MachineSMEABI::assignBundleZAStates(const EdgeBundles &Bundles, return BundleStates; } +std::pair<MachineBasicBlock::iterator, LiveRegs> +MachineSMEABI::findStateChangeInsertionPoint( + MachineBasicBlock &MBB, const BlockInfo &Block, + SmallVectorImpl<InstInfo>::const_iterator Inst) { + LiveRegs PhysLiveRegs; + MachineBasicBlock::iterator InsertPt; + if (Inst != Block.Insts.end()) { + InsertPt = Inst->InsertPt; + PhysLiveRegs = Inst->PhysLiveRegs; + } else { + InsertPt = MBB.getFirstTerminator(); + PhysLiveRegs = Block.PhysLiveRegsAtExit; + } + + if (!(PhysLiveRegs & LiveRegs::NZCV)) + return {InsertPt, PhysLiveRegs}; // Nothing to do (no live flags). + + // Find the previous state change. We can not move before this point. + MachineBasicBlock::iterator PrevStateChangeI; + if (Inst == Block.Insts.begin()) { + PrevStateChangeI = MBB.begin(); + } else { + // Note: `std::prev(Inst)` is the previous InstInfo. We only create an + // InstInfo object for instructions that require a specific ZA state, so the + // InstInfo is the site of the previous state change in the block (which can + // be several MIs earlier). + PrevStateChangeI = std::prev(Inst)->InsertPt; + } + + // Note: LiveUnits will only accurately track X0 and NZCV. + LiveRegUnits LiveUnits(*TRI); + setPhysLiveRegs(LiveUnits, PhysLiveRegs); + for (MachineBasicBlock::iterator I = InsertPt; I != PrevStateChangeI; --I) { + // Don't move before/into a call (which may have a state change before it). + if (I->getOpcode() == TII->getCallFrameDestroyOpcode() || I->isCall()) + break; + LiveUnits.stepBackward(*I); + if (LiveUnits.available(AArch64::NZCV)) + return {I, getPhysLiveRegs(LiveUnits)}; + } + return {InsertPt, PhysLiveRegs}; +} + void MachineSMEABI::insertStateChanges(EmitContext &Context, const FunctionInfo &FnInfo, const EdgeBundles &Bundles, @@ -490,10 +548,13 @@ void MachineSMEABI::insertStateChanges(EmitContext &Context, CurrentState = InState; for (auto &Inst : Block.Insts) { - if (CurrentState != Inst.NeededState) - emitStateChange(Context, MBB, Inst.InsertPt, CurrentState, - Inst.NeededState, Inst.PhysLiveRegs); - CurrentState = Inst.NeededState; + if (CurrentState != Inst.NeededState) { + auto [InsertPt, PhysLiveRegs] = + findStateChangeInsertionPoint(MBB, Block, &Inst); + emitStateChange(Context, MBB, InsertPt, CurrentState, Inst.NeededState, + PhysLiveRegs); + CurrentState = Inst.NeededState; + } } if (MBB.succ_empty()) @@ -501,9 +562,12 @@ void MachineSMEABI::insertStateChanges(EmitContext &Context, ZAState OutState = BundleStates[Bundles.getBundle(MBB.getNumber(), /*Out=*/true)]; - if (CurrentState != OutState) - emitStateChange(Context, MBB, MBB.getFirstTerminator(), CurrentState, - OutState, Block.PhysLiveRegsAtExit); + if (CurrentState != OutState) { + auto [InsertPt, PhysLiveRegs] = + findStateChangeInsertionPoint(MBB, Block, Block.Insts.end()); + emitStateChange(Context, MBB, InsertPt, CurrentState, OutState, + PhysLiveRegs); + } } } diff --git a/llvm/lib/Target/AMDGPU/SIISelLowering.cpp b/llvm/lib/Target/AMDGPU/SIISelLowering.cpp index 80e985d..a2841c11 100644 --- a/llvm/lib/Target/AMDGPU/SIISelLowering.cpp +++ b/llvm/lib/Target/AMDGPU/SIISelLowering.cpp @@ -18168,7 +18168,7 @@ Align SITargetLowering::getPrefLoopAlignment(MachineLoop *ML) const { return CacheLineAlign; } -LLVM_ATTRIBUTE_UNUSED +[[maybe_unused]] static bool isCopyFromRegOfInlineAsm(const SDNode *N) { assert(N->getOpcode() == ISD::CopyFromReg); do { diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp index a44a247..d516330 100644 --- a/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp +++ b/llvm/lib/Target/AMDGPU/SIInstrInfo.cpp @@ -865,22 +865,16 @@ void SIInstrInfo::copyPhysReg(MachineBasicBlock &MBB, return; } - if (DestReg == AMDGPU::VCC_LO) { - if (AMDGPU::SReg_32RegClass.contains(SrcReg)) { - BuildMI(MBB, MI, DL, get(AMDGPU::S_MOV_B32), AMDGPU::VCC_LO) - .addReg(SrcReg, getKillRegState(KillSrc)); - } else { + if (!AMDGPU::SReg_32RegClass.contains(SrcReg)) { + if (DestReg == AMDGPU::VCC_LO) { // FIXME: Hack until VReg_1 removed. assert(AMDGPU::VGPR_32RegClass.contains(SrcReg)); BuildMI(MBB, MI, DL, get(AMDGPU::V_CMP_NE_U32_e32)) - .addImm(0) - .addReg(SrcReg, getKillRegState(KillSrc)); + .addImm(0) + .addReg(SrcReg, getKillRegState(KillSrc)); + return; } - return; - } - - if (!AMDGPU::SReg_32RegClass.contains(SrcReg)) { reportIllegalCopy(this, MBB, MI, DL, DestReg, SrcReg, KillSrc); return; } @@ -898,22 +892,16 @@ void SIInstrInfo::copyPhysReg(MachineBasicBlock &MBB, return; } - if (DestReg == AMDGPU::VCC) { - if (AMDGPU::SReg_64_EncodableRegClass.contains(SrcReg)) { - BuildMI(MBB, MI, DL, get(AMDGPU::S_MOV_B64), AMDGPU::VCC) - .addReg(SrcReg, getKillRegState(KillSrc)); - } else { + if (!AMDGPU::SReg_64_EncodableRegClass.contains(SrcReg)) { + if (DestReg == AMDGPU::VCC) { // FIXME: Hack until VReg_1 removed. assert(AMDGPU::VGPR_32RegClass.contains(SrcReg)); BuildMI(MBB, MI, DL, get(AMDGPU::V_CMP_NE_U32_e32)) - .addImm(0) - .addReg(SrcReg, getKillRegState(KillSrc)); + .addImm(0) + .addReg(SrcReg, getKillRegState(KillSrc)); + return; } - return; - } - - if (!AMDGPU::SReg_64_EncodableRegClass.contains(SrcReg)) { reportIllegalCopy(this, MBB, MI, DL, DestReg, SrcReg, KillSrc); return; } diff --git a/llvm/lib/Target/AMDGPU/SIPreEmitPeephole.cpp b/llvm/lib/Target/AMDGPU/SIPreEmitPeephole.cpp index c684f9e..01a40c1 100644 --- a/llvm/lib/Target/AMDGPU/SIPreEmitPeephole.cpp +++ b/llvm/lib/Target/AMDGPU/SIPreEmitPeephole.cpp @@ -654,7 +654,6 @@ void SIPreEmitPeephole::collectUnpackingCandidates( if (TotalCyclesBetweenCandidates < NumMFMACycles - 1) InstrsToUnpack.insert(&Instr); } - return; } void SIPreEmitPeephole::performF32Unpacking(MachineInstr &I) { @@ -681,7 +680,6 @@ void SIPreEmitPeephole::performF32Unpacking(MachineInstr &I) { HiDstOp.setIsRenamable(DstOp.isRenamable()); I.eraseFromParent(); - return; } MachineInstrBuilder SIPreEmitPeephole::createUnpackedMI(MachineInstr &I, diff --git a/llvm/lib/Target/ARM/ARMExpandPseudoInsts.cpp b/llvm/lib/Target/ARM/ARMExpandPseudoInsts.cpp index 9945ecc..0d7b6d1 100644 --- a/llvm/lib/Target/ARM/ARMExpandPseudoInsts.cpp +++ b/llvm/lib/Target/ARM/ARMExpandPseudoInsts.cpp @@ -161,8 +161,8 @@ namespace { friend bool operator<(const NEONLdStTableEntry &TE, unsigned PseudoOpc) { return TE.PseudoOpc < PseudoOpc; } - friend bool LLVM_ATTRIBUTE_UNUSED operator<(unsigned PseudoOpc, - const NEONLdStTableEntry &TE) { + [[maybe_unused]] friend bool operator<(unsigned PseudoOpc, + const NEONLdStTableEntry &TE) { return PseudoOpc < TE.PseudoOpc; } }; diff --git a/llvm/lib/Target/CSKY/Disassembler/CSKYDisassembler.cpp b/llvm/lib/Target/CSKY/Disassembler/CSKYDisassembler.cpp index 39e651d..8945ec3 100644 --- a/llvm/lib/Target/CSKY/Disassembler/CSKYDisassembler.cpp +++ b/llvm/lib/Target/CSKY/Disassembler/CSKYDisassembler.cpp @@ -166,7 +166,7 @@ static DecodeStatus DecodeFPR64RegisterClass(MCInst &Inst, uint64_t RegNo, } // TODO -LLVM_ATTRIBUTE_UNUSED +[[maybe_unused]] static DecodeStatus DecodesFPR128RegisterClass(MCInst &Inst, uint64_t RegNo, uint64_t Address, const MCDisassembler *Decoder) { diff --git a/llvm/lib/Target/DirectX/DXContainerGlobals.cpp b/llvm/lib/Target/DirectX/DXContainerGlobals.cpp index ca81d30..8ace2d2 100644 --- a/llvm/lib/Target/DirectX/DXContainerGlobals.cpp +++ b/llvm/lib/Target/DirectX/DXContainerGlobals.cpp @@ -28,6 +28,7 @@ #include "llvm/Support/MD5.h" #include "llvm/TargetParser/Triple.h" #include "llvm/Transforms/Utils/ModuleUtils.h" +#include <cstdint> #include <optional> using namespace llvm; @@ -193,7 +194,12 @@ void DXContainerGlobals::addResourcesForPSV(Module &M, PSVRuntimeInfo &PSV) { dxbc::PSV::v2::ResourceBindInfo BindInfo; BindInfo.Type = Type; BindInfo.LowerBound = Binding.LowerBound; - BindInfo.UpperBound = Binding.LowerBound + Binding.Size - 1; + assert(Binding.Size == UINT32_MAX || + (uint64_t)Binding.LowerBound + Binding.Size - 1 <= UINT32_MAX && + "Resource range is too large"); + BindInfo.UpperBound = (Binding.Size == UINT32_MAX) + ? UINT32_MAX + : Binding.LowerBound + Binding.Size - 1; BindInfo.Space = Binding.Space; BindInfo.Kind = static_cast<dxbc::PSV::ResourceKind>(Kind); BindInfo.Flags = Flags; diff --git a/llvm/lib/Target/DirectX/DXILWriter/DXILBitcodeWriter.cpp b/llvm/lib/Target/DirectX/DXILWriter/DXILBitcodeWriter.cpp index 82c43ff..26a8728 100644 --- a/llvm/lib/Target/DirectX/DXILWriter/DXILBitcodeWriter.cpp +++ b/llvm/lib/Target/DirectX/DXILWriter/DXILBitcodeWriter.cpp @@ -1165,12 +1165,15 @@ void DXILBitcodeWriter::writeValueSymbolTableForwardDecl() {} /// Returns the bit offset to backpatch with the location of the real VST. void DXILBitcodeWriter::writeModuleInfo() { // Emit various pieces of data attached to a module. - if (!M.getTargetTriple().empty()) - writeStringRecord(Stream, bitc::MODULE_CODE_TRIPLE, - M.getTargetTriple().str(), 0 /*TODO*/); - const std::string &DL = M.getDataLayoutStr(); - if (!DL.empty()) - writeStringRecord(Stream, bitc::MODULE_CODE_DATALAYOUT, DL, 0 /*TODO*/); + + // We need to hardcode a triple and datalayout that's compatible with the + // historical DXIL triple and datalayout from DXC. + StringRef Triple = "dxil-ms-dx"; + StringRef DL = "e-m:e-p:32:32-i1:8-i8:8-i16:32-i32:32-i64:64-" + "f16:32-f32:32-f64:64-n8:16:32:64"; + writeStringRecord(Stream, bitc::MODULE_CODE_TRIPLE, Triple, 0 /*TODO*/); + writeStringRecord(Stream, bitc::MODULE_CODE_DATALAYOUT, DL, 0 /*TODO*/); + if (!M.getModuleInlineAsm().empty()) writeStringRecord(Stream, bitc::MODULE_CODE_ASM, M.getModuleInlineAsm(), 0 /*TODO*/); diff --git a/llvm/lib/Target/DirectX/DXILWriter/DXILWriterPass.cpp b/llvm/lib/Target/DirectX/DXILWriter/DXILWriterPass.cpp index 1eb03bf..725f2b1 100644 --- a/llvm/lib/Target/DirectX/DXILWriter/DXILWriterPass.cpp +++ b/llvm/lib/Target/DirectX/DXILWriter/DXILWriterPass.cpp @@ -149,11 +149,6 @@ public: std::string Data; llvm::raw_string_ostream OS(Data); - Triple OriginalTriple = M.getTargetTriple(); - // Set to DXIL triple when write to bitcode. - // Only the output bitcode need to be DXIL triple. - M.setTargetTriple(Triple("dxil-ms-dx")); - // Perform late legalization of lifetime intrinsics that would otherwise // fail the Module Verifier if performed in an earlier pass legalizeLifetimeIntrinsics(M); @@ -165,9 +160,6 @@ public: // not-so-legal legalizations removeLifetimeIntrinsics(M); - // Recover triple. - M.setTargetTriple(OriginalTriple); - Constant *ModuleConstant = ConstantDataArray::get(M.getContext(), arrayRefFromStringRef(Data)); auto *GV = new llvm::GlobalVariable(M, ModuleConstant->getType(), true, diff --git a/llvm/lib/Target/Hexagon/Disassembler/HexagonDisassembler.cpp b/llvm/lib/Target/Hexagon/Disassembler/HexagonDisassembler.cpp index 974f653..5f180d6 100644 --- a/llvm/lib/Target/Hexagon/Disassembler/HexagonDisassembler.cpp +++ b/llvm/lib/Target/Hexagon/Disassembler/HexagonDisassembler.cpp @@ -667,11 +667,10 @@ static DecodeStatus DecodeHvxWRRegisterClass(MCInst &Inst, unsigned RegNo, return DecodeRegisterClass(Inst, RegNo, HvxWRDecoderTable); } -LLVM_ATTRIBUTE_UNUSED // Suppress warning temporarily. - static DecodeStatus - DecodeHvxVQRRegisterClass(MCInst &Inst, unsigned RegNo, - uint64_t /*Address*/, - const MCDisassembler *Decoder) { +[[maybe_unused]] // Suppress warning temporarily. +static DecodeStatus DecodeHvxVQRRegisterClass(MCInst &Inst, unsigned RegNo, + uint64_t /*Address*/, + const MCDisassembler *Decoder) { static const MCPhysReg HvxVQRDecoderTable[] = { Hexagon::VQ0, Hexagon::VQ1, Hexagon::VQ2, Hexagon::VQ3, Hexagon::VQ4, Hexagon::VQ5, Hexagon::VQ6, Hexagon::VQ7}; diff --git a/llvm/lib/Target/Hexagon/HexagonBitSimplify.cpp b/llvm/lib/Target/Hexagon/HexagonBitSimplify.cpp index 52e6b0b..68f5312 100644 --- a/llvm/lib/Target/Hexagon/HexagonBitSimplify.cpp +++ b/llvm/lib/Target/Hexagon/HexagonBitSimplify.cpp @@ -174,8 +174,8 @@ namespace { const TargetRegisterInfo *TRI; }; - raw_ostream &operator<< (raw_ostream &OS, const PrintRegSet &P) - LLVM_ATTRIBUTE_UNUSED; + [[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, + const PrintRegSet &P); raw_ostream &operator<< (raw_ostream &OS, const PrintRegSet &P) { OS << '{'; for (unsigned R = P.RS.find_first(); R; R = P.RS.find_next(R)) diff --git a/llvm/lib/Target/Hexagon/HexagonCommonGEP.cpp b/llvm/lib/Target/Hexagon/HexagonCommonGEP.cpp index 14b6bb3..9087f9d 100644 --- a/llvm/lib/Target/Hexagon/HexagonCommonGEP.cpp +++ b/llvm/lib/Target/Hexagon/HexagonCommonGEP.cpp @@ -272,15 +272,14 @@ namespace { OS << *I << ' ' << **I << '\n'; } - raw_ostream &operator<< (raw_ostream &OS, - const NodeVect &S) LLVM_ATTRIBUTE_UNUSED; + [[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, const NodeVect &S); raw_ostream &operator<< (raw_ostream &OS, const NodeVect &S) { dump_node_container(OS, S); return OS; } - raw_ostream &operator<< (raw_ostream &OS, - const NodeToUsesMap &M) LLVM_ATTRIBUTE_UNUSED; + [[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, + const NodeToUsesMap &M); raw_ostream &operator<< (raw_ostream &OS, const NodeToUsesMap &M){ for (const auto &I : M) { const UseSet &Us = I.second; @@ -914,9 +913,8 @@ namespace { const NodeToValueMap ⤅ }; - raw_ostream &operator<< (raw_ostream &OS, - const LocationAsBlock &Loc) LLVM_ATTRIBUTE_UNUSED ; - raw_ostream &operator<< (raw_ostream &OS, const LocationAsBlock &Loc) { + [[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, + const LocationAsBlock &Loc) { for (const auto &I : Loc.Map) { OS << I.first << " -> "; if (BasicBlock *B = cast_or_null<BasicBlock>(I.second)) diff --git a/llvm/lib/Target/Hexagon/HexagonConstExtenders.cpp b/llvm/lib/Target/Hexagon/HexagonConstExtenders.cpp index 5dde47a..a3296e0 100644 --- a/llvm/lib/Target/Hexagon/HexagonConstExtenders.cpp +++ b/llvm/lib/Target/Hexagon/HexagonConstExtenders.cpp @@ -419,8 +419,8 @@ namespace { using HCE = HexagonConstExtenders; - LLVM_ATTRIBUTE_UNUSED - raw_ostream &operator<< (raw_ostream &OS, const OffsetRange &OR) { + [[maybe_unused]] + raw_ostream &operator<<(raw_ostream &OS, const OffsetRange &OR) { if (OR.Min > OR.Max) OS << '!'; OS << '[' << OR.Min << ',' << OR.Max << "]a" << unsigned(OR.Align) @@ -435,8 +435,8 @@ namespace { const HexagonRegisterInfo &HRI; }; - LLVM_ATTRIBUTE_UNUSED - raw_ostream &operator<< (raw_ostream &OS, const PrintRegister &P) { + [[maybe_unused]] + raw_ostream &operator<<(raw_ostream &OS, const PrintRegister &P) { if (P.Rs.Reg != 0) OS << printReg(P.Rs.Reg, &P.HRI, P.Rs.Sub); else @@ -451,8 +451,8 @@ namespace { const HexagonRegisterInfo &HRI; }; - LLVM_ATTRIBUTE_UNUSED - raw_ostream &operator<< (raw_ostream &OS, const PrintExpr &P) { + [[maybe_unused]] + raw_ostream &operator<<(raw_ostream &OS, const PrintExpr &P) { OS << "## " << (P.Ex.Neg ? "- " : "+ "); if (P.Ex.Rs.Reg != 0) OS << printReg(P.Ex.Rs.Reg, &P.HRI, P.Ex.Rs.Sub); @@ -469,15 +469,15 @@ namespace { const HexagonRegisterInfo &HRI; }; - LLVM_ATTRIBUTE_UNUSED - raw_ostream &operator<< (raw_ostream &OS, const PrintInit &P) { + [[maybe_unused]] + raw_ostream &operator<<(raw_ostream &OS, const PrintInit &P) { OS << '[' << P.ExtI.first << ", " << PrintExpr(P.ExtI.second, P.HRI) << ']'; return OS; } - LLVM_ATTRIBUTE_UNUSED - raw_ostream &operator<< (raw_ostream &OS, const HCE::ExtDesc &ED) { + [[maybe_unused]] + raw_ostream &operator<<(raw_ostream &OS, const HCE::ExtDesc &ED) { assert(ED.OpNum != -1u); const MachineBasicBlock &MBB = *ED.getOp().getParent()->getParent(); const MachineFunction &MF = *MBB.getParent(); @@ -493,8 +493,8 @@ namespace { return OS; } - LLVM_ATTRIBUTE_UNUSED - raw_ostream &operator<< (raw_ostream &OS, const HCE::ExtRoot &ER) { + [[maybe_unused]] + raw_ostream &operator<<(raw_ostream &OS, const HCE::ExtRoot &ER) { switch (ER.Kind) { case MachineOperand::MO_Immediate: OS << "imm:" << ER.V.ImmVal; @@ -527,8 +527,8 @@ namespace { return OS; } - LLVM_ATTRIBUTE_UNUSED - raw_ostream &operator<< (raw_ostream &OS, const HCE::ExtValue &EV) { + [[maybe_unused]] + raw_ostream &operator<<(raw_ostream &OS, const HCE::ExtValue &EV) { OS << HCE::ExtRoot(EV) << " off:" << EV.Offset; return OS; } @@ -540,8 +540,8 @@ namespace { const HexagonRegisterInfo &HRI; }; - LLVM_ATTRIBUTE_UNUSED - raw_ostream &operator<< (raw_ostream &OS, const PrintIMap &P) { + [[maybe_unused]] + raw_ostream &operator<<(raw_ostream &OS, const PrintIMap &P) { OS << "{\n"; for (const std::pair<const HCE::ExtenderInit, HCE::IndexList> &Q : P.IMap) { OS << " " << PrintInit(Q.first, P.HRI) << " -> {"; diff --git a/llvm/lib/Target/Hexagon/HexagonEarlyIfConv.cpp b/llvm/lib/Target/Hexagon/HexagonEarlyIfConv.cpp index 14a7ae7..3900aac 100644 --- a/llvm/lib/Target/Hexagon/HexagonEarlyIfConv.cpp +++ b/llvm/lib/Target/Hexagon/HexagonEarlyIfConv.cpp @@ -132,8 +132,7 @@ namespace { const TargetRegisterInfo &TRI; friend raw_ostream &operator<< (raw_ostream &OS, const PrintFP &P); }; - raw_ostream &operator<<(raw_ostream &OS, - const PrintFP &P) LLVM_ATTRIBUTE_UNUSED; + [[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, const PrintFP &P); raw_ostream &operator<<(raw_ostream &OS, const PrintFP &P) { OS << "{ SplitB:" << PrintMB(P.FP.SplitB) << ", PredR:" << printReg(P.FP.PredR, &P.TRI) diff --git a/llvm/lib/Target/Hexagon/HexagonGenPredicate.cpp b/llvm/lib/Target/Hexagon/HexagonGenPredicate.cpp index f9fdab4..9c81e96 100644 --- a/llvm/lib/Target/Hexagon/HexagonGenPredicate.cpp +++ b/llvm/lib/Target/Hexagon/HexagonGenPredicate.cpp @@ -51,11 +51,11 @@ private: const TargetRegisterInfo &TRI; }; - raw_ostream &operator<< (raw_ostream &OS, const PrintRegister &PR) - LLVM_ATTRIBUTE_UNUSED; - raw_ostream &operator<< (raw_ostream &OS, const PrintRegister &PR) { - return OS << printReg(PR.Reg.Reg, &PR.TRI, PR.Reg.SubReg); - } +[[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, + const PrintRegister &PR); +raw_ostream &operator<<(raw_ostream &OS, const PrintRegister &PR) { + return OS << printReg(PR.Reg.Reg, &PR.TRI, PR.Reg.SubReg); +} class HexagonGenPredicate : public MachineFunctionPass { public: diff --git a/llvm/lib/Target/Hexagon/HexagonISelDAGToDAGHVX.cpp b/llvm/lib/Target/Hexagon/HexagonISelDAGToDAGHVX.cpp index 4d96cfa..c7a4f68 100644 --- a/llvm/lib/Target/Hexagon/HexagonISelDAGToDAGHVX.cpp +++ b/llvm/lib/Target/Hexagon/HexagonISelDAGToDAGHVX.cpp @@ -789,7 +789,7 @@ struct ShuffleMask { } }; -LLVM_ATTRIBUTE_UNUSED +[[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, const ShuffleMask &SM) { SM.print(OS); return OS; diff --git a/llvm/lib/Target/Hexagon/HexagonVectorCombine.cpp b/llvm/lib/Target/Hexagon/HexagonVectorCombine.cpp index 87d052b..e4c0a16 100644 --- a/llvm/lib/Target/Hexagon/HexagonVectorCombine.cpp +++ b/llvm/lib/Target/Hexagon/HexagonVectorCombine.cpp @@ -364,7 +364,7 @@ private: const HexagonVectorCombine &HVC; }; -LLVM_ATTRIBUTE_UNUSED +[[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, const AlignVectors::AddrInfo &AI) { OS << "Inst: " << AI.Inst << " " << *AI.Inst << '\n'; OS << "Addr: " << *AI.Addr << '\n'; @@ -375,7 +375,7 @@ raw_ostream &operator<<(raw_ostream &OS, const AlignVectors::AddrInfo &AI) { return OS; } -LLVM_ATTRIBUTE_UNUSED +[[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, const AlignVectors::MoveGroup &MG) { OS << "IsLoad:" << (MG.IsLoad ? "yes" : "no"); OS << ", IsHvx:" << (MG.IsHvx ? "yes" : "no") << '\n'; @@ -394,7 +394,7 @@ raw_ostream &operator<<(raw_ostream &OS, const AlignVectors::MoveGroup &MG) { return OS; } -LLVM_ATTRIBUTE_UNUSED +[[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, const AlignVectors::ByteSpan::Block &B) { OS << " @" << B.Pos << " [" << B.Seg.Start << ',' << B.Seg.Size << "] "; @@ -408,7 +408,7 @@ raw_ostream &operator<<(raw_ostream &OS, return OS; } -LLVM_ATTRIBUTE_UNUSED +[[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, const AlignVectors::ByteSpan &BS) { OS << "ByteSpan[size=" << BS.size() << ", extent=" << BS.extent() << '\n'; for (const AlignVectors::ByteSpan::Block &B : BS) diff --git a/llvm/lib/Target/Hexagon/HexagonVectorLoopCarriedReuse.cpp b/llvm/lib/Target/Hexagon/HexagonVectorLoopCarriedReuse.cpp index fa8ae60..2ff5843 100644 --- a/llvm/lib/Target/Hexagon/HexagonVectorLoopCarriedReuse.cpp +++ b/llvm/lib/Target/Hexagon/HexagonVectorLoopCarriedReuse.cpp @@ -111,7 +111,7 @@ namespace { friend raw_ostream &operator<< (raw_ostream &OS, const DepChain &D); }; - LLVM_ATTRIBUTE_UNUSED + [[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, const DepChain &D) { const ChainOfDependences &CD = D.Chain; int ChainSize = CD.size(); @@ -144,7 +144,7 @@ namespace { bool isDefined() { return Inst2Replace != nullptr; } }; - LLVM_ATTRIBUTE_UNUSED + [[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, const ReuseValue &RU) { OS << "** ReuseValue ***\n"; OS << "Instruction to Replace: " << *(RU.Inst2Replace) << "\n"; diff --git a/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonBaseInfo.h b/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonBaseInfo.h index ca98269..e3094b4 100644 --- a/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonBaseInfo.h +++ b/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonBaseInfo.h @@ -275,7 +275,7 @@ namespace HexagonII { INST_ICLASS_ALU32_3 = 0xf0000000 }; - LLVM_ATTRIBUTE_UNUSED + [[maybe_unused]] static unsigned getMemAccessSizeInBytes(MemAccessSize S) { switch (S) { case ByteAccess: return 1; diff --git a/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonMCTargetDesc.cpp b/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonMCTargetDesc.cpp index bfea50e..6b48a21 100644 --- a/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonMCTargetDesc.cpp +++ b/llvm/lib/Target/Hexagon/MCTargetDesc/HexagonMCTargetDesc.cpp @@ -422,12 +422,12 @@ static MCTargetStreamer *createHexagonNullTargetStreamer(MCStreamer &S) { return new HexagonTargetStreamer(S); } -static void LLVM_ATTRIBUTE_UNUSED clearFeature(MCSubtargetInfo* STI, uint64_t F) { +[[maybe_unused]] static void clearFeature(MCSubtargetInfo *STI, uint64_t F) { if (STI->hasFeature(F)) STI->ToggleFeature(F); } -static bool LLVM_ATTRIBUTE_UNUSED checkFeature(MCSubtargetInfo* STI, uint64_t F) { +[[maybe_unused]] static bool checkFeature(MCSubtargetInfo *STI, uint64_t F) { return STI->hasFeature(F); } diff --git a/llvm/lib/Target/PowerPC/PPCISelLowering.cpp b/llvm/lib/Target/PowerPC/PPCISelLowering.cpp index 944a1e2..8bf0d11 100644 --- a/llvm/lib/Target/PowerPC/PPCISelLowering.cpp +++ b/llvm/lib/Target/PowerPC/PPCISelLowering.cpp @@ -9702,6 +9702,10 @@ SDValue PPCTargetLowering::LowerBUILD_VECTOR(SDValue Op, } return SDV; } + // Recognize build vector patterns to emit VSX vector instructions + // instead of loading value from memory. + if (SDValue VecPat = combineBVLoadsSpecialValue(Op, DAG)) + return VecPat; } // Check if this is a splat of a constant value. APInt APSplatBits, APSplatUndef; @@ -15696,6 +15700,142 @@ combineElementTruncationToVectorTruncation(SDNode *N, return SDValue(); } +// LXVKQ instruction load VSX vector with a special quadword value +// based on an immediate value. This helper method returns the details of the +// match as a tuple of {LXVKQ unsigned IMM Value, right_shift_amount} +// to help generate the LXVKQ instruction and the subsequent shift instruction +// required to match the original build vector pattern. + +// LXVKQPattern: {LXVKQ unsigned IMM Value, right_shift_amount} +using LXVKQPattern = std::tuple<uint32_t, uint8_t>; + +static std::optional<LXVKQPattern> getPatternInfo(const APInt &FullVal) { + + // LXVKQ instruction loads the Quadword value: + // 0x8000_0000_0000_0000_0000_0000_0000_0000 when imm = 0b10000 + static const APInt BasePattern = APInt(128, 0x8000000000000000ULL) << 64; + static const uint32_t Uim = 16; + + // Check for direct LXVKQ match (no shift needed) + if (FullVal == BasePattern) + return std::make_tuple(Uim, uint8_t{0}); + + // Check if FullValue is 1 (the result of the base pattern >> 127) + if (FullVal == APInt(128, 1)) + return std::make_tuple(Uim, uint8_t{127}); + + return std::nullopt; +} + +/// Combine vector loads to a single load (using lxvkq) or splat with shift of a +/// constant (xxspltib + vsrq) by recognising patterns in the Build Vector. +/// LXVKQ instruction load VSX vector with a special quadword value based on an +/// immediate value. if UIM=0b10000 then LXVKQ loads VSR[32×TX+T] with value +/// 0x8000_0000_0000_0000_0000_0000_0000_0000. +/// This can be used to inline the build vector constants that have the +/// following patterns: +/// +/// 0x8000_0000_0000_0000_0000_0000_0000_0000 (MSB set pattern) +/// 0x0000_0000_0000_0000_0000_0000_0000_0001 (LSB set pattern) +/// MSB pattern can directly loaded using LXVKQ while LSB is loaded using a +/// combination of splatting and right shift instructions. + +SDValue PPCTargetLowering::combineBVLoadsSpecialValue(SDValue Op, + SelectionDAG &DAG) const { + + assert((Op.getNode() && Op.getOpcode() == ISD::BUILD_VECTOR) && + "Expected a BuildVectorSDNode in combineBVLoadsSpecialValue"); + + // This transformation is only supported if we are loading either a byte, + // halfword, word, or doubleword. + EVT VT = Op.getValueType(); + if (!(VT == MVT::v8i16 || VT == MVT::v16i8 || VT == MVT::v4i32 || + VT == MVT::v2i64)) + return SDValue(); + + LLVM_DEBUG(llvm::dbgs() << "\ncombineBVLoadsSpecialValue: Build vector (" + << VT.getEVTString() << "): "; + Op->dump()); + + unsigned NumElems = VT.getVectorNumElements(); + unsigned ElemBits = VT.getScalarSizeInBits(); + + bool IsLittleEndian = DAG.getDataLayout().isLittleEndian(); + + // Check for Non-constant operand in the build vector. + for (const SDValue &Operand : Op.getNode()->op_values()) { + if (!isa<ConstantSDNode>(Operand)) + return SDValue(); + } + + // Assemble build vector operands as a 128-bit register value + // We need to reconstruct what the 128-bit register pattern would be + // that produces this vector when interpreted with the current endianness + APInt FullVal = APInt::getZero(128); + + for (unsigned Index = 0; Index < NumElems; ++Index) { + auto *C = cast<ConstantSDNode>(Op.getOperand(Index)); + + // Get element value as raw bits (zero-extended) + uint64_t ElemValue = C->getZExtValue(); + + // Mask to element size to ensure we only get the relevant bits + if (ElemBits < 64) + ElemValue &= ((1ULL << ElemBits) - 1); + + // Calculate bit position for this element in the 128-bit register + unsigned BitPos = + (IsLittleEndian) ? (Index * ElemBits) : (128 - (Index + 1) * ElemBits); + + // Create APInt for the element value and shift it to correct position + APInt ElemAPInt(128, ElemValue); + ElemAPInt <<= BitPos; + + // Place the element value at the correct bit position + FullVal |= ElemAPInt; + } + + if (FullVal.isZero() || FullVal.isAllOnes()) + return SDValue(); + + if (auto UIMOpt = getPatternInfo(FullVal)) { + const auto &[Uim, ShiftAmount] = *UIMOpt; + SDLoc Dl(Op); + + // Generate LXVKQ instruction if the shift amount is zero. + if (ShiftAmount == 0) { + SDValue UimVal = DAG.getTargetConstant(Uim, Dl, MVT::i32); + SDValue LxvkqInstr = + SDValue(DAG.getMachineNode(PPC::LXVKQ, Dl, VT, UimVal), 0); + LLVM_DEBUG(llvm::dbgs() + << "combineBVLoadsSpecialValue: Instruction Emitted "; + LxvkqInstr.dump()); + return LxvkqInstr; + } + + assert(ShiftAmount == 127 && "Unexpected lxvkq shift amount value"); + + // The right shifted pattern can be constructed using a combination of + // XXSPLTIB and VSRQ instruction. VSRQ uses the shift amount from the lower + // 7 bits of byte 15. This can be specified using XXSPLTIB with immediate + // value 255. + SDValue ShiftAmountVec = + SDValue(DAG.getMachineNode(PPC::XXSPLTIB, Dl, MVT::v4i32, + DAG.getTargetConstant(255, Dl, MVT::i32)), + 0); + // Generate appropriate right shift instruction + SDValue ShiftVec = SDValue( + DAG.getMachineNode(PPC::VSRQ, Dl, VT, ShiftAmountVec, ShiftAmountVec), + 0); + LLVM_DEBUG(llvm::dbgs() + << "\n combineBVLoadsSpecialValue: Instruction Emitted "; + ShiftVec.dump()); + return ShiftVec; + } + // No patterns matched for build vectors. + return SDValue(); +} + /// Reduce the number of loads when building a vector. /// /// Building a vector out of multiple loads can be converted to a load diff --git a/llvm/lib/Target/PowerPC/PPCISelLowering.h b/llvm/lib/Target/PowerPC/PPCISelLowering.h index 59f3387..880aca7 100644 --- a/llvm/lib/Target/PowerPC/PPCISelLowering.h +++ b/llvm/lib/Target/PowerPC/PPCISelLowering.h @@ -1472,6 +1472,9 @@ namespace llvm { combineElementTruncationToVectorTruncation(SDNode *N, DAGCombinerInfo &DCI) const; + SDValue combineBVLoadsSpecialValue(SDValue Operand, + SelectionDAG &DAG) const; + /// lowerToVINSERTH - Return the SDValue if this VECTOR_SHUFFLE can be /// handled by the VINSERTH instruction introduced in ISA 3.0. This is /// essentially any shuffle of v8i16 vectors that just inserts one element diff --git a/llvm/lib/Target/PowerPC/PPCInstrP10.td b/llvm/lib/Target/PowerPC/PPCInstrP10.td index 2384959..2d8c633 100644 --- a/llvm/lib/Target/PowerPC/PPCInstrP10.td +++ b/llvm/lib/Target/PowerPC/PPCInstrP10.td @@ -2404,6 +2404,190 @@ multiclass XXEvalTernarySelectOr<ValueType Vt> { 126>; } +// ============================================================================= +// XXEVAL Ternary Pattern Multiclass: XXEvalTernarySelectNor +// This class matches the equivalent Ternary Operation: A ? f(B,C) : NOR(B,C) +// and emit the corresponding xxeval instruction with the imm value. +// +// The patterns implement xxeval vector select operations where: +// - A is the selector vector +// - f(B,C) is the "true" case op in set {B, C, AND(B,C), XOR(B,C), NOT(C), +// NOT(B), NAND(B,C)} +// - C is the "false" case op NOR(B,C) +// ============================================================================= +multiclass XXEvalTernarySelectNor<ValueType Vt>{ + // Pattern: (A ? AND(B,C) : NOR(B,C)) XXEVAL immediate value: 129 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VAnd Vt:$vB, Vt:$vC), (VNor Vt:$vB, Vt:$vC)), + 129>; + + // Pattern: (A ? B : NOR(B,C)) XXEVAL immediate value: 131 + def : XXEvalPattern<Vt, (vselect Vt:$vA, Vt:$vB, (VNor Vt:$vB, Vt:$vC)),131>; + + // Pattern: (A ? C : NOR(B,C)) XXEVAL immediate value: 133 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, Vt:$vC, (VNor Vt:$vB, Vt:$vC)), + 133>; + + // Pattern: (A ? XOR(B,C) : NOR(B,C)) XXEVAL immediate value: 134 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VXor Vt:$vB, Vt:$vC), (VNor Vt:$vB, Vt:$vC)), + 134>; + + // Pattern: (A ? NOT(C) : NOR(B,C)) XXEVAL immediate value: 138 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VNot Vt:$vC), (VNor Vt:$vB, Vt:$vC)), + 138>; + + // Pattern: (A ? NOT(B) : NOR(B,C)) XXEVAL immediate value: 140 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VNot Vt:$vB), (VNor Vt:$vB, Vt:$vC)), + 140>; + + // Pattern: (A ? NAND(B,C) : NOR(B,C)) XXEVAL immediate value: 142 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VNand Vt:$vB, Vt:$vC), (VNor Vt:$vB, Vt:$vC)), + 142>; +} + +// ============================================================================= +// XXEVAL Ternary Pattern Multiclass: XXEvalTernarySelectEqv +// This class matches the equivalent Ternary Operation: A ? f(B,C) : EQV(B,C) +// and emit the corresponding xxeval instruction with the imm value. +// +// The patterns implement xxeval vector select operations where: +// - A is the selector vector +// - f(B,C) is the "true" case op in set {OR(B,C), NOR(B,C), NAND(B,C), NOT(B), +// NOT(C)} +// - C is the "false" case op EQV(B,C) +// ============================================================================= +multiclass XXEvalTernarySelectEqv<ValueType Vt>{ + // Pattern: (A ? OR(B,C) : EQV(B,C)) XXEVAL immediate value: 151 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VOr Vt:$vB, Vt:$vC), (VEqv Vt:$vB, Vt:$vC)), + 151>; + + // Pattern: (A ? NOR(B,C) : EQV(B,C)) XXEVAL immediate value: 152 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VNor Vt:$vB, Vt:$vC), (VEqv Vt:$vB, Vt:$vC)), + 152>; + + // Pattern: (A ? NOT(C) : EQV(B,C)) XXEVAL immediate value: 154 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VNot Vt:$vC), (VEqv Vt:$vB, Vt:$vC)), + 154>; + + // Pattern: (A ? NAND(B,C) : EQV(B,C)) XXEVAL immediate value: 158 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VNand Vt:$vB, Vt:$vC), (VEqv Vt:$vB, Vt:$vC)), + 158>; +} + +// ============================================================================= +// XXEVAL Ternary Pattern Multiclass: XXEvalTernarySelectNotC +// This class matches the equivalent Ternary Operation: A ? f(B,C) : NOT(C) +// and emit the corresponding xxeval instruction with the imm value. +// +// The patterns implement xxeval vector select operations where: +// - A is the selector vector +// - f(B,C) is the "true" case op in set {AND(B,C), OR(B,C), XOR(B,C), NAND(B,C), +// B, NOT(B)} +// - C is the "false" case op NOT(C) +// ============================================================================= +multiclass XXEvalTernarySelectNotC<ValueType Vt>{ + // Pattern: (A ? AND(B,C) : NOT(C)) XXEVAL immediate value: 161 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VAnd Vt:$vB, Vt:$vC), (VNot Vt:$vC)), 161>; + + // Pattern: (A ? B : NOT(C)) XXEVAL immediate value: 163 + def : XXEvalPattern<Vt, (vselect Vt:$vA, Vt:$vB, (VNot Vt:$vC)), 163>; + + // Pattern: (A ? XOR(B,C) : NOT(C)) XXEVAL immediate value: 166 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VXor Vt:$vB, Vt:$vC), (VNot Vt:$vC)), 166>; + + // Pattern: (A ? OR(B,C) : NOT(C)) XXEVAL immediate value: 167 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VOr Vt:$vB, Vt:$vC), (VNot Vt:$vC)), 167>; + + // Pattern: (A ? NOT(B) : NOT(C)) XXEVAL immediate value: 172 + def : XXEvalPattern<Vt, (vselect Vt:$vA, (VNot Vt:$vB), (VNot Vt:$vC)), 172>; + + // Pattern: (A ? NAND(B,C) : NOT(C)) XXEVAL immediate value: 174 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VNand Vt:$vB, Vt:$vC), (VNot Vt:$vC)), 174>; +} + +// ============================================================================= +// XXEVAL Ternary Pattern Multiclass: XXEvalTernarySelectNotB +// This class matches the equivalent Ternary Operation: A ? f(B,C) : NOT(B) +// and emit the corresponding xxeval instruction with the imm value. +// +// The patterns implement xxeval vector select operations where: +// - A is the selector vector +// - f(B,C) is the "true" case op in set {AND(B,C), OR(B,C), XOR(B,C), NAND(B,C), +// C, NOT(B)} +// - C is the "false" case op NOT(B) +// ============================================================================= +multiclass XXEvalTernarySelectNotB<ValueType Vt>{ + // Pattern: (A ? AND(B,C) : NOT(B)) XXEVAL immediate value: 193 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VAnd Vt:$vB, Vt:$vC), (VNot Vt:$vB)), 193>; + + // Pattern: (A ? C : NOT(B)) XXEVAL immediate value: 197 + def : XXEvalPattern<Vt, (vselect Vt:$vA, Vt:$vC, (VNot Vt:$vB)), 197>; + + // Pattern: (A ? XOR(B,C) : NOT(B)) XXEVAL immediate value: 198 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VXor Vt:$vB, Vt:$vC), (VNot Vt:$vB)), 198>; + + // Pattern: (A ? OR(B,C) : NOT(B)) XXEVAL immediate value: 199 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VOr Vt:$vB, Vt:$vC), (VNot Vt:$vB)), 199>; + + // Pattern: (A ? NOT(C) : NOT(B)) XXEVAL immediate value: 202 + def : XXEvalPattern<Vt, (vselect Vt:$vA, (VNot Vt:$vC), (VNot Vt:$vB)), 202>; + + // Pattern: (A ? NAND(B,C) : NOT(B)) XXEVAL immediate value: 206 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VNand Vt:$vB, Vt:$vC), (VNot Vt:$vB)), 206>; +} + +// ============================================================================= +// XXEVAL Ternary Pattern Multiclass: XXEvalTernarySelectNand +// This class matches the equivalent Ternary Operation: A ? f(B,C) : NAND(B,C) +// and emit the corresponding xxeval instruction with the imm value. +// +// The patterns implement xxeval vector select operations where: +// - A is the selector vector +// - f(B,C) is the "true" case op in set {B, C, XOR(B,C), OR(B,C), EQV(B,C)} +// - C is the "false" case op NAND(B,C) +// ============================================================================= +multiclass XXEvalTernarySelectNand<ValueType Vt>{ + // Pattern: (A ? B : NAND(B,C)) XXEVAL immediate value: 227 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, Vt:$vB, (VNand Vt:$vB, Vt:$vC)), 227>; + + // Pattern: (A ? C : NAND(B,C)) XXEVAL immediate value: 229 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, Vt:$vC, (VNand Vt:$vB, Vt:$vC)), 229>; + + // Pattern: (A ? XOR(B,C) : NAND(B,C)) XXEVAL immediate value: 230 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VXor Vt:$vB, Vt:$vC), (VNand Vt:$vB, Vt:$vC)), + 230>; + + // Pattern: (A ? OR(B,C) : NAND(B,C)) XXEVAL immediate value: 231 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VOr Vt:$vB, Vt:$vC), (VNand Vt:$vB, Vt:$vC)), + 231>; + + // Pattern: (A ? EQV(B,C) : NAND(B,C)) XXEVAL immediate value: 233 + def : XXEvalPattern< + Vt, (vselect Vt:$vA, (VEqv Vt:$vB, Vt:$vC), (VNand Vt:$vB, Vt:$vC)), + 233>; +} + let Predicates = [PrefixInstrs, HasP10Vector] in { let AddedComplexity = 400 in { def : Pat<(v4i32 (build_vector i32immNonAllOneNonZero:$A, @@ -2519,6 +2703,11 @@ let Predicates = [PrefixInstrs, HasP10Vector] in { defm : XXEvalTernarySelectC<Ty>; defm : XXEvalTernarySelectXor<Ty>; defm : XXEvalTernarySelectOr<Ty>; + defm : XXEvalTernarySelectNor<Ty>; + defm : XXEvalTernarySelectEqv<Ty>; + defm : XXEvalTernarySelectNotC<Ty>; + defm : XXEvalTernarySelectNotB<Ty>; + defm : XXEvalTernarySelectNand<Ty>; } // Anonymous patterns to select prefixed VSX loads and stores. diff --git a/llvm/lib/Target/RISCV/AsmParser/RISCVAsmParser.cpp b/llvm/lib/Target/RISCV/AsmParser/RISCVAsmParser.cpp index 8851a0f..e857b2d 100644 --- a/llvm/lib/Target/RISCV/AsmParser/RISCVAsmParser.cpp +++ b/llvm/lib/Target/RISCV/AsmParser/RISCVAsmParser.cpp @@ -3356,10 +3356,10 @@ bool RISCVAsmParser::parseDirectiveAttribute() { bool isValidInsnFormat(StringRef Format, const MCSubtargetInfo &STI) { return StringSwitch<bool>(Format) - .Cases("r", "r4", "i", "b", "sb", "u", "j", "uj", "s", true) - .Cases("cr", "ci", "ciw", "css", "cl", "cs", "ca", "cb", "cj", + .Cases({"r", "r4", "i", "b", "sb", "u", "j", "uj", "s"}, true) + .Cases({"cr", "ci", "ciw", "css", "cl", "cs", "ca", "cb", "cj"}, STI.hasFeature(RISCV::FeatureStdExtZca)) - .Cases("qc.eai", "qc.ei", "qc.eb", "qc.ej", "qc.es", + .Cases({"qc.eai", "qc.ei", "qc.eb", "qc.ej", "qc.es"}, !STI.hasFeature(RISCV::Feature64Bit)) .Default(false); } diff --git a/llvm/lib/Target/RISCV/MCA/RISCVCustomBehaviour.cpp b/llvm/lib/Target/RISCV/MCA/RISCVCustomBehaviour.cpp index 50730c6..ab93bba 100644 --- a/llvm/lib/Target/RISCV/MCA/RISCVCustomBehaviour.cpp +++ b/llvm/lib/Target/RISCV/MCA/RISCVCustomBehaviour.cpp @@ -43,7 +43,7 @@ const llvm::StringRef RISCVLMULInstrument::DESC_NAME = "RISCV-LMUL"; bool RISCVLMULInstrument::isDataValid(llvm::StringRef Data) { // Return true if not one of the valid LMUL strings return StringSwitch<bool>(Data) - .Cases("M1", "M2", "M4", "M8", "MF2", "MF4", "MF8", true) + .Cases({"M1", "M2", "M4", "M8", "MF2", "MF4", "MF8"}, true) .Default(false); } diff --git a/llvm/lib/Target/RISCV/RISCVSchedSiFive7.td b/llvm/lib/Target/RISCV/RISCVSchedSiFive7.td index f863392a..637d61fe 100644 --- a/llvm/lib/Target/RISCV/RISCVSchedSiFive7.td +++ b/llvm/lib/Target/RISCV/RISCVSchedSiFive7.td @@ -270,7 +270,7 @@ class SiFive7AnyToGPRBypass<SchedRead read, int cycles = 2> // and floating point computation. // The V pipeline is modeled by the VCQ, VA, VL, and VS resources. There can // be one or two VA (Vector Arithmetic). -multiclass SiFive7ProcResources<bit extraVALU = false> { +multiclass SiFive7ProcResources<bit dualVALU = false> { let BufferSize = 0 in { def PipeA : ProcResource<1>; def PipeB : ProcResource<1>; @@ -279,7 +279,7 @@ multiclass SiFive7ProcResources<bit extraVALU = false> { def FDiv : ProcResource<1>; // FP Division/Sqrt // Arithmetic sequencer(s) - if extraVALU then { + if dualVALU then { // VA1 can handle any vector airthmetic instruction. def VA1 : ProcResource<1>; // VA2 generally can only handle simple vector arithmetic. @@ -305,7 +305,7 @@ multiclass SiFive7ProcResources<bit extraVALU = false> { def PipeAB : ProcResGroup<[!cast<ProcResource>(NAME#"PipeA"), !cast<ProcResource>(NAME#"PipeB")]>; - if extraVALU then + if dualVALU then def VA1OrVA2 : ProcResGroup<[!cast<ProcResource>(NAME#"VA1"), !cast<ProcResource>(NAME#"VA2")]>; } @@ -1550,10 +1550,10 @@ multiclass SiFive7ReadAdvance { /// This multiclass is a "bundle" of (1) processor resources (i.e. pipes) and /// (2) WriteRes entries. It's parameterized by config values that will /// eventually be supplied by different SchedMachineModels. -multiclass SiFive7SchedResources<int vlen, bit extraVALU, +multiclass SiFive7SchedResources<int vlen, bit dualVALU, SiFive7FPLatencies fpLatencies, bit hasFastGather> { - defm SiFive7 : SiFive7ProcResources<extraVALU>; + defm SiFive7 : SiFive7ProcResources<dualVALU>; // Pull out defs from SiFive7ProcResources so we can refer to them by name. defvar SiFive7PipeA = !cast<ProcResource>(NAME # SiFive7PipeA); @@ -1562,10 +1562,10 @@ multiclass SiFive7SchedResources<int vlen, bit extraVALU, defvar SiFive7IDiv = !cast<ProcResource>(NAME # SiFive7IDiv); defvar SiFive7FDiv = !cast<ProcResource>(NAME # SiFive7FDiv); // Pass SiFive7VA for VA1 and VA1OrVA2 if there is only 1 VALU. - defvar SiFive7VA1 = !if (extraVALU, + defvar SiFive7VA1 = !if (dualVALU, !cast<ProcResource>(NAME # SiFive7VA1), !cast<ProcResource>(NAME # SiFive7VA)); - defvar SiFive7VA1OrVA2 = !if (extraVALU, + defvar SiFive7VA1OrVA2 = !if (dualVALU, !cast<ProcResGroup>(NAME # SiFive7VA1OrVA2), !cast<ProcResource>(NAME # SiFive7VA)); defvar SiFive7VA = !cast<ProcResource>(NAME # SiFive7VA); @@ -1608,7 +1608,7 @@ class SiFive7SchedMachineModel<int vlen> : SchedMachineModel { HasStdExtZknh, HasStdExtZksed, HasStdExtZksh, HasStdExtZkr]; int VLEN = vlen; - bit HasExtraVALU = false; + bit HasDualVALU = false; SiFive7FPLatencies FPLatencies; bit HasFastGather = false; @@ -1635,7 +1635,7 @@ def SiFive7VLEN512Model : SiFive7SchedMachineModel<512> { } def SiFive7VLEN1024X300Model : SiFive7SchedMachineModel<1024> { - let HasExtraVALU = true; + let HasDualVALU = true; let FPLatencies = SiFive7LowFPLatencies; let HasFastGather = true; } @@ -1643,7 +1643,7 @@ def SiFive7VLEN1024X300Model : SiFive7SchedMachineModel<1024> { /// Binding models to their scheduling resources. foreach model = [SiFive7VLEN512Model, SiFive7VLEN1024X300Model] in { let SchedModel = model in - defm model.Name : SiFive7SchedResources<model.VLEN, model.HasExtraVALU, + defm model.Name : SiFive7SchedResources<model.VLEN, model.HasDualVALU, model.FPLatencies, model.HasFastGather>; } diff --git a/llvm/lib/Target/RISCV/RISCVVLOptimizer.cpp b/llvm/lib/Target/RISCV/RISCVVLOptimizer.cpp index 96ad5c6..0a8838c 100644 --- a/llvm/lib/Target/RISCV/RISCVVLOptimizer.cpp +++ b/llvm/lib/Target/RISCV/RISCVVLOptimizer.cpp @@ -156,13 +156,13 @@ FunctionPass *llvm::createRISCVVLOptimizerPass() { return new RISCVVLOptimizer(); } -LLVM_ATTRIBUTE_UNUSED +[[maybe_unused]] static raw_ostream &operator<<(raw_ostream &OS, const OperandInfo &OI) { OI.print(OS); return OS; } -LLVM_ATTRIBUTE_UNUSED +[[maybe_unused]] static raw_ostream &operator<<(raw_ostream &OS, const std::optional<OperandInfo> &OI) { if (OI) diff --git a/llvm/lib/Target/SPIRV/SPIRVInstructionSelector.cpp b/llvm/lib/Target/SPIRV/SPIRVInstructionSelector.cpp index a466ab2..a0cff4d 100644 --- a/llvm/lib/Target/SPIRV/SPIRVInstructionSelector.cpp +++ b/llvm/lib/Target/SPIRV/SPIRVInstructionSelector.cpp @@ -3765,7 +3765,6 @@ void SPIRVInstructionSelector::decorateUsesAsNonUniform( SPIRV::Decoration::NonUniformEXT, {}); } } - return; } bool SPIRVInstructionSelector::extractSubvector( diff --git a/llvm/lib/Target/Sparc/SparcFrameLowering.cpp b/llvm/lib/Target/Sparc/SparcFrameLowering.cpp index 2934c88..fa08d44 100644 --- a/llvm/lib/Target/Sparc/SparcFrameLowering.cpp +++ b/llvm/lib/Target/Sparc/SparcFrameLowering.cpp @@ -246,8 +246,7 @@ SparcFrameLowering::getFrameIndexReference(const MachineFunction &MF, int FI, } } -static bool LLVM_ATTRIBUTE_UNUSED verifyLeafProcRegUse(MachineRegisterInfo *MRI) -{ +[[maybe_unused]] static bool verifyLeafProcRegUse(MachineRegisterInfo *MRI) { for (unsigned reg = SP::I0; reg <= SP::I7; ++reg) if (MRI->isPhysRegUsed(reg)) diff --git a/llvm/lib/Target/WebAssembly/MCTargetDesc/WebAssemblyMCTypeUtilities.cpp b/llvm/lib/Target/WebAssembly/MCTargetDesc/WebAssemblyMCTypeUtilities.cpp index d9c8e22..6e99fc3 100644 --- a/llvm/lib/Target/WebAssembly/MCTargetDesc/WebAssemblyMCTypeUtilities.cpp +++ b/llvm/lib/Target/WebAssembly/MCTargetDesc/WebAssemblyMCTypeUtilities.cpp @@ -23,7 +23,7 @@ std::optional<wasm::ValType> WebAssembly::parseType(StringRef Type) { .Case("i64", wasm::ValType::I64) .Case("f32", wasm::ValType::F32) .Case("f64", wasm::ValType::F64) - .Cases("v128", "i8x16", "i16x8", "i32x4", "i64x2", "f32x4", "f64x2", + .Cases({"v128", "i8x16", "i16x8", "i32x4", "i64x2", "f32x4", "f64x2"}, wasm::ValType::V128) .Case("funcref", wasm::ValType::FUNCREF) .Case("externref", wasm::ValType::EXTERNREF) diff --git a/llvm/lib/Target/WebAssembly/WebAssemblyISelLowering.cpp b/llvm/lib/Target/WebAssembly/WebAssemblyISelLowering.cpp index 47c24fc..f973949 100644 --- a/llvm/lib/Target/WebAssembly/WebAssemblyISelLowering.cpp +++ b/llvm/lib/Target/WebAssembly/WebAssemblyISelLowering.cpp @@ -601,6 +601,29 @@ static MachineBasicBlock *LowerMemcpy(MachineInstr &MI, DebugLoc DL, MachineOperand Src = MI.getOperand(3); MachineOperand Len = MI.getOperand(4); + // If the length is a constant, we don't actually need the check. + if (MachineInstr *Def = MRI.getVRegDef(Len.getReg())) { + if (Def->getOpcode() == WebAssembly::CONST_I32 || + Def->getOpcode() == WebAssembly::CONST_I64) { + if (Def->getOperand(1).getImm() == 0) { + // A zero-length memcpy is a no-op. + MI.eraseFromParent(); + return BB; + } + // A non-zero-length memcpy doesn't need a zero check. + unsigned MemoryCopy = + Int64 ? WebAssembly::MEMORY_COPY_A64 : WebAssembly::MEMORY_COPY_A32; + BuildMI(*BB, MI, DL, TII.get(MemoryCopy)) + .add(DstMem) + .add(SrcMem) + .add(Dst) + .add(Src) + .add(Len); + MI.eraseFromParent(); + return BB; + } + } + // We're going to add an extra use to `Len` to test if it's zero; that // use shouldn't be a kill, even if the original use is. MachineOperand NoKillLen = Len; @@ -669,6 +692,28 @@ static MachineBasicBlock *LowerMemset(MachineInstr &MI, DebugLoc DL, MachineOperand Val = MI.getOperand(2); MachineOperand Len = MI.getOperand(3); + // If the length is a constant, we don't actually need the check. + if (MachineInstr *Def = MRI.getVRegDef(Len.getReg())) { + if (Def->getOpcode() == WebAssembly::CONST_I32 || + Def->getOpcode() == WebAssembly::CONST_I64) { + if (Def->getOperand(1).getImm() == 0) { + // A zero-length memset is a no-op. + MI.eraseFromParent(); + return BB; + } + // A non-zero-length memset doesn't need a zero check. + unsigned MemoryFill = + Int64 ? WebAssembly::MEMORY_FILL_A64 : WebAssembly::MEMORY_FILL_A32; + BuildMI(*BB, MI, DL, TII.get(MemoryFill)) + .add(Mem) + .add(Dst) + .add(Val) + .add(Len); + MI.eraseFromParent(); + return BB; + } + } + // We're going to add an extra use to `Len` to test if it's zero; that // use shouldn't be a kill, even if the original use is. MachineOperand NoKillLen = Len; diff --git a/llvm/lib/Target/X86/AsmParser/X86AsmParser.cpp b/llvm/lib/Target/X86/AsmParser/X86AsmParser.cpp index a8908d4..ac251fd 100644 --- a/llvm/lib/Target/X86/AsmParser/X86AsmParser.cpp +++ b/llvm/lib/Target/X86/AsmParser/X86AsmParser.cpp @@ -3514,15 +3514,16 @@ bool X86AsmParser::parseInstruction(ParseInstructionInfo &Info, StringRef Name, // xacquire <insn> ; xacquire must be accompanied by 'lock' bool IsPrefix = StringSwitch<bool>(Name) - .Cases("cs", "ds", "es", "fs", "gs", "ss", true) - .Cases("rex64", "data32", "data16", "addr32", "addr16", true) - .Cases("xacquire", "xrelease", true) - .Cases("acquire", "release", isParsingIntelSyntax()) + .Cases({"cs", "ds", "es", "fs", "gs", "ss"}, true) + .Cases({"rex64", "data32", "data16", "addr32", "addr16"}, true) + .Cases({"xacquire", "xrelease"}, true) + .Cases({"acquire", "release"}, isParsingIntelSyntax()) .Default(false); auto isLockRepeatNtPrefix = [](StringRef N) { return StringSwitch<bool>(N) - .Cases("lock", "rep", "repe", "repz", "repne", "repnz", "notrack", true) + .Cases({"lock", "rep", "repe", "repz", "repne", "repnz", "notrack"}, + true) .Default(false); }; diff --git a/llvm/lib/Target/X86/GISel/X86LegalizerInfo.cpp b/llvm/lib/Target/X86/GISel/X86LegalizerInfo.cpp index 28fa2cd..b81641f 100644 --- a/llvm/lib/Target/X86/GISel/X86LegalizerInfo.cpp +++ b/llvm/lib/Target/X86/GISel/X86LegalizerInfo.cpp @@ -414,6 +414,8 @@ X86LegalizerInfo::X86LegalizerInfo(const X86Subtarget &STI, getActionDefinitionsBuilder(G_SEXT_INREG).lower(); + getActionDefinitionsBuilder(G_IS_FPCLASS).lower(); + // fp constants getActionDefinitionsBuilder(G_FCONSTANT) .legalFor({s32, s64}) diff --git a/llvm/lib/Target/X86/X86.td b/llvm/lib/Target/X86/X86.td index 2bf016a..6db780f 100644 --- a/llvm/lib/Target/X86/X86.td +++ b/llvm/lib/Target/X86/X86.td @@ -1338,7 +1338,6 @@ def ProcessorFeatures { list<SubtargetFeature> PTLFeatures = !listremove(ARLSFeatures, [FeatureWIDEKL]); - // Clearwaterforest list<SubtargetFeature> CWFAdditionalFeatures = [FeaturePREFETCHI, FeatureAVXVNNIINT16, @@ -1880,8 +1879,10 @@ def : ProcModel<P, AlderlakePModel, } def : ProcModel<"lunarlake", LunarlakePModel, ProcessorFeatures.ARLSFeatures, ProcessorFeatures.ADLTuning>; -def : ProcModel<"pantherlake", AlderlakePModel, +foreach P = ["pantherlake", "wildcatlake"] in { +def : ProcModel<P, AlderlakePModel, ProcessorFeatures.PTLFeatures, ProcessorFeatures.ADLTuning>; +} def : ProcModel<"clearwaterforest", AlderlakePModel, ProcessorFeatures.CWFFeatures, ProcessorFeatures.ADLTuning>; def : ProcModel<"emeraldrapids", SapphireRapidsModel, diff --git a/llvm/lib/Target/X86/X86FloatingPoint.cpp b/llvm/lib/Target/X86/X86FloatingPoint.cpp index e0991aa..9f88fda 100644 --- a/llvm/lib/Target/X86/X86FloatingPoint.cpp +++ b/llvm/lib/Target/X86/X86FloatingPoint.cpp @@ -602,8 +602,7 @@ namespace { friend bool operator<(const TableEntry &TE, unsigned V) { return TE.from < V; } - friend bool LLVM_ATTRIBUTE_UNUSED operator<(unsigned V, - const TableEntry &TE) { + [[maybe_unused]] friend bool operator<(unsigned V, const TableEntry &TE) { return V < TE.from; } }; diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp index eea84a2..a0b64ff 100644 --- a/llvm/lib/Target/X86/X86ISelLowering.cpp +++ b/llvm/lib/Target/X86/X86ISelLowering.cpp @@ -3624,6 +3624,16 @@ X86TargetLowering::getJumpConditionMergingParams(Instruction::BinaryOps Opc, match(Lhs, m_SpecificICmp(ICmpInst::ICMP_EQ, m_Value(), m_Value())) && match(Rhs, m_SpecificICmp(ICmpInst::ICMP_EQ, m_Value(), m_Value()))) BaseCost += 1; + + // For OR conditions with EQ comparisons, prefer splitting into branches + // (unless CCMP is available). OR+EQ cannot be optimized via bitwise ops, + // unlike OR+NE which becomes (P|Q)!=0. Similarly, don't split signed + // comparisons (SLT, SGT) that can be optimized. + if (BaseCost >= 0 && !Subtarget.hasCCMP() && Opc == Instruction::Or && + match(Lhs, m_SpecificICmp(ICmpInst::ICMP_EQ, m_Value(), m_Value())) && + match(Rhs, m_SpecificICmp(ICmpInst::ICMP_EQ, m_Value(), m_Value()))) + return {-1, -1, -1}; + return {BaseCost, BrMergingLikelyBias.getValue(), BrMergingUnlikelyBias.getValue()}; } @@ -3787,7 +3797,7 @@ static bool isUndefOrZeroOrInRange(ArrayRef<int> Mask, int Low, int Hi) { /// Return true if every element in Mask, is an in-place blend/select mask or is /// undef. -LLVM_ATTRIBUTE_UNUSED static bool isBlendOrUndef(ArrayRef<int> Mask) { +[[maybe_unused]] static bool isBlendOrUndef(ArrayRef<int> Mask) { unsigned NumElts = Mask.size(); for (auto [I, M] : enumerate(Mask)) if (!isUndefOrEqual(M, I) && !isUndefOrEqual(M, I + NumElts)) @@ -8096,7 +8106,7 @@ static SDValue LowerBUILD_VECTORvXi1(SDValue Op, const SDLoc &dl, return DstVec; } -LLVM_ATTRIBUTE_UNUSED static bool isHorizOp(unsigned Opcode) { +[[maybe_unused]] static bool isHorizOp(unsigned Opcode) { switch (Opcode) { case X86ISD::PACKSS: case X86ISD::PACKUS: @@ -20813,7 +20823,7 @@ SDValue X86TargetLowering::FP_TO_INTHelper(SDValue Op, SelectionDAG &DAG, // for DAG type consistency we have to match the FP operand type. APFloat Thresh(APFloat::IEEEsingle(), APInt(32, 0x5f000000)); - LLVM_ATTRIBUTE_UNUSED APFloat::opStatus Status = APFloat::opOK; + [[maybe_unused]] APFloat::opStatus Status = APFloat::opOK; bool LosesInfo = false; if (TheVT == MVT::f64) // The rounding mode is irrelevant as the conversion should be exact. @@ -22856,7 +22866,7 @@ static SDValue combineVectorSizedSetCCEquality(EVT VT, SDValue X, SDValue Y, // be generated by the memcmp expansion pass with oversized integer compares // (see PR33325). bool IsOrXorXorTreeCCZero = isNullConstant(Y) && isOrXorXorTree(X); - if (isNullConstant(Y) && !IsOrXorXorTreeCCZero) + if (isNullConstant(Y) && OpSize == 128 && !IsOrXorXorTreeCCZero) return SDValue(); // Don't perform this combine if constructing the vector will be expensive. @@ -58332,11 +58342,12 @@ static SDValue combineX86CloadCstore(SDNode *N, SelectionDAG &DAG) { } else if (Op1.getOpcode() == ISD::AND && Sub.getValue(0).use_empty()) { SDValue Src = Op1; SDValue Op10 = Op1.getOperand(0); - if (Op10.getOpcode() == ISD::XOR && isAllOnesConstant(Op10.getOperand(1))) { - // res, flags2 = sub 0, (and (xor X, -1), Y) + if (Op10.getOpcode() == ISD::XOR && isAllOnesConstant(Op10.getOperand(1)) && + llvm::isOneConstant(Op1.getOperand(1))) { + // res, flags2 = sub 0, (and (xor X, -1), 1) // cload/cstore ..., cond_ne, flag2 // -> - // res, flags2 = sub 0, (and X, Y) + // res, flags2 = sub 0, (and X, 1) // cload/cstore ..., cond_e, flag2 Src = DAG.getNode(ISD::AND, DL, Op1.getValueType(), Op10.getOperand(0), Op1.getOperand(1)); diff --git a/llvm/lib/Target/X86/X86InstrCompiler.td b/llvm/lib/Target/X86/X86InstrCompiler.td index 0fd44b7..ec31675 100644 --- a/llvm/lib/Target/X86/X86InstrCompiler.td +++ b/llvm/lib/Target/X86/X86InstrCompiler.td @@ -1256,8 +1256,17 @@ def : Pat<(i64 (X86Wrapper tconstpool :$dst)), (MOV64ri32 tconstpool :$dst)>, Requires<[KernelCode]>; def : Pat<(i64 (X86Wrapper tjumptable :$dst)), (MOV64ri32 tjumptable :$dst)>, Requires<[KernelCode]>; -def : Pat<(i64 (X86Wrapper tglobaladdr :$dst)), - (MOV64ri32 tglobaladdr :$dst)>, Requires<[KernelCode]>; + +// If the globaladdr is an absolute_symbol, don't bother using the sign extending +// instruction since there's no benefit to using it with absolute symbols. +def globalAddrNoAbsSym : PatLeaf<(tglobaladdr:$dst), [{ + auto *GA = cast<GlobalAddressSDNode>(N); + return !GA->getGlobal()->getAbsoluteSymbolRange(); +}]>; +def : Pat<(i64 (X86Wrapper globalAddrNoAbsSym:$dst)), + (MOV64ri32 tglobaladdr:$dst)>, + Requires<[KernelCode]>; + def : Pat<(i64 (X86Wrapper texternalsym:$dst)), (MOV64ri32 texternalsym:$dst)>, Requires<[KernelCode]>; def : Pat<(i64 (X86Wrapper mcsym:$dst)), diff --git a/llvm/lib/TargetParser/ARMTargetParserCommon.cpp b/llvm/lib/TargetParser/ARMTargetParserCommon.cpp index 89d5e0d..f6cea85 100644 --- a/llvm/lib/TargetParser/ARMTargetParserCommon.cpp +++ b/llvm/lib/TargetParser/ARMTargetParserCommon.cpp @@ -22,13 +22,13 @@ StringRef ARM::getArchSynonym(StringRef Arch) { .Case("v5e", "v5te") .Case("v6j", "v6") .Case("v6hl", "v6k") - .Cases("v6m", "v6sm", "v6s-m", "v6-m") - .Cases("v6z", "v6zk", "v6kz") - .Cases("v7", "v7a", "v7hl", "v7l", "v7-a") + .Cases({"v6m", "v6sm", "v6s-m"}, "v6-m") + .Cases({"v6z", "v6zk"}, "v6kz") + .Cases({"v7", "v7a", "v7hl", "v7l"}, "v7-a") .Case("v7r", "v7-r") .Case("v7m", "v7-m") .Case("v7em", "v7e-m") - .Cases("v8", "v8a", "v8l", "aarch64", "arm64", "v8-a") + .Cases({"v8", "v8a", "v8l", "aarch64", "arm64"}, "v8-a") .Case("v8.1a", "v8.1-a") .Case("v8.2a", "v8.2-a") .Case("v8.3a", "v8.3-a") @@ -39,7 +39,7 @@ StringRef ARM::getArchSynonym(StringRef Arch) { .Case("v8.8a", "v8.8-a") .Case("v8.9a", "v8.9-a") .Case("v8r", "v8-r") - .Cases("v9", "v9a", "v9-a") + .Cases({"v9", "v9a"}, "v9-a") .Case("v9.1a", "v9.1-a") .Case("v9.2a", "v9.2-a") .Case("v9.3a", "v9.3-a") diff --git a/llvm/lib/TargetParser/Host.cpp b/llvm/lib/TargetParser/Host.cpp index a5bdc9d..3479106 100644 --- a/llvm/lib/TargetParser/Host.cpp +++ b/llvm/lib/TargetParser/Host.cpp @@ -70,8 +70,8 @@ using namespace llvm; -static std::unique_ptr<llvm::MemoryBuffer> - LLVM_ATTRIBUTE_UNUSED getProcCpuinfoContent() { +[[maybe_unused]] static std::unique_ptr<llvm::MemoryBuffer> +getProcCpuinfoContent() { const char *CPUInfoFile = "/proc/cpuinfo"; if (const char *CpuinfoIntercept = std::getenv("LLVM_CPUINFO")) CPUInfoFile = CpuinfoIntercept; @@ -964,6 +964,13 @@ static StringRef getIntelProcessorTypeAndSubtype(unsigned Family, *Subtype = X86::INTEL_COREI7_PANTHERLAKE; break; + // Wildcatlake: + case 0xd5: + CPU = "wildcatlake"; + *Type = X86::INTEL_COREI7; + *Subtype = X86::INTEL_COREI7_PANTHERLAKE; + break; + // Graniterapids: case 0xad: CPU = "graniterapids"; diff --git a/llvm/lib/TargetParser/TargetDataLayout.cpp b/llvm/lib/TargetParser/TargetDataLayout.cpp index 950bb2b..d765d9c 100644 --- a/llvm/lib/TargetParser/TargetDataLayout.cpp +++ b/llvm/lib/TargetParser/TargetDataLayout.cpp @@ -548,8 +548,11 @@ std::string Triple::computeDataLayout(StringRef ABIName) const { case Triple::csky: return computeCSKYDataLayout(*this); case Triple::dxil: + // TODO: We need to align vectors on the element size generally, but for now + // we hard code this for 3-element 32- and 64-bit vectors as a workaround. + // See https://github.com/llvm/llvm-project/issues/123968 return "e-m:e-p:32:32-i1:32-i8:8-i16:16-i32:32-i64:64-f16:16-" - "f32:32-f64:64-n8:16:32:64"; + "f32:32-f64:64-n8:16:32:64-v48:16:16-v96:32:32-v192:64:64"; case Triple::hexagon: return "e-m:e-p:32:32:32-a:0-n16:32-" "i64:64:64-i32:32:32-i16:16:16-i1:8:8-f32:32:32-f64:64:64-" diff --git a/llvm/lib/TargetParser/Triple.cpp b/llvm/lib/TargetParser/Triple.cpp index f021094..1068ce4 100644 --- a/llvm/lib/TargetParser/Triple.cpp +++ b/llvm/lib/TargetParser/Triple.cpp @@ -579,87 +579,89 @@ static Triple::ArchType parseARMArch(StringRef ArchName) { } static Triple::ArchType parseArch(StringRef ArchName) { - auto AT = StringSwitch<Triple::ArchType>(ArchName) - .Cases("i386", "i486", "i586", "i686", Triple::x86) - // FIXME: Do we need to support these? - .Cases("i786", "i886", "i986", Triple::x86) - .Cases("amd64", "x86_64", "x86_64h", Triple::x86_64) - .Cases("powerpc", "powerpcspe", "ppc", "ppc32", Triple::ppc) - .Cases("powerpcle", "ppcle", "ppc32le", Triple::ppcle) - .Cases("powerpc64", "ppu", "ppc64", Triple::ppc64) - .Cases("powerpc64le", "ppc64le", Triple::ppc64le) - .Case("xscale", Triple::arm) - .Case("xscaleeb", Triple::armeb) - .Case("aarch64", Triple::aarch64) - .Case("aarch64_be", Triple::aarch64_be) - .Case("aarch64_32", Triple::aarch64_32) - .Case("arc", Triple::arc) - .Case("arm64", Triple::aarch64) - .Case("arm64_32", Triple::aarch64_32) - .Case("arm64e", Triple::aarch64) - .Case("arm64ec", Triple::aarch64) - .Case("arm", Triple::arm) - .Case("armeb", Triple::armeb) - .Case("thumb", Triple::thumb) - .Case("thumbeb", Triple::thumbeb) - .Case("avr", Triple::avr) - .Case("m68k", Triple::m68k) - .Case("msp430", Triple::msp430) - .Cases("mips", "mipseb", "mipsallegrex", "mipsisa32r6", - "mipsr6", Triple::mips) - .Cases("mipsel", "mipsallegrexel", "mipsisa32r6el", "mipsr6el", - Triple::mipsel) - .Cases("mips64", "mips64eb", "mipsn32", "mipsisa64r6", - "mips64r6", "mipsn32r6", Triple::mips64) - .Cases("mips64el", "mipsn32el", "mipsisa64r6el", "mips64r6el", - "mipsn32r6el", Triple::mips64el) - .Case("r600", Triple::r600) - .Case("amdgcn", Triple::amdgcn) - .Case("riscv32", Triple::riscv32) - .Case("riscv64", Triple::riscv64) - .Case("riscv32be", Triple::riscv32be) - .Case("riscv64be", Triple::riscv64be) - .Case("hexagon", Triple::hexagon) - .Cases("s390x", "systemz", Triple::systemz) - .Case("sparc", Triple::sparc) - .Case("sparcel", Triple::sparcel) - .Cases("sparcv9", "sparc64", Triple::sparcv9) - .Case("tce", Triple::tce) - .Case("tcele", Triple::tcele) - .Case("xcore", Triple::xcore) - .Case("nvptx", Triple::nvptx) - .Case("nvptx64", Triple::nvptx64) - .Case("amdil", Triple::amdil) - .Case("amdil64", Triple::amdil64) - .Case("hsail", Triple::hsail) - .Case("hsail64", Triple::hsail64) - .Case("spir", Triple::spir) - .Case("spir64", Triple::spir64) - .Cases("spirv", "spirv1.5", "spirv1.6", Triple::spirv) - .Cases("spirv32", "spirv32v1.0", "spirv32v1.1", "spirv32v1.2", - "spirv32v1.3", "spirv32v1.4", "spirv32v1.5", - "spirv32v1.6", Triple::spirv32) - .Cases("spirv64", "spirv64v1.0", "spirv64v1.1", "spirv64v1.2", - "spirv64v1.3", "spirv64v1.4", "spirv64v1.5", - "spirv64v1.6", Triple::spirv64) - .StartsWith("kalimba", Triple::kalimba) - .Case("lanai", Triple::lanai) - .Case("renderscript32", Triple::renderscript32) - .Case("renderscript64", Triple::renderscript64) - .Case("shave", Triple::shave) - .Case("ve", Triple::ve) - .Case("wasm32", Triple::wasm32) - .Case("wasm64", Triple::wasm64) - .Case("csky", Triple::csky) - .Case("loongarch32", Triple::loongarch32) - .Case("loongarch64", Triple::loongarch64) - .Cases("dxil", "dxilv1.0", "dxilv1.1", "dxilv1.2", "dxilv1.3", - "dxilv1.4", "dxilv1.5", "dxilv1.6", "dxilv1.7", - "dxilv1.8", Triple::dxil) - // Note: Cases has max limit of 10. - .Case("dxilv1.9", Triple::dxil) - .Case("xtensa", Triple::xtensa) - .Default(Triple::UnknownArch); + auto AT = + StringSwitch<Triple::ArchType>(ArchName) + .Cases({"i386", "i486", "i586", "i686"}, Triple::x86) + // FIXME: Do we need to support these? + .Cases({"i786", "i886", "i986"}, Triple::x86) + .Cases({"amd64", "x86_64", "x86_64h"}, Triple::x86_64) + .Cases({"powerpc", "powerpcspe", "ppc", "ppc32"}, Triple::ppc) + .Cases({"powerpcle", "ppcle", "ppc32le"}, Triple::ppcle) + .Cases({"powerpc64", "ppu", "ppc64"}, Triple::ppc64) + .Cases({"powerpc64le", "ppc64le"}, Triple::ppc64le) + .Case("xscale", Triple::arm) + .Case("xscaleeb", Triple::armeb) + .Case("aarch64", Triple::aarch64) + .Case("aarch64_be", Triple::aarch64_be) + .Case("aarch64_32", Triple::aarch64_32) + .Case("arc", Triple::arc) + .Case("arm64", Triple::aarch64) + .Case("arm64_32", Triple::aarch64_32) + .Case("arm64e", Triple::aarch64) + .Case("arm64ec", Triple::aarch64) + .Case("arm", Triple::arm) + .Case("armeb", Triple::armeb) + .Case("thumb", Triple::thumb) + .Case("thumbeb", Triple::thumbeb) + .Case("avr", Triple::avr) + .Case("m68k", Triple::m68k) + .Case("msp430", Triple::msp430) + .Cases({"mips", "mipseb", "mipsallegrex", "mipsisa32r6", "mipsr6"}, + Triple::mips) + .Cases({"mipsel", "mipsallegrexel", "mipsisa32r6el", "mipsr6el"}, + Triple::mipsel) + .Cases({"mips64", "mips64eb", "mipsn32", "mipsisa64r6", "mips64r6", + "mipsn32r6"}, + Triple::mips64) + .Cases({"mips64el", "mipsn32el", "mipsisa64r6el", "mips64r6el", + "mipsn32r6el"}, + Triple::mips64el) + .Case("r600", Triple::r600) + .Case("amdgcn", Triple::amdgcn) + .Case("riscv32", Triple::riscv32) + .Case("riscv64", Triple::riscv64) + .Case("riscv32be", Triple::riscv32be) + .Case("riscv64be", Triple::riscv64be) + .Case("hexagon", Triple::hexagon) + .Cases({"s390x", "systemz"}, Triple::systemz) + .Case("sparc", Triple::sparc) + .Case("sparcel", Triple::sparcel) + .Cases({"sparcv9", "sparc64"}, Triple::sparcv9) + .Case("tce", Triple::tce) + .Case("tcele", Triple::tcele) + .Case("xcore", Triple::xcore) + .Case("nvptx", Triple::nvptx) + .Case("nvptx64", Triple::nvptx64) + .Case("amdil", Triple::amdil) + .Case("amdil64", Triple::amdil64) + .Case("hsail", Triple::hsail) + .Case("hsail64", Triple::hsail64) + .Case("spir", Triple::spir) + .Case("spir64", Triple::spir64) + .Cases({"spirv", "spirv1.5", "spirv1.6"}, Triple::spirv) + .Cases({"spirv32", "spirv32v1.0", "spirv32v1.1", "spirv32v1.2", + "spirv32v1.3", "spirv32v1.4", "spirv32v1.5", "spirv32v1.6"}, + Triple::spirv32) + .Cases({"spirv64", "spirv64v1.0", "spirv64v1.1", "spirv64v1.2", + "spirv64v1.3", "spirv64v1.4", "spirv64v1.5", "spirv64v1.6"}, + Triple::spirv64) + .StartsWith("kalimba", Triple::kalimba) + .Case("lanai", Triple::lanai) + .Case("renderscript32", Triple::renderscript32) + .Case("renderscript64", Triple::renderscript64) + .Case("shave", Triple::shave) + .Case("ve", Triple::ve) + .Case("wasm32", Triple::wasm32) + .Case("wasm64", Triple::wasm64) + .Case("csky", Triple::csky) + .Case("loongarch32", Triple::loongarch32) + .Case("loongarch64", Triple::loongarch64) + .Cases({"dxil", "dxilv1.0", "dxilv1.1", "dxilv1.2", "dxilv1.3", + "dxilv1.4", "dxilv1.5", "dxilv1.6", "dxilv1.7", "dxilv1.8", + "dxilv1.9"}, + Triple::dxil) + .Case("xtensa", Triple::xtensa) + .Default(Triple::UnknownArch); // Some architectures require special parsing logic just to compute the // ArchType result. @@ -1071,7 +1073,7 @@ Triple::Triple(std::string &&Str) : Data(std::move(Str)) { .StartsWith("mips64", Triple::GNUABI64) .StartsWith("mipsisa64", Triple::GNUABI64) .StartsWith("mipsisa32", Triple::GNU) - .Cases("mips", "mipsel", "mipsr6", "mipsr6el", Triple::GNU) + .Cases({"mips", "mipsel", "mipsr6", "mipsr6el"}, Triple::GNU) .Default(UnknownEnvironment); } } diff --git a/llvm/lib/TargetParser/X86TargetParser.cpp b/llvm/lib/TargetParser/X86TargetParser.cpp index 1932a3a..e382cfe 100644 --- a/llvm/lib/TargetParser/X86TargetParser.cpp +++ b/llvm/lib/TargetParser/X86TargetParser.cpp @@ -378,6 +378,7 @@ constexpr ProcInfo Processors[] = { { {"gracemont"}, CK_Gracemont, FEATURE_AVX2, FeaturesAlderlake, 'p', false }, // Pantherlake microarchitecture based processors. { {"pantherlake"}, CK_Lunarlake, FEATURE_AVX2, FeaturesPantherlake, 'p', false }, + { {"wildcatlake"}, CK_Lunarlake, FEATURE_AVX2, FeaturesPantherlake, 'p', false }, // Sierraforest microarchitecture based processors. { {"sierraforest"}, CK_Sierraforest, FEATURE_AVX2, FeaturesSierraforest, 'p', false }, // Grandridge microarchitecture based processors. diff --git a/llvm/lib/Transforms/InstCombine/InstCombineSelect.cpp b/llvm/lib/Transforms/InstCombine/InstCombineSelect.cpp index 6b67b48..09cb225 100644 --- a/llvm/lib/Transforms/InstCombine/InstCombineSelect.cpp +++ b/llvm/lib/Transforms/InstCombine/InstCombineSelect.cpp @@ -2979,10 +2979,14 @@ Instruction *InstCombinerImpl::foldAndOrOfSelectUsingImpliedCond(Value *Op, "Op must be either i1 or vector of i1."); if (SI.getCondition()->getType() != Op->getType()) return nullptr; - if (Value *V = simplifyNestedSelectsUsingImpliedCond(SI, Op, IsAnd, DL)) - return SelectInst::Create(Op, - IsAnd ? V : ConstantInt::getTrue(Op->getType()), - IsAnd ? ConstantInt::getFalse(Op->getType()) : V); + if (Value *V = simplifyNestedSelectsUsingImpliedCond(SI, Op, IsAnd, DL)) { + Instruction *MDFrom = nullptr; + if (!ProfcheckDisableMetadataFixes) + MDFrom = &SI; + return SelectInst::Create( + Op, IsAnd ? V : ConstantInt::getTrue(Op->getType()), + IsAnd ? ConstantInt::getFalse(Op->getType()) : V, "", nullptr, MDFrom); + } return nullptr; } diff --git a/llvm/lib/Transforms/Instrumentation/ControlHeightReduction.cpp b/llvm/lib/Transforms/Instrumentation/ControlHeightReduction.cpp index 7c78eb3..444b390 100644 --- a/llvm/lib/Transforms/Instrumentation/ControlHeightReduction.cpp +++ b/llvm/lib/Transforms/Instrumentation/ControlHeightReduction.cpp @@ -396,9 +396,8 @@ class CHR { } // end anonymous namespace -static inline -raw_ostream LLVM_ATTRIBUTE_UNUSED &operator<<(raw_ostream &OS, - const CHRStats &Stats) { +[[maybe_unused]] static inline raw_ostream &operator<<(raw_ostream &OS, + const CHRStats &Stats) { Stats.print(OS); return OS; } @@ -425,8 +424,8 @@ static bool shouldApply(Function &F, ProfileSummaryInfo &PSI) { return PSI.isFunctionEntryHot(&F); } -static void LLVM_ATTRIBUTE_UNUSED dumpIR(Function &F, const char *Label, - CHRStats *Stats) { +[[maybe_unused]] static void dumpIR(Function &F, const char *Label, + CHRStats *Stats) { StringRef FuncName = F.getName(); StringRef ModuleName = F.getParent()->getName(); (void)(FuncName); // Unused in release build. @@ -1622,7 +1621,7 @@ static void insertTrivialPHIs(CHRScope *Scope, } // Assert that all the CHR regions of the scope have a biased branch or select. -static void LLVM_ATTRIBUTE_UNUSED +[[maybe_unused]] static void assertCHRRegionsHaveBiasedBranchOrSelect(CHRScope *Scope) { #ifndef NDEBUG auto HasBiasedBranchOrSelect = [](RegInfo &RI, CHRScope *Scope) { @@ -1644,8 +1643,9 @@ assertCHRRegionsHaveBiasedBranchOrSelect(CHRScope *Scope) { // Assert that all the condition values of the biased branches and selects have // been hoisted to the pre-entry block or outside of the scope. -static void LLVM_ATTRIBUTE_UNUSED assertBranchOrSelectConditionHoisted( - CHRScope *Scope, BasicBlock *PreEntryBlock) { +[[maybe_unused]] static void +assertBranchOrSelectConditionHoisted(CHRScope *Scope, + BasicBlock *PreEntryBlock) { CHR_DEBUG(dbgs() << "Biased regions condition values \n"); for (RegInfo &RI : Scope->CHRRegions) { Region *R = RI.R; @@ -2007,8 +2007,8 @@ void CHR::transformScopes(SmallVectorImpl<CHRScope *> &CHRScopes) { } } -static void LLVM_ATTRIBUTE_UNUSED -dumpScopes(SmallVectorImpl<CHRScope *> &Scopes, const char *Label) { +[[maybe_unused]] static void dumpScopes(SmallVectorImpl<CHRScope *> &Scopes, + const char *Label) { dbgs() << Label << " " << Scopes.size() << "\n"; for (CHRScope *Scope : Scopes) { dbgs() << *Scope << "\n"; diff --git a/llvm/lib/Transforms/ObjCARC/ObjCARCOpts.cpp b/llvm/lib/Transforms/ObjCARC/ObjCARCOpts.cpp index 09db464..386e48f 100644 --- a/llvm/lib/Transforms/ObjCARC/ObjCARCOpts.cpp +++ b/llvm/lib/Transforms/ObjCARC/ObjCARCOpts.cpp @@ -326,8 +326,7 @@ const unsigned BBState::OverflowOccurredValue = 0xffffffff; namespace llvm { -raw_ostream &operator<<(raw_ostream &OS, - BBState &BBState) LLVM_ATTRIBUTE_UNUSED; +[[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, BBState &BBState); } // end namespace llvm diff --git a/llvm/lib/Transforms/ObjCARC/PtrState.h b/llvm/lib/Transforms/ObjCARC/PtrState.h index 232db2b..5cc4212 100644 --- a/llvm/lib/Transforms/ObjCARC/PtrState.h +++ b/llvm/lib/Transforms/ObjCARC/PtrState.h @@ -47,8 +47,7 @@ enum Sequence { S_MovableRelease ///< objc_release(x), !clang.imprecise_release. }; -raw_ostream &operator<<(raw_ostream &OS, - const Sequence S) LLVM_ATTRIBUTE_UNUSED; +[[maybe_unused]] raw_ostream &operator<<(raw_ostream &OS, const Sequence S); /// Unidirectional information about either a /// retain-decrement-use-release sequence or release-use-decrement-retain diff --git a/llvm/lib/Transforms/Scalar/DFAJumpThreading.cpp b/llvm/lib/Transforms/Scalar/DFAJumpThreading.cpp index ff5f390..66e45ec 100644 --- a/llvm/lib/Transforms/Scalar/DFAJumpThreading.cpp +++ b/llvm/lib/Transforms/Scalar/DFAJumpThreading.cpp @@ -266,8 +266,7 @@ void DFAJumpThreading::unfold(DomTreeUpdater *DTU, LoopInfo *LI, if (!ProfcheckDisableMetadataFixes) BI->setMetadata(LLVMContext::MD_prof, SI->getMetadata(LLVMContext::MD_prof)); - DTU->applyUpdates({{DominatorTree::Insert, StartBlock, EndBlock}, - {DominatorTree::Insert, StartBlock, NewBlock}}); + DTU->applyUpdates({{DominatorTree::Insert, StartBlock, NewBlock}}); } else { BasicBlock *EndBlock = SIUse->getParent(); BasicBlock *NewBlockT = BasicBlock::Create( @@ -1479,10 +1478,13 @@ bool DFAJumpThreading::run(Function &F) { DTU->flush(); #ifdef EXPENSIVE_CHECKS - assert(DTU->getDomTree().verify(DominatorTree::VerificationLevel::Full)); verifyFunction(F, &dbgs()); #endif + if (MadeChanges && VerifyDomInfo) + assert(DTU->getDomTree().verify(DominatorTree::VerificationLevel::Full) && + "Failed to maintain validity of domtree!"); + return MadeChanges; } diff --git a/llvm/lib/Transforms/Scalar/DeadStoreElimination.cpp b/llvm/lib/Transforms/Scalar/DeadStoreElimination.cpp index 7ad710d..6141b6d 100644 --- a/llvm/lib/Transforms/Scalar/DeadStoreElimination.cpp +++ b/llvm/lib/Transforms/Scalar/DeadStoreElimination.cpp @@ -77,6 +77,7 @@ #include "llvm/Support/DebugCounter.h" #include "llvm/Support/ErrorHandling.h" #include "llvm/Support/raw_ostream.h" +#include "llvm/Transforms/Scalar.h" #include "llvm/Transforms/Utils/AssumeBundleBuilder.h" #include "llvm/Transforms/Utils/BuildLibCalls.h" #include "llvm/Transforms/Utils/Local.h" @@ -805,9 +806,8 @@ tryToMergePartialOverlappingStores(StoreInst *KillingI, StoreInst *DeadI, return nullptr; } -namespace { // Returns true if \p I is an intrinsic that does not read or write memory. -bool isNoopIntrinsic(Instruction *I) { +static bool isNoopIntrinsic(Instruction *I) { if (const IntrinsicInst *II = dyn_cast<IntrinsicInst>(I)) { switch (II->getIntrinsicID()) { case Intrinsic::lifetime_start: @@ -828,7 +828,7 @@ bool isNoopIntrinsic(Instruction *I) { } // Check if we can ignore \p D for DSE. -bool canSkipDef(MemoryDef *D, bool DefVisibleToCaller) { +static bool canSkipDef(MemoryDef *D, bool DefVisibleToCaller) { Instruction *DI = D->getMemoryInst(); // Calls that only access inaccessible memory cannot read or write any memory // locations we consider for elimination. @@ -856,6 +856,8 @@ bool canSkipDef(MemoryDef *D, bool DefVisibleToCaller) { return false; } +namespace { + // A memory location wrapper that represents a MemoryLocation, `MemLoc`, // defined by `MemDef`. struct MemoryLocationWrapper { @@ -889,23 +891,25 @@ struct MemoryDefWrapper { SmallVector<MemoryLocationWrapper, 1> DefinedLocations; }; -bool hasInitializesAttr(Instruction *I) { - CallBase *CB = dyn_cast<CallBase>(I); - return CB && CB->getArgOperandWithAttribute(Attribute::Initializes); -} - struct ArgumentInitInfo { unsigned Idx; bool IsDeadOrInvisibleOnUnwind; ConstantRangeList Inits; }; +} // namespace + +static bool hasInitializesAttr(Instruction *I) { + CallBase *CB = dyn_cast<CallBase>(I); + return CB && CB->getArgOperandWithAttribute(Attribute::Initializes); +} // Return the intersected range list of the initializes attributes of "Args". // "Args" are call arguments that alias to each other. // If any argument in "Args" doesn't have dead_on_unwind attr and // "CallHasNoUnwindAttr" is false, return empty. -ConstantRangeList getIntersectedInitRangeList(ArrayRef<ArgumentInitInfo> Args, - bool CallHasNoUnwindAttr) { +static ConstantRangeList +getIntersectedInitRangeList(ArrayRef<ArgumentInitInfo> Args, + bool CallHasNoUnwindAttr) { if (Args.empty()) return {}; @@ -925,6 +929,8 @@ ConstantRangeList getIntersectedInitRangeList(ArrayRef<ArgumentInitInfo> Args, return IntersectedIntervals; } +namespace { + struct DSEState { Function &F; AliasAnalysis &AA; @@ -2328,10 +2334,11 @@ struct DSEState { // change state: whether make any change. bool eliminateDeadDefs(const MemoryDefWrapper &KillingDefWrapper); }; +} // namespace // Return true if "Arg" is function local and isn't captured before "CB". -bool isFuncLocalAndNotCaptured(Value *Arg, const CallBase *CB, - EarliestEscapeAnalysis &EA) { +static bool isFuncLocalAndNotCaptured(Value *Arg, const CallBase *CB, + EarliestEscapeAnalysis &EA) { const Value *UnderlyingObj = getUnderlyingObject(Arg); return isIdentifiedFunctionLocal(UnderlyingObj) && capturesNothing( @@ -2627,7 +2634,6 @@ static bool eliminateDeadStores(Function &F, AliasAnalysis &AA, MemorySSA &MSSA, return MadeChange; } -} // end anonymous namespace //===----------------------------------------------------------------------===// // DSE Pass @@ -2728,8 +2734,6 @@ INITIALIZE_PASS_DEPENDENCY(AssumptionCacheTracker) INITIALIZE_PASS_END(DSELegacyPass, "dse", "Dead Store Elimination", false, false) -namespace llvm { -LLVM_ABI FunctionPass *createDeadStoreEliminationPass() { +LLVM_ABI FunctionPass *llvm::createDeadStoreEliminationPass() { return new DSELegacyPass(); } -} // namespace llvm diff --git a/llvm/lib/Transforms/Scalar/FlattenCFGPass.cpp b/llvm/lib/Transforms/Scalar/FlattenCFGPass.cpp index 213d0f3..1335665 100644 --- a/llvm/lib/Transforms/Scalar/FlattenCFGPass.cpp +++ b/llvm/lib/Transforms/Scalar/FlattenCFGPass.cpp @@ -39,10 +39,11 @@ public: private: AliasAnalysis *AA; }; +} // namespace /// iterativelyFlattenCFG - Call FlattenCFG on all the blocks in the function, /// iterating until no more changes are made. -bool iterativelyFlattenCFG(Function &F, AliasAnalysis *AA) { +static bool iterativelyFlattenCFG(Function &F, AliasAnalysis *AA) { bool Changed = false; bool LocalChange = true; @@ -67,7 +68,6 @@ bool iterativelyFlattenCFG(Function &F, AliasAnalysis *AA) { } return Changed; } -} // namespace char FlattenCFGLegacyPass::ID = 0; diff --git a/llvm/lib/Transforms/Scalar/GVNSink.cpp b/llvm/lib/Transforms/Scalar/GVNSink.cpp index 1c88532..b9534def 100644 --- a/llvm/lib/Transforms/Scalar/GVNSink.cpp +++ b/llvm/lib/Transforms/Scalar/GVNSink.cpp @@ -73,24 +73,17 @@ #include <utility> using namespace llvm; +using namespace llvm::GVNExpression; #define DEBUG_TYPE "gvn-sink" STATISTIC(NumRemoved, "Number of instructions removed"); -namespace llvm { -namespace GVNExpression { - LLVM_DUMP_METHOD void Expression::dump() const { print(dbgs()); dbgs() << "\n"; } -} // end namespace GVNExpression -} // end namespace llvm - -namespace { - static bool isMemoryInst(const Instruction *I) { return isa<LoadInst>(I) || isa<StoreInst>(I) || (isa<InvokeInst>(I) && !cast<InvokeInst>(I)->doesNotAccessMemory()) || @@ -99,6 +92,8 @@ static bool isMemoryInst(const Instruction *I) { //===----------------------------------------------------------------------===// +namespace { + /// Candidate solution for sinking. There may be different ways to /// sink instructions, differing in the number of instructions sunk, /// the number of predecessors sunk from and the number of PHIs @@ -125,14 +120,6 @@ struct SinkingInstructionCandidate { } }; -#ifndef NDEBUG -raw_ostream &operator<<(raw_ostream &OS, const SinkingInstructionCandidate &C) { - OS << "<Candidate Cost=" << C.Cost << " #Blocks=" << C.NumBlocks - << " #Insts=" << C.NumInstructions << " #PHIs=" << C.NumPHIs << ">"; - return OS; -} -#endif - //===----------------------------------------------------------------------===// /// Describes a PHI node that may or may not exist. These track the PHIs @@ -256,8 +243,18 @@ public: return Values == Other.Values && Blocks == Other.Blocks; } }; +} // namespace -template <typename ModelledPHI> struct DenseMapInfo { +#ifndef NDEBUG +static raw_ostream &operator<<(raw_ostream &OS, + const SinkingInstructionCandidate &C) { + OS << "<Candidate Cost=" << C.Cost << " #Blocks=" << C.NumBlocks + << " #Insts=" << C.NumInstructions << " #PHIs=" << C.NumPHIs << ">"; + return OS; +} +#endif + +template <> struct llvm::DenseMapInfo<ModelledPHI> { static inline ModelledPHI &getEmptyKey() { static ModelledPHI Dummy = ModelledPHI::createDummy(0); return Dummy; @@ -275,7 +272,9 @@ template <typename ModelledPHI> struct DenseMapInfo { } }; -using ModelledPHISet = DenseSet<ModelledPHI, DenseMapInfo<ModelledPHI>>; +using ModelledPHISet = DenseSet<ModelledPHI>; + +namespace { //===----------------------------------------------------------------------===// // ValueTable @@ -290,7 +289,7 @@ using ModelledPHISet = DenseSet<ModelledPHI, DenseMapInfo<ModelledPHI>>; /// /// This class also contains fields for discriminators used when determining /// equivalence of instructions with sideeffects. -class InstructionUseExpr : public GVNExpression::BasicExpression { +class InstructionUseExpr : public BasicExpression { unsigned MemoryUseOrder = -1; bool Volatile = false; ArrayRef<int> ShuffleMask; @@ -298,7 +297,7 @@ class InstructionUseExpr : public GVNExpression::BasicExpression { public: InstructionUseExpr(Instruction *I, ArrayRecycler<Value *> &R, BumpPtrAllocator &A) - : GVNExpression::BasicExpression(I->getNumUses()) { + : BasicExpression(I->getNumUses()) { allocateOperands(R, A); setOpcode(I->getOpcode()); setType(I->getType()); @@ -315,8 +314,8 @@ public: void setVolatile(bool V) { Volatile = V; } hash_code getHashValue() const override { - return hash_combine(GVNExpression::BasicExpression::getHashValue(), - MemoryUseOrder, Volatile, ShuffleMask); + return hash_combine(BasicExpression::getHashValue(), MemoryUseOrder, + Volatile, ShuffleMask); } template <typename Function> hash_code getHashValue(Function MapFn) { @@ -332,7 +331,7 @@ using BasicBlocksSet = SmallPtrSet<const BasicBlock *, 32>; class ValueTable { DenseMap<Value *, uint32_t> ValueNumbering; - DenseMap<GVNExpression::Expression *, uint32_t> ExpressionNumbering; + DenseMap<Expression *, uint32_t> ExpressionNumbering; DenseMap<size_t, uint32_t> HashNumbering; BumpPtrAllocator Allocator; ArrayRecycler<Value *> Recycler; @@ -594,6 +593,7 @@ private: } } }; +} // namespace std::optional<SinkingInstructionCandidate> GVNSink::analyzeInstructionForSinking(LockstepReverseIterator<false> &LRI, @@ -851,8 +851,6 @@ void GVNSink::sinkLastInstruction(ArrayRef<BasicBlock *> Blocks, NumRemoved += Insts.size() - 1; } -} // end anonymous namespace - PreservedAnalyses GVNSinkPass::run(Function &F, FunctionAnalysisManager &AM) { GVNSink G; if (!G.run(F)) diff --git a/llvm/lib/Transforms/Scalar/SROA.cpp b/llvm/lib/Transforms/Scalar/SROA.cpp index b9d332b..a692009 100644 --- a/llvm/lib/Transforms/Scalar/SROA.cpp +++ b/llvm/lib/Transforms/Scalar/SROA.cpp @@ -118,9 +118,13 @@ STATISTIC( STATISTIC(NumDeleted, "Number of instructions deleted"); STATISTIC(NumVectorized, "Number of vectorized aggregates"); +namespace llvm { /// Disable running mem2reg during SROA in order to test or debug SROA. static cl::opt<bool> SROASkipMem2Reg("sroa-skip-mem2reg", cl::init(false), cl::Hidden); +extern cl::opt<bool> ProfcheckDisableMetadataFixes; +} // namespace llvm + namespace { class AllocaSliceRewriter; @@ -547,12 +551,10 @@ public: } /// Support comparison with a single offset to allow binary searches. - friend LLVM_ATTRIBUTE_UNUSED bool operator<(const Slice &LHS, - uint64_t RHSOffset) { + [[maybe_unused]] friend bool operator<(const Slice &LHS, uint64_t RHSOffset) { return LHS.beginOffset() < RHSOffset; } - friend LLVM_ATTRIBUTE_UNUSED bool operator<(uint64_t LHSOffset, - const Slice &RHS) { + [[maybe_unused]] friend bool operator<(uint64_t LHSOffset, const Slice &RHS) { return LHSOffset < RHS.beginOffset(); } @@ -1777,7 +1779,8 @@ static void speculateSelectInstLoads(SelectInst &SI, LoadInst &LI, } Value *V = IRB.CreateSelect(SI.getCondition(), TL, FL, - LI.getName() + ".sroa.speculated"); + LI.getName() + ".sroa.speculated", + ProfcheckDisableMetadataFixes ? nullptr : &SI); LLVM_DEBUG(dbgs() << " speculated to: " << *V << "\n"); LI.replaceAllUsesWith(V); @@ -2662,7 +2665,9 @@ static Value *insertVector(IRBuilderTy &IRB, Value *Old, Value *V, for (unsigned i = 0; i != cast<FixedVectorType>(VecTy)->getNumElements(); ++i) Mask2.push_back(IRB.getInt1(i >= BeginIndex && i < EndIndex)); - V = IRB.CreateSelect(ConstantVector::get(Mask2), V, Old, Name + "blend"); + // No profiling support for vector selects. + V = IRB.CreateSelectWithUnknownProfile(ConstantVector::get(Mask2), V, Old, + DEBUG_TYPE, Name + "blend"); LLVM_DEBUG(dbgs() << " blend: " << *V << "\n"); return V; @@ -4360,10 +4365,13 @@ private: }; Value *Cond, *True, *False; + Instruction *MDFrom = nullptr; if (auto *SI = dyn_cast<SelectInst>(Sel)) { Cond = SI->getCondition(); True = SI->getTrueValue(); False = SI->getFalseValue(); + if (!ProfcheckDisableMetadataFixes) + MDFrom = SI; } else { Cond = Sel->getOperand(0); True = ConstantInt::get(Sel->getType(), 1); @@ -4383,8 +4391,12 @@ private: IRB.CreateGEP(Ty, FalseOps[0], ArrayRef(FalseOps).drop_front(), False->getName() + ".sroa.gep", NW); - Value *NSel = - IRB.CreateSelect(Cond, NTrue, NFalse, Sel->getName() + ".sroa.sel"); + Value *NSel = MDFrom + ? IRB.CreateSelect(Cond, NTrue, NFalse, + Sel->getName() + ".sroa.sel", MDFrom) + : IRB.CreateSelectWithUnknownProfile( + Cond, NTrue, NFalse, DEBUG_TYPE, + Sel->getName() + ".sroa.sel"); Visited.erase(&GEPI); GEPI.replaceAllUsesWith(NSel); GEPI.eraseFromParent(); diff --git a/llvm/lib/Transforms/Scalar/StructurizeCFG.cpp b/llvm/lib/Transforms/Scalar/StructurizeCFG.cpp index 2ee91a9..0f3978f 100644 --- a/llvm/lib/Transforms/Scalar/StructurizeCFG.cpp +++ b/llvm/lib/Transforms/Scalar/StructurizeCFG.cpp @@ -47,6 +47,7 @@ #include "llvm/Transforms/Utils/BasicBlockUtils.h" #include "llvm/Transforms/Utils/Local.h" #include "llvm/Transforms/Utils/SSAUpdater.h" +#include "llvm/Transforms/Utils/SSAUpdaterBulk.h" #include <cassert> #include <utility> @@ -321,7 +322,7 @@ class StructurizeCFG { void collectInfos(); - void insertConditions(bool Loops); + void insertConditions(bool Loops, SSAUpdaterBulk &PhiInserter); void simplifyConditions(); @@ -671,10 +672,9 @@ void StructurizeCFG::collectInfos() { } /// Insert the missing branch conditions -void StructurizeCFG::insertConditions(bool Loops) { +void StructurizeCFG::insertConditions(bool Loops, SSAUpdaterBulk &PhiInserter) { BranchVector &Conds = Loops ? LoopConds : Conditions; Value *Default = Loops ? BoolTrue : BoolFalse; - SSAUpdater PhiInserter; for (BranchInst *Term : Conds) { assert(Term->isConditional()); @@ -683,8 +683,9 @@ void StructurizeCFG::insertConditions(bool Loops) { BasicBlock *SuccTrue = Term->getSuccessor(0); BasicBlock *SuccFalse = Term->getSuccessor(1); - PhiInserter.Initialize(Boolean, ""); - PhiInserter.AddAvailableValue(Loops ? SuccFalse : Parent, Default); + unsigned Variable = PhiInserter.AddVariable("", Boolean); + PhiInserter.AddAvailableValue(Variable, Loops ? SuccFalse : Parent, + Default); BBPredicates &Preds = Loops ? LoopPreds[SuccFalse] : Predicates[SuccTrue]; @@ -697,7 +698,7 @@ void StructurizeCFG::insertConditions(bool Loops) { ParentInfo = PI; break; } - PhiInserter.AddAvailableValue(BB, PI.Pred); + PhiInserter.AddAvailableValue(Variable, BB, PI.Pred); Dominator.addAndRememberBlock(BB); } @@ -706,9 +707,9 @@ void StructurizeCFG::insertConditions(bool Loops) { CondBranchWeights::setMetadata(*Term, ParentInfo.Weights); } else { if (!Dominator.resultIsRememberedBlock()) - PhiInserter.AddAvailableValue(Dominator.result(), Default); + PhiInserter.AddAvailableValue(Variable, Dominator.result(), Default); - Term->setCondition(PhiInserter.GetValueInMiddleOfBlock(Parent)); + PhiInserter.AddUse(Variable, &Term->getOperandUse(0)); } } } @@ -1414,8 +1415,12 @@ bool StructurizeCFG::run(Region *R, DominatorTree *DT, orderNodes(); collectInfos(); createFlow(); - insertConditions(false); - insertConditions(true); + + SSAUpdaterBulk PhiInserter; + insertConditions(false, PhiInserter); + insertConditions(true, PhiInserter); + PhiInserter.RewriteAndOptimizeAllUses(*DT); + setPhiValues(); simplifyHoistedPhis(); simplifyConditions(); diff --git a/llvm/lib/Transforms/Utils/LoopUnroll.cpp b/llvm/lib/Transforms/Utils/LoopUnroll.cpp index b18acea..4fe736a 100644 --- a/llvm/lib/Transforms/Utils/LoopUnroll.cpp +++ b/llvm/lib/Transforms/Utils/LoopUnroll.cpp @@ -1106,7 +1106,6 @@ llvm::UnrollLoop(Loop *L, UnrollLoopOptions ULO, LoopInfo *LI, } Phi.replaceAllUsesWith(RdxResult); - continue; } } diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp index b96d29e..62a81ba 100644 --- a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp +++ b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp @@ -8240,14 +8240,14 @@ VPlanPtr LoopVectorizationPlanner::tryToBuildVPlanWithVPRecipes( // the vector loop or when not folding the tail. In the later case, we know // that the canonical induction increment will not overflow as the vector trip // count is >= increment and a multiple of the increment. + VPRegionBlock *LoopRegion = Plan->getVectorLoopRegion(); bool HasNUW = !IVUpdateMayOverflow || Style == TailFoldingStyle::None; if (!HasNUW) { - auto *IVInc = Plan->getVectorLoopRegion() - ->getExitingBasicBlock() - ->getTerminator() - ->getOperand(0); - assert(match(IVInc, m_VPInstruction<Instruction::Add>( - m_Specific(Plan->getCanonicalIV()), m_VPValue())) && + auto *IVInc = + LoopRegion->getExitingBasicBlock()->getTerminator()->getOperand(0); + assert(match(IVInc, + m_VPInstruction<Instruction::Add>( + m_Specific(LoopRegion->getCanonicalIV()), m_VPValue())) && "Did not find the canonical IV increment"); cast<VPRecipeWithIRFlags>(IVInc)->dropPoisonGeneratingFlags(); } @@ -8293,7 +8293,6 @@ VPlanPtr LoopVectorizationPlanner::tryToBuildVPlanWithVPRecipes( // Scan the body of the loop in a topological order to visit each basic block // after having visited its predecessor basic blocks. - VPRegionBlock *LoopRegion = Plan->getVectorLoopRegion(); VPBasicBlock *HeaderVPBB = LoopRegion->getEntryBasicBlock(); ReversePostOrderTraversal<VPBlockShallowTraversalWrapper<VPBlockBase *>> RPOT( HeaderVPBB); @@ -8377,8 +8376,8 @@ VPlanPtr LoopVectorizationPlanner::tryToBuildVPlanWithVPRecipes( for (VPValue *Old : Old2New.keys()) Old->getDefiningRecipe()->eraseFromParent(); - assert(isa<VPRegionBlock>(Plan->getVectorLoopRegion()) && - !Plan->getVectorLoopRegion()->getEntryBasicBlock()->empty() && + assert(isa<VPRegionBlock>(LoopRegion) && + !LoopRegion->getEntryBasicBlock()->empty() && "entry block must be set to a VPRegionBlock having a non-empty entry " "VPBasicBlock"); @@ -9326,8 +9325,9 @@ static void preparePlanForMainVectorLoop(VPlan &MainPlan, VPlan &EpiPlan) { if (ResumePhiIter == MainScalarPH->phis().end()) { VPBuilder ScalarPHBuilder(MainScalarPH, MainScalarPH->begin()); ResumePhi = ScalarPHBuilder.createScalarPhi( - {VectorTC, MainPlan.getCanonicalIV()->getStartValue()}, {}, - "vec.epilog.resume.val"); + {VectorTC, + MainPlan.getVectorLoopRegion()->getCanonicalIV()->getStartValue()}, + {}, "vec.epilog.resume.val"); } else { ResumePhi = cast<VPPhi>(&*ResumePhiIter); if (MainScalarPH->begin() == MainScalarPH->end()) @@ -9354,7 +9354,7 @@ static SmallVector<Instruction *> preparePlanForEpilogueVectorLoop( VPBasicBlock *Header = VectorLoop->getEntryBasicBlock(); Header->setName("vec.epilog.vector.body"); - VPCanonicalIVPHIRecipe *IV = Plan.getCanonicalIV(); + VPCanonicalIVPHIRecipe *IV = VectorLoop->getCanonicalIV(); // When vectorizing the epilogue loop, the canonical induction needs to be // adjusted by the value after the main vector loop. Find the resume value // created during execution of the main VPlan. It must be the first phi in the diff --git a/llvm/lib/Transforms/Vectorize/VPlan.h b/llvm/lib/Transforms/Vectorize/VPlan.h index 8ca3bed..0e0b042 100644 --- a/llvm/lib/Transforms/Vectorize/VPlan.h +++ b/llvm/lib/Transforms/Vectorize/VPlan.h @@ -24,12 +24,9 @@ #ifndef LLVM_TRANSFORMS_VECTORIZE_VPLAN_H #define LLVM_TRANSFORMS_VECTORIZE_VPLAN_H -#include "VPlanAnalysis.h" #include "VPlanValue.h" #include "llvm/ADT/DenseMap.h" -#include "llvm/ADT/SmallBitVector.h" #include "llvm/ADT/SmallPtrSet.h" -#include "llvm/ADT/SmallSet.h" #include "llvm/ADT/SmallVector.h" #include "llvm/ADT/Twine.h" #include "llvm/ADT/ilist.h" @@ -41,10 +38,11 @@ #include "llvm/IR/Operator.h" #include "llvm/Support/Compiler.h" #include "llvm/Support/InstructionCost.h" -#include <algorithm> #include <cassert> #include <cstddef> +#include <functional> #include <string> +#include <utility> namespace llvm { @@ -346,13 +344,6 @@ public: /// Return the cost of the block. virtual InstructionCost cost(ElementCount VF, VPCostContext &Ctx) = 0; - /// Return true if it is legal to hoist instructions into this block. - bool isLegalToHoistInto() { - // There are currently no constraints that prevent an instruction to be - // hoisted into a VPBlockBase. - return true; - } - #if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP) void printAsOperand(raw_ostream &OS, bool PrintType = false) const { OS << getName(); @@ -1021,6 +1012,8 @@ public: // part if scalar. In the latter case, the recipe will be removed during // unrolling. ExtractLastElement, + // Extracts the last lane for each part from its operand. + ExtractLastLanePerPart, // Extracts the second-to-last lane from its operand or the second-to-last // part if it is scalar. In the latter case, the recipe will be removed // during unrolling. @@ -4067,6 +4060,19 @@ public: /// Remove the current region from its VPlan, connecting its predecessor to /// its entry, and its exiting block to its successor. void dissolveToCFGLoop(); + + /// Returns the canonical induction recipe of the region. + VPCanonicalIVPHIRecipe *getCanonicalIV() { + VPBasicBlock *EntryVPBB = getEntryBasicBlock(); + if (EntryVPBB->empty()) { + // VPlan native path. TODO: Unify both code paths. + EntryVPBB = cast<VPBasicBlock>(EntryVPBB->getSingleSuccessor()); + } + return cast<VPCanonicalIVPHIRecipe>(&*EntryVPBB->begin()); + } + const VPCanonicalIVPHIRecipe *getCanonicalIV() const { + return const_cast<VPRegionBlock *>(this)->getCanonicalIV(); + } }; /// VPlan models a candidate for vectorization, encoding various decisions take @@ -4261,12 +4267,14 @@ public: BackedgeTakenCount = new VPValue(); return BackedgeTakenCount; } + VPValue *getBackedgeTakenCount() const { return BackedgeTakenCount; } /// The vector trip count. VPValue &getVectorTripCount() { return VectorTripCount; } /// Returns the VF of the vector loop region. VPValue &getVF() { return VF; }; + const VPValue &getVF() const { return VF; }; /// Returns VF * UF of the vector loop region. VPValue &getVFxUF() { return VFxUF; } @@ -4378,16 +4386,6 @@ public: LLVM_DUMP_METHOD void dump() const; #endif - /// Returns the canonical induction recipe of the vector loop. - VPCanonicalIVPHIRecipe *getCanonicalIV() { - VPBasicBlock *EntryVPBB = getVectorLoopRegion()->getEntryBasicBlock(); - if (EntryVPBB->empty()) { - // VPlan native path. - EntryVPBB = cast<VPBasicBlock>(EntryVPBB->getSingleSuccessor()); - } - return cast<VPCanonicalIVPHIRecipe>(&*EntryVPBB->begin()); - } - VPValue *getSCEVExpansion(const SCEV *S) const { return SCEVToExpansion.lookup(S); } diff --git a/llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp b/llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp index 07bfe7a..f413c63 100644 --- a/llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp +++ b/llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp @@ -116,6 +116,7 @@ Type *VPTypeAnalysis::inferScalarTypeForRecipe(const VPInstruction *R) { case VPInstruction::FirstActiveLane: return Type::getIntNTy(Ctx, 64); case VPInstruction::ExtractLastElement: + case VPInstruction::ExtractLastLanePerPart: case VPInstruction::ExtractPenultimateElement: { Type *BaseTy = inferScalarType(R->getOperand(0)); if (auto *VecTy = dyn_cast<VectorType>(BaseTy)) diff --git a/llvm/lib/Transforms/Vectorize/VPlanConstruction.cpp b/llvm/lib/Transforms/Vectorize/VPlanConstruction.cpp index c0147ce..332791a 100644 --- a/llvm/lib/Transforms/Vectorize/VPlanConstruction.cpp +++ b/llvm/lib/Transforms/Vectorize/VPlanConstruction.cpp @@ -658,9 +658,11 @@ void VPlanTransforms::attachCheckBlock(VPlan &Plan, Value *Cond, } VPIRMetadata VPBranchWeights; - auto *Term = VPBuilder(CheckBlockVPBB) - .createNaryOp(VPInstruction::BranchOnCond, {CondVPV}, - Plan.getCanonicalIV()->getDebugLoc()); + auto *Term = + VPBuilder(CheckBlockVPBB) + .createNaryOp( + VPInstruction::BranchOnCond, {CondVPV}, + Plan.getVectorLoopRegion()->getCanonicalIV()->getDebugLoc()); if (AddBranchWeights) { MDBuilder MDB(Plan.getContext()); MDNode *BranchWeights = @@ -921,8 +923,8 @@ bool VPlanTransforms::handleMaxMinNumReductions(VPlan &Plan) { if (auto *DerivedIV = dyn_cast<VPDerivedIVRecipe>(VecV)) { if (DerivedIV->getNumUsers() == 1 && DerivedIV->getOperand(1) == &Plan.getVectorTripCount()) { - auto *NewSel = Builder.createSelect(AnyNaN, Plan.getCanonicalIV(), - &Plan.getVectorTripCount()); + auto *NewSel = Builder.createSelect( + AnyNaN, LoopRegion->getCanonicalIV(), &Plan.getVectorTripCount()); DerivedIV->moveAfter(&*Builder.getInsertPoint()); DerivedIV->setOperand(1, NewSel); continue; @@ -935,7 +937,8 @@ bool VPlanTransforms::handleMaxMinNumReductions(VPlan &Plan) { "FMaxNum/FMinNum reduction.\n"); return false; } - auto *NewSel = Builder.createSelect(AnyNaN, Plan.getCanonicalIV(), VecV); + auto *NewSel = + Builder.createSelect(AnyNaN, LoopRegion->getCanonicalIV(), VecV); ResumeR->setOperand(0, NewSel); } diff --git a/llvm/lib/Transforms/Vectorize/VPlanPatternMatch.h b/llvm/lib/Transforms/Vectorize/VPlanPatternMatch.h index b42b049..ff286f7 100644 --- a/llvm/lib/Transforms/Vectorize/VPlanPatternMatch.h +++ b/llvm/lib/Transforms/Vectorize/VPlanPatternMatch.h @@ -372,6 +372,12 @@ m_ExtractLastElement(const Op0_t &Op0) { return m_VPInstruction<VPInstruction::ExtractLastElement>(Op0); } +template <typename Op0_t> +inline VPInstruction_match<VPInstruction::ExtractLastLanePerPart, Op0_t> +m_ExtractLastLanePerPart(const Op0_t &Op0) { + return m_VPInstruction<VPInstruction::ExtractLastLanePerPart>(Op0); +} + template <typename Op0_t, typename Op1_t, typename Op2_t> inline VPInstruction_match<VPInstruction::ActiveLaneMask, Op0_t, Op1_t, Op2_t> m_ActiveLaneMask(const Op0_t &Op0, const Op1_t &Op1, const Op2_t &Op2) { @@ -394,6 +400,12 @@ m_AnyOf(const Op0_t &Op0) { return m_VPInstruction<VPInstruction::AnyOf>(Op0); } +template <typename Op0_t> +inline VPInstruction_match<VPInstruction::FirstActiveLane, Op0_t> +m_FirstActiveLane(const Op0_t &Op0) { + return m_VPInstruction<VPInstruction::FirstActiveLane>(Op0); +} + template <unsigned Opcode, typename Op0_t> inline AllRecipe_match<Opcode, Op0_t> m_Unary(const Op0_t &Op0) { return AllRecipe_match<Opcode, Op0_t>(Op0); diff --git a/llvm/lib/Transforms/Vectorize/VPlanPredicator.cpp b/llvm/lib/Transforms/Vectorize/VPlanPredicator.cpp index 0c27d53..fb17d5d 100644 --- a/llvm/lib/Transforms/Vectorize/VPlanPredicator.cpp +++ b/llvm/lib/Transforms/Vectorize/VPlanPredicator.cpp @@ -168,7 +168,8 @@ void VPPredicator::createHeaderMask(VPBasicBlock *HeaderVPBB, bool FoldTail) { // non-phi instructions. auto &Plan = *HeaderVPBB->getPlan(); - auto *IV = new VPWidenCanonicalIVRecipe(Plan.getCanonicalIV()); + auto *IV = + new VPWidenCanonicalIVRecipe(HeaderVPBB->getParent()->getCanonicalIV()); Builder.setInsertPoint(HeaderVPBB, HeaderVPBB->getFirstNonPhi()); Builder.insert(IV); diff --git a/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp b/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp index 2368d18..775837f 100644 --- a/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp +++ b/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp @@ -511,6 +511,7 @@ unsigned VPInstruction::getNumOperandsForOpcode(unsigned Opcode) { case VPInstruction::CanonicalIVIncrementForPart: case VPInstruction::ExplicitVectorLength: case VPInstruction::ExtractLastElement: + case VPInstruction::ExtractLastLanePerPart: case VPInstruction::ExtractPenultimateElement: case VPInstruction::FirstActiveLane: case VPInstruction::Not: @@ -878,9 +879,11 @@ Value *VPInstruction::generate(VPTransformState &State) { return ReducedPartRdx; } + case VPInstruction::ExtractLastLanePerPart: case VPInstruction::ExtractLastElement: case VPInstruction::ExtractPenultimateElement: { - unsigned Offset = getOpcode() == VPInstruction::ExtractLastElement ? 1 : 2; + unsigned Offset = + getOpcode() == VPInstruction::ExtractPenultimateElement ? 2 : 1; Value *Res; if (State.VF.isVector()) { assert(Offset <= State.VF.getKnownMinValue() && @@ -1166,6 +1169,7 @@ InstructionCost VPInstruction::computeCost(ElementCount VF, bool VPInstruction::isVectorToScalar() const { return getOpcode() == VPInstruction::ExtractLastElement || + getOpcode() == VPInstruction::ExtractLastLanePerPart || getOpcode() == VPInstruction::ExtractPenultimateElement || getOpcode() == Instruction::ExtractElement || getOpcode() == VPInstruction::ExtractLane || @@ -1229,6 +1233,7 @@ bool VPInstruction::opcodeMayReadOrWriteFromMemory() const { case VPInstruction::CanonicalIVIncrementForPart: case VPInstruction::ExtractLane: case VPInstruction::ExtractLastElement: + case VPInstruction::ExtractLastLanePerPart: case VPInstruction::ExtractPenultimateElement: case VPInstruction::ActiveLaneMask: case VPInstruction::FirstActiveLane: @@ -1376,6 +1381,9 @@ void VPInstruction::print(raw_ostream &O, const Twine &Indent, case VPInstruction::ExtractLastElement: O << "extract-last-element"; break; + case VPInstruction::ExtractLastLanePerPart: + O << "extract-last-lane-per-part"; + break; case VPInstruction::ExtractPenultimateElement: O << "extract-penultimate-element"; break; @@ -2344,7 +2352,7 @@ bool VPWidenIntOrFpInductionRecipe::isCanonical() const { return false; auto *StepC = dyn_cast<ConstantInt>(getStepValue()->getLiveInIRValue()); auto *StartC = dyn_cast<ConstantInt>(getStartValue()->getLiveInIRValue()); - auto *CanIV = cast<VPCanonicalIVPHIRecipe>(&*getParent()->begin()); + auto *CanIV = getParent()->getParent()->getCanonicalIV(); return StartC && StartC->isZero() && StepC && StepC->isOne() && getScalarType() == CanIV->getScalarType(); } diff --git a/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp b/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp index 40b7e8d..8d76b2d8 100644 --- a/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp +++ b/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp @@ -501,7 +501,8 @@ static void removeRedundantInductionCasts(VPlan &Plan) { /// Try to replace VPWidenCanonicalIVRecipes with a widened canonical IV /// recipe, if it exists. static void removeRedundantCanonicalIVs(VPlan &Plan) { - VPCanonicalIVPHIRecipe *CanonicalIV = Plan.getCanonicalIV(); + VPRegionBlock *LoopRegion = Plan.getVectorLoopRegion(); + VPCanonicalIVPHIRecipe *CanonicalIV = LoopRegion->getCanonicalIV(); VPWidenCanonicalIVRecipe *WidenNewIV = nullptr; for (VPUser *U : CanonicalIV->users()) { WidenNewIV = dyn_cast<VPWidenCanonicalIVRecipe>(U); @@ -512,7 +513,7 @@ static void removeRedundantCanonicalIVs(VPlan &Plan) { if (!WidenNewIV) return; - VPBasicBlock *HeaderVPBB = Plan.getVectorLoopRegion()->getEntryBasicBlock(); + VPBasicBlock *HeaderVPBB = LoopRegion->getEntryBasicBlock(); for (VPRecipeBase &Phi : HeaderVPBB->phis()) { auto *WidenOriginalIV = dyn_cast<VPWidenIntOrFpInductionRecipe>(&Phi); @@ -582,8 +583,9 @@ createScalarIVSteps(VPlan &Plan, InductionDescriptor::InductionKind Kind, FPMathOperator *FPBinOp, Instruction *TruncI, VPValue *StartV, VPValue *Step, DebugLoc DL, VPBuilder &Builder) { - VPBasicBlock *HeaderVPBB = Plan.getVectorLoopRegion()->getEntryBasicBlock(); - VPCanonicalIVPHIRecipe *CanonicalIV = Plan.getCanonicalIV(); + VPRegionBlock *LoopRegion = Plan.getVectorLoopRegion(); + VPBasicBlock *HeaderVPBB = LoopRegion->getEntryBasicBlock(); + VPCanonicalIVPHIRecipe *CanonicalIV = LoopRegion->getCanonicalIV(); VPSingleDefRecipe *BaseIV = Builder.createDerivedIV( Kind, FPBinOp, StartV, CanonicalIV, Step, "offset.idx"); @@ -786,9 +788,7 @@ static VPValue *optimizeEarlyExitInductionUser(VPlan &Plan, ScalarEvolution &SE) { VPValue *Incoming, *Mask; if (!match(Op, m_VPInstruction<VPInstruction::ExtractLane>( - m_VPInstruction<VPInstruction::FirstActiveLane>( - m_VPValue(Mask)), - m_VPValue(Incoming)))) + m_FirstActiveLane(m_VPValue(Mask)), m_VPValue(Incoming)))) return nullptr; auto *WideIV = getOptimizableIVOf(Incoming, SE); @@ -800,8 +800,9 @@ static VPValue *optimizeEarlyExitInductionUser(VPlan &Plan, return nullptr; // Calculate the final index. - VPValue *EndValue = Plan.getCanonicalIV(); - auto CanonicalIVType = Plan.getCanonicalIV()->getScalarType(); + VPRegionBlock *LoopRegion = Plan.getVectorLoopRegion(); + auto *CanonicalIV = LoopRegion->getCanonicalIV(); + Type *CanonicalIVType = CanonicalIV->getScalarType(); VPBuilder B(cast<VPBasicBlock>(PredVPBB)); DebugLoc DL = cast<VPInstruction>(Op)->getDebugLoc(); @@ -810,7 +811,8 @@ static VPValue *optimizeEarlyExitInductionUser(VPlan &Plan, Type *FirstActiveLaneType = TypeInfo.inferScalarType(FirstActiveLane); FirstActiveLane = B.createScalarZExtOrTrunc(FirstActiveLane, CanonicalIVType, FirstActiveLaneType, DL); - EndValue = B.createNaryOp(Instruction::Add, {EndValue, FirstActiveLane}, DL); + VPValue *EndValue = + B.createNaryOp(Instruction::Add, {CanonicalIV, FirstActiveLane}, DL); // `getOptimizableIVOf()` always returns the pre-incremented IV, so if it // changed it means the exit is using the incremented value, so we need to @@ -1205,7 +1207,8 @@ static void simplifyRecipe(VPRecipeBase &R, VPTypeAnalysis &TypeInfo) { } // Look through ExtractLastElement (BuildVector ....). - if (match(&R, m_ExtractLastElement(m_BuildVector()))) { + if (match(&R, m_CombineOr(m_ExtractLastElement(m_BuildVector()), + m_ExtractLastLanePerPart(m_BuildVector())))) { auto *BuildVector = cast<VPInstruction>(R.getOperand(0)); Def->replaceAllUsesWith( BuildVector->getOperand(BuildVector->getNumOperands() - 1)); @@ -1271,13 +1274,15 @@ static void simplifyRecipe(VPRecipeBase &R, VPTypeAnalysis &TypeInfo) { return; } - if (match(Def, m_ExtractLastElement(m_Broadcast(m_VPValue(A))))) { + if (match(Def, + m_CombineOr(m_ExtractLastElement(m_Broadcast(m_VPValue(A))), + m_ExtractLastLanePerPart(m_Broadcast(m_VPValue(A)))))) { Def->replaceAllUsesWith(A); return; } - if (match(Def, - m_VPInstruction<VPInstruction::ExtractLastElement>(m_VPValue(A))) && + if (match(Def, m_CombineOr(m_ExtractLastElement(m_VPValue(A)), + m_ExtractLastLanePerPart(m_VPValue(A)))) && ((isa<VPInstruction>(A) && vputils::isSingleScalar(A)) || (isa<VPReplicateRecipe>(A) && cast<VPReplicateRecipe>(A)->isSingleScalar())) && @@ -1285,6 +1290,12 @@ static void simplifyRecipe(VPRecipeBase &R, VPTypeAnalysis &TypeInfo) { [Def, A](VPUser *U) { return U->usesScalars(A) || Def == U; })) { return Def->replaceAllUsesWith(A); } + + if (Plan->getUF() == 1 && + match(Def, m_ExtractLastLanePerPart(m_VPValue(A)))) { + return Def->replaceAllUsesWith( + Builder.createNaryOp(VPInstruction::ExtractLastElement, {A})); + } } void VPlanTransforms::simplifyRecipes(VPlan &Plan) { @@ -1322,8 +1333,11 @@ static void narrowToSingleScalarRecipes(VPlan &Plan) { RepOrWidenR->getUnderlyingInstr(), RepOrWidenR->operands(), true /*IsSingleScalar*/, nullptr /*Mask*/, *RepR /*Metadata*/); Clone->insertBefore(RepOrWidenR); - auto *Ext = new VPInstruction(VPInstruction::ExtractLastElement, - {Clone->getOperand(0)}); + unsigned ExtractOpc = + vputils::isUniformAcrossVFsAndUFs(RepR->getOperand(1)) + ? VPInstruction::ExtractLastElement + : VPInstruction::ExtractLastLanePerPart; + auto *Ext = new VPInstruction(ExtractOpc, {Clone->getOperand(0)}); Ext->insertBefore(Clone); Clone->setOperand(0, Ext); RepR->eraseFromParent(); @@ -1337,7 +1351,8 @@ static void narrowToSingleScalarRecipes(VPlan &Plan) { !all_of(RepOrWidenR->users(), [RepOrWidenR](const VPUser *U) { return U->usesScalars(RepOrWidenR) || match(cast<VPRecipeBase>(U), - m_ExtractLastElement(m_VPValue())); + m_CombineOr(m_ExtractLastElement(m_VPValue()), + m_ExtractLastLanePerPart(m_VPValue()))); })) continue; @@ -1530,7 +1545,7 @@ static bool isConditionTrueViaVFAndUF(VPValue *Cond, VPlan &Plan, return isConditionTrueViaVFAndUF(C, Plan, BestVF, BestUF, SE); }); - auto *CanIV = Plan.getCanonicalIV(); + auto *CanIV = Plan.getVectorLoopRegion()->getCanonicalIV(); if (!match(Cond, m_SpecificICmp(CmpInst::ICMP_EQ, m_Specific(CanIV->getBackedgeValue()), m_Specific(&Plan.getVectorTripCount())))) @@ -2319,7 +2334,7 @@ static VPActiveLaneMaskPHIRecipe *addVPLaneMaskPhiAndUpdateExitBranch( VPlan &Plan, bool DataAndControlFlowWithoutRuntimeCheck) { VPRegionBlock *TopRegion = Plan.getVectorLoopRegion(); VPBasicBlock *EB = TopRegion->getExitingBasicBlock(); - auto *CanonicalIVPHI = Plan.getCanonicalIV(); + auto *CanonicalIVPHI = TopRegion->getCanonicalIV(); VPValue *StartV = CanonicalIVPHI->getStartValue(); auto *CanonicalIVIncrement = @@ -2358,7 +2373,7 @@ static VPActiveLaneMaskPHIRecipe *addVPLaneMaskPhiAndUpdateExitBranch( // Create the active lane mask instruction in the VPlan preheader. VPValue *ALMMultiplier = Plan.getOrAddLiveIn( - ConstantInt::get(Plan.getCanonicalIV()->getScalarType(), 1)); + ConstantInt::get(TopRegion->getCanonicalIV()->getScalarType(), 1)); auto *EntryALM = Builder.createNaryOp(VPInstruction::ActiveLaneMask, {EntryIncrement, TC, ALMMultiplier}, DL, "active.lane.mask.entry"); @@ -2394,13 +2409,15 @@ static VPActiveLaneMaskPHIRecipe *addVPLaneMaskPhiAndUpdateExitBranch( /// TODO: Introduce explicit recipe for header-mask instead of searching /// for the header-mask pattern manually. static VPSingleDefRecipe *findHeaderMask(VPlan &Plan) { + VPRegionBlock *LoopRegion = Plan.getVectorLoopRegion(); SmallVector<VPValue *> WideCanonicalIVs; - auto *FoundWidenCanonicalIVUser = find_if(Plan.getCanonicalIV()->users(), - IsaPred<VPWidenCanonicalIVRecipe>); - assert(count_if(Plan.getCanonicalIV()->users(), + auto *FoundWidenCanonicalIVUser = find_if( + LoopRegion->getCanonicalIV()->users(), IsaPred<VPWidenCanonicalIVRecipe>); + assert(count_if(LoopRegion->getCanonicalIV()->users(), IsaPred<VPWidenCanonicalIVRecipe>) <= 1 && "Must have at most one VPWideCanonicalIVRecipe"); - if (FoundWidenCanonicalIVUser != Plan.getCanonicalIV()->users().end()) { + if (FoundWidenCanonicalIVUser != + LoopRegion->getCanonicalIV()->users().end()) { auto *WideCanonicalIV = cast<VPWidenCanonicalIVRecipe>(*FoundWidenCanonicalIVUser); WideCanonicalIVs.push_back(WideCanonicalIV); @@ -2408,7 +2425,7 @@ static VPSingleDefRecipe *findHeaderMask(VPlan &Plan) { // Also include VPWidenIntOrFpInductionRecipes that represent a widened // version of the canonical induction. - VPBasicBlock *HeaderVPBB = Plan.getVectorLoopRegion()->getEntryBasicBlock(); + VPBasicBlock *HeaderVPBB = LoopRegion->getEntryBasicBlock(); for (VPRecipeBase &Phi : HeaderVPBB->phis()) { auto *WidenOriginalIV = dyn_cast<VPWidenIntOrFpInductionRecipe>(&Phi); if (WidenOriginalIV && WidenOriginalIV->isCanonical()) @@ -2441,8 +2458,9 @@ void VPlanTransforms::addActiveLaneMask( "DataAndControlFlowWithoutRuntimeCheck implies " "UseActiveLaneMaskForControlFlow"); - auto *FoundWidenCanonicalIVUser = find_if(Plan.getCanonicalIV()->users(), - IsaPred<VPWidenCanonicalIVRecipe>); + VPRegionBlock *LoopRegion = Plan.getVectorLoopRegion(); + auto *FoundWidenCanonicalIVUser = find_if( + LoopRegion->getCanonicalIV()->users(), IsaPred<VPWidenCanonicalIVRecipe>); assert(FoundWidenCanonicalIVUser && "Must have widened canonical IV when tail folding!"); VPSingleDefRecipe *HeaderMask = findHeaderMask(Plan); @@ -2455,7 +2473,7 @@ void VPlanTransforms::addActiveLaneMask( } else { VPBuilder B = VPBuilder::getToInsertAfter(WideCanonicalIV); VPValue *ALMMultiplier = Plan.getOrAddLiveIn( - ConstantInt::get(Plan.getCanonicalIV()->getScalarType(), 1)); + ConstantInt::get(LoopRegion->getCanonicalIV()->getScalarType(), 1)); LaneMask = B.createNaryOp(VPInstruction::ActiveLaneMask, {WideCanonicalIV, Plan.getTripCount(), ALMMultiplier}, @@ -2565,9 +2583,10 @@ static void transformRecipestoEVLRecipes(VPlan &Plan, VPValue &EVL) { }); assert(all_of(Plan.getVFxUF().users(), - [&Plan](VPUser *U) { - return match(U, m_c_Add(m_Specific(Plan.getCanonicalIV()), - m_Specific(&Plan.getVFxUF()))) || + [&LoopRegion, &Plan](VPUser *U) { + return match(U, + m_c_Add(m_Specific(LoopRegion->getCanonicalIV()), + m_Specific(&Plan.getVFxUF()))) || isa<VPWidenPointerInductionRecipe>(U); }) && "Only users of VFxUF should be VPWidenPointerInductionRecipe and the " @@ -2722,9 +2741,10 @@ void VPlanTransforms::addExplicitVectorLength( VPlan &Plan, const std::optional<unsigned> &MaxSafeElements) { if (Plan.hasScalarVFOnly()) return; - VPBasicBlock *Header = Plan.getVectorLoopRegion()->getEntryBasicBlock(); + VPRegionBlock *LoopRegion = Plan.getVectorLoopRegion(); + VPBasicBlock *Header = LoopRegion->getEntryBasicBlock(); - auto *CanonicalIVPHI = Plan.getCanonicalIV(); + auto *CanonicalIVPHI = LoopRegion->getCanonicalIV(); auto *CanIVTy = CanonicalIVPHI->getScalarType(); VPValue *StartV = CanonicalIVPHI->getStartValue(); @@ -4164,7 +4184,7 @@ void VPlanTransforms::narrowInterleaveGroups(VPlan &Plan, ElementCount VF, // Adjust induction to reflect that the transformed plan only processes one // original iteration. - auto *CanIV = Plan.getCanonicalIV(); + auto *CanIV = VectorLoop->getCanonicalIV(); auto *Inc = cast<VPInstruction>(CanIV->getBackedgeValue()); VPBuilder PHBuilder(Plan.getVectorPreheader()); diff --git a/llvm/lib/Transforms/Vectorize/VPlanUnroll.cpp b/llvm/lib/Transforms/Vectorize/VPlanUnroll.cpp index 1c4adfc..5aeda3e 100644 --- a/llvm/lib/Transforms/Vectorize/VPlanUnroll.cpp +++ b/llvm/lib/Transforms/Vectorize/VPlanUnroll.cpp @@ -69,7 +69,8 @@ class UnrollState { VPBasicBlock::iterator InsertPtForPhi); VPValue *getConstantVPV(unsigned Part) { - Type *CanIVIntTy = Plan.getCanonicalIV()->getScalarType(); + Type *CanIVIntTy = + Plan.getVectorLoopRegion()->getCanonicalIV()->getScalarType(); return Plan.getOrAddLiveIn(ConstantInt::get(CanIVIntTy, Part)); } @@ -351,8 +352,7 @@ void UnrollState::unrollBlock(VPBlockBase *VPB) { // Compute*Result which combine all parts to compute the final value. VPValue *Op1; if (match(&R, m_VPInstruction<VPInstruction::AnyOf>(m_VPValue(Op1))) || - match(&R, m_VPInstruction<VPInstruction::FirstActiveLane>( - m_VPValue(Op1))) || + match(&R, m_FirstActiveLane(m_VPValue(Op1))) || match(&R, m_VPInstruction<VPInstruction::ComputeAnyOfResult>( m_VPValue(), m_VPValue(), m_VPValue(Op1))) || match(&R, m_VPInstruction<VPInstruction::ComputeReductionResult>( diff --git a/llvm/lib/Transforms/Vectorize/VPlanUtils.cpp b/llvm/lib/Transforms/Vectorize/VPlanUtils.cpp index 66748c5..8b1b0e5 100644 --- a/llvm/lib/Transforms/Vectorize/VPlanUtils.cpp +++ b/llvm/lib/Transforms/Vectorize/VPlanUtils.cpp @@ -53,7 +53,7 @@ VPValue *vputils::getOrCreateVPValueForSCEVExpr(VPlan &Plan, const SCEV *Expr) { return Expanded; } -bool vputils::isHeaderMask(const VPValue *V, VPlan &Plan) { +bool vputils::isHeaderMask(const VPValue *V, const VPlan &Plan) { if (isa<VPActiveLaneMaskPHIRecipe>(V)) return true; @@ -67,12 +67,14 @@ bool vputils::isHeaderMask(const VPValue *V, VPlan &Plan) { if (match(V, m_ActiveLaneMask(m_VPValue(A), m_VPValue(B), m_One()))) return B == Plan.getTripCount() && - (match(A, m_ScalarIVSteps(m_Specific(Plan.getCanonicalIV()), m_One(), - m_Specific(&Plan.getVF()))) || + (match(A, + m_ScalarIVSteps( + m_Specific(Plan.getVectorLoopRegion()->getCanonicalIV()), + m_One(), m_Specific(&Plan.getVF()))) || IsWideCanonicalIV(A)); return match(V, m_ICmp(m_VPValue(A), m_VPValue(B))) && IsWideCanonicalIV(A) && - B == Plan.getOrCreateBackedgeTakenCount(); + B == Plan.getBackedgeTakenCount(); } const SCEV *vputils::getSCEVExprForVPValue(VPValue *V, ScalarEvolution &SE) { @@ -102,7 +104,8 @@ bool vputils::isUniformAcrossVFsAndUFs(VPValue *V) { return all_of(R->operands(), isUniformAcrossVFsAndUFs); } - auto *CanonicalIV = R->getParent()->getPlan()->getCanonicalIV(); + auto *CanonicalIV = + R->getParent()->getEnclosingLoopRegion()->getCanonicalIV(); // Canonical IV chain is uniform. if (V == CanonicalIV || V == CanonicalIV->getBackedgeValue()) return true; diff --git a/llvm/lib/Transforms/Vectorize/VPlanUtils.h b/llvm/lib/Transforms/Vectorize/VPlanUtils.h index 0222b0a..cf95ac0 100644 --- a/llvm/lib/Transforms/Vectorize/VPlanUtils.h +++ b/llvm/lib/Transforms/Vectorize/VPlanUtils.h @@ -90,7 +90,7 @@ inline bool isSingleScalar(const VPValue *VPV) { } /// Return true if \p V is a header mask in \p Plan. -bool isHeaderMask(const VPValue *V, VPlan &Plan); +bool isHeaderMask(const VPValue *V, const VPlan &Plan); /// Checks if \p V is uniform across all VF lanes and UF parts. It is considered /// as such if it is either loop invariant (defined outside the vector region) diff --git a/llvm/lib/Transforms/Vectorize/VPlanVerifier.cpp b/llvm/lib/Transforms/Vectorize/VPlanVerifier.cpp index 5262af6..91734a1 100644 --- a/llvm/lib/Transforms/Vectorize/VPlanVerifier.cpp +++ b/llvm/lib/Transforms/Vectorize/VPlanVerifier.cpp @@ -298,11 +298,16 @@ bool VPlanVerifier::verifyVPBasicBlock(const VPBasicBlock *VPBB) { return false; } } - if (const auto *EVL = dyn_cast<VPInstruction>(&R)) { - if (EVL->getOpcode() == VPInstruction::ExplicitVectorLength && - !verifyEVLRecipe(*EVL)) { - errs() << "EVL VPValue is not used correctly\n"; - return false; + if (const auto *VPI = dyn_cast<VPInstruction>(&R)) { + switch (VPI->getOpcode()) { + case VPInstruction::ExplicitVectorLength: + if (!verifyEVLRecipe(*VPI)) { + errs() << "EVL VPValue is not used correctly\n"; + return false; + } + break; + default: + break; } } } diff --git a/llvm/test/Analysis/BasicAA/intrinsics.ll b/llvm/test/Analysis/BasicAA/intrinsics.ll index f8b30df..56d762b 100644 --- a/llvm/test/Analysis/BasicAA/intrinsics.ll +++ b/llvm/test/Analysis/BasicAA/intrinsics.ll @@ -1,3 +1,4 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 6 ; RUN: opt -aa-pipeline=basic-aa -passes=gvn -S < %s | FileCheck %s target datalayout = "e-p:32:32:32-i1:8:32-i8:8:32-i16:16:32-i32:32:32-i64:32:32-f32:32:32-f64:32:32-v64:32:64-v128:32:128-a0:0:32-n32" @@ -5,12 +6,15 @@ target datalayout = "e-p:32:32:32-i1:8:32-i8:8:32-i16:16:32-i32:32:32-i64:32:32- ; BasicAA should prove that these calls don't interfere, since they are ; IntrArgReadMem and have noalias pointers. -; CHECK: define <8 x i16> @test0(ptr noalias %p, ptr noalias %q, <8 x i16> %y, <8 x i1> %m, <8 x i16> %pt) { -; CHECK-NEXT: entry: -; CHECK-NEXT: %a = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr %p, i32 16, <8 x i1> %m, <8 x i16> %pt) [[ATTR:#[0-9]+]] -; CHECK-NEXT: call void @llvm.masked.store.v8i16.p0(<8 x i16> %y, ptr %q, i32 16, <8 x i1> %m) -; CHECK-NEXT: %c = add <8 x i16> %a, %a define <8 x i16> @test0(ptr noalias %p, ptr noalias %q, <8 x i16> %y, <8 x i1> %m, <8 x i16> %pt) { +; CHECK-LABEL: define <8 x i16> @test0( +; CHECK-SAME: ptr noalias [[P:%.*]], ptr noalias [[Q:%.*]], <8 x i16> [[Y:%.*]], <8 x i1> [[M:%.*]], <8 x i16> [[PT:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[A:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[P]], i32 16, <8 x i1> [[M]], <8 x i16> [[PT]]) #[[ATTR2:[0-9]+]] +; CHECK-NEXT: call void @llvm.masked.store.v8i16.p0(<8 x i16> [[Y]], ptr [[Q]], i32 16, <8 x i1> [[M]]) +; CHECK-NEXT: [[C:%.*]] = add <8 x i16> [[A]], [[A]] +; CHECK-NEXT: ret <8 x i16> [[C]] +; entry: %a = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr %p, i32 16, <8 x i1> %m, <8 x i16> %pt) nounwind call void @llvm.masked.store.v8i16.p0(<8 x i16> %y, ptr %q, i32 16, <8 x i1> %m) @@ -24,4 +28,3 @@ declare void @llvm.masked.store.v8i16.p0(<8 x i16>, ptr, i32, <8 x i1>) nounwind ; CHECK: attributes #0 = { nocallback nofree nosync nounwind willreturn memory(argmem: read) } ; CHECK: attributes #1 = { nocallback nofree nosync nounwind willreturn memory(argmem: write) } -; CHECK: attributes [[ATTR]] = { nounwind } diff --git a/llvm/test/Analysis/BasicAA/scalable-dse-aa.ll b/llvm/test/Analysis/BasicAA/scalable-dse-aa.ll index 7e980c9..ffd8259 100644 --- a/llvm/test/Analysis/BasicAA/scalable-dse-aa.ll +++ b/llvm/test/Analysis/BasicAA/scalable-dse-aa.ll @@ -1,10 +1,23 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 6 ; RUN: opt < %s -aa-pipeline=basic-aa -passes=dse -S | FileCheck %s define <vscale x 4 x float> @dead_scalable_store(ptr %0) { ; CHECK-LABEL: define <vscale x 4 x float> @dead_scalable_store( -; CHECK: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> %load.0.16, ptr nonnull %gep.arr.16, i32 1, <vscale x 4 x i1> %mask) -; CHECK-NOT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> %load.0.32, ptr nonnull %gep.arr.32, i32 1, <vscale x 4 x i1> %mask) -; CHECK: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> %load.0.48, ptr nonnull %gep.arr.48, i32 1, <vscale x 4 x i1> %mask) +; CHECK-SAME: ptr [[TMP0:%.*]]) { +; CHECK-NEXT: [[ARR:%.*]] = alloca [64 x i32], align 4 +; CHECK-NEXT: [[MASK:%.*]] = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 0, i32 4) +; CHECK-NEXT: [[GEP_0_16:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP0]], i64 16 +; CHECK-NEXT: [[GEP_0_48:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP0]], i64 48 +; CHECK-NEXT: [[GEP_ARR_16:%.*]] = getelementptr inbounds nuw i8, ptr [[ARR]], i64 16 +; CHECK-NEXT: [[GEP_ARR_48:%.*]] = getelementptr inbounds nuw i8, ptr [[ARR]], i64 48 +; CHECK-NEXT: [[LOAD_0_16:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_0_16]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[LOAD_0_16]], ptr nonnull [[GEP_ARR_16]], i32 1, <vscale x 4 x i1> [[MASK]]) +; CHECK-NEXT: [[LOAD_0_48:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_0_48]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[LOAD_0_48]], ptr nonnull [[GEP_ARR_48]], i32 1, <vscale x 4 x i1> [[MASK]]) +; CHECK-NEXT: [[FADDOP0:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_ARR_16]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: [[FADDOP1:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_ARR_48]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: [[FADD:%.*]] = fadd <vscale x 4 x float> [[FADDOP0]], [[FADDOP1]] +; CHECK-NEXT: ret <vscale x 4 x float> [[FADD]] ; %arr = alloca [64 x i32], align 4 %mask = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 0, i32 4) @@ -34,9 +47,21 @@ define <vscale x 4 x float> @dead_scalable_store(ptr %0) { define <4 x float> @dead_scalable_store_fixed(ptr %0) { ; CHECK-LABEL: define <4 x float> @dead_scalable_store_fixed( -; CHECK: call void @llvm.masked.store.v4f32.p0(<4 x float> %load.0.16, ptr nonnull %gep.arr.16, i32 1, <4 x i1> %mask) -; CHECK-NOT: call void @llvm.masked.store.v4f32.p0(<4 x float> %load.0.32, ptr nonnull %gep.arr.36, i32 1, <4 x i1> %mask2) -; CHECK: call void @llvm.masked.store.v4f32.p0(<4 x float> %load.0.48, ptr nonnull %gep.arr.48, i32 1, <4 x i1> %mask) +; CHECK-SAME: ptr [[TMP0:%.*]]) { +; CHECK-NEXT: [[ARR:%.*]] = alloca [64 x i32], align 4 +; CHECK-NEXT: [[MASK:%.*]] = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 0, i32 4) +; CHECK-NEXT: [[GEP_0_16:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP0]], i64 16 +; CHECK-NEXT: [[GEP_0_48:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP0]], i64 48 +; CHECK-NEXT: [[GEP_ARR_16:%.*]] = getelementptr inbounds nuw i8, ptr [[ARR]], i64 16 +; CHECK-NEXT: [[GEP_ARR_48:%.*]] = getelementptr inbounds nuw i8, ptr [[ARR]], i64 48 +; CHECK-NEXT: [[LOAD_0_16:%.*]] = call <4 x float> @llvm.masked.load.v4f32.p0(ptr nonnull [[GEP_0_16]], i32 1, <4 x i1> [[MASK]], <4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.v4f32.p0(<4 x float> [[LOAD_0_16]], ptr nonnull [[GEP_ARR_16]], i32 1, <4 x i1> [[MASK]]) +; CHECK-NEXT: [[LOAD_0_48:%.*]] = call <4 x float> @llvm.masked.load.v4f32.p0(ptr nonnull [[GEP_0_48]], i32 1, <4 x i1> [[MASK]], <4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.v4f32.p0(<4 x float> [[LOAD_0_48]], ptr nonnull [[GEP_ARR_48]], i32 1, <4 x i1> [[MASK]]) +; CHECK-NEXT: [[FADDOP0:%.*]] = call <4 x float> @llvm.masked.load.v4f32.p0(ptr nonnull [[GEP_ARR_16]], i32 1, <4 x i1> [[MASK]], <4 x float> zeroinitializer) +; CHECK-NEXT: [[FADDOP1:%.*]] = call <4 x float> @llvm.masked.load.v4f32.p0(ptr nonnull [[GEP_ARR_48]], i32 1, <4 x i1> [[MASK]], <4 x float> zeroinitializer) +; CHECK-NEXT: [[FADD:%.*]] = fadd <4 x float> [[FADDOP0]], [[FADDOP1]] +; CHECK-NEXT: ret <4 x float> [[FADD]] ; %arr = alloca [64 x i32], align 4 %mask = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 0, i32 4) @@ -67,9 +92,25 @@ define <4 x float> @dead_scalable_store_fixed(ptr %0) { define <vscale x 4 x float> @scalable_store_partial_overwrite(ptr %0) { ; CHECK-LABEL: define <vscale x 4 x float> @scalable_store_partial_overwrite( -; CHECK: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> %load.0.16, ptr nonnull %gep.arr.16, i32 1, <vscale x 4 x i1> %mask) -; CHECK: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> %load.0.30, ptr nonnull %gep.arr.30, i32 1, <vscale x 4 x i1> %mask) -; CHECK: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> %load.0.48, ptr nonnull %gep.arr.48, i32 1, <vscale x 4 x i1> %mask) +; CHECK-SAME: ptr [[TMP0:%.*]]) { +; CHECK-NEXT: [[ARR:%.*]] = alloca [64 x i32], align 4 +; CHECK-NEXT: [[MASK:%.*]] = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 0, i32 4) +; CHECK-NEXT: [[GEP_0_16:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP0]], i64 16 +; CHECK-NEXT: [[GEP_0_30:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP0]], i64 30 +; CHECK-NEXT: [[GEP_0_48:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP0]], i64 48 +; CHECK-NEXT: [[GEP_ARR_16:%.*]] = getelementptr inbounds nuw i8, ptr [[ARR]], i64 16 +; CHECK-NEXT: [[GEP_ARR_30:%.*]] = getelementptr inbounds nuw i8, ptr [[ARR]], i64 30 +; CHECK-NEXT: [[GEP_ARR_48:%.*]] = getelementptr inbounds nuw i8, ptr [[ARR]], i64 48 +; CHECK-NEXT: [[LOAD_0_16:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_0_16]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[LOAD_0_16]], ptr nonnull [[GEP_ARR_16]], i32 1, <vscale x 4 x i1> [[MASK]]) +; CHECK-NEXT: [[LOAD_0_30:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_0_30]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[LOAD_0_30]], ptr nonnull [[GEP_ARR_30]], i32 1, <vscale x 4 x i1> [[MASK]]) +; CHECK-NEXT: [[LOAD_0_48:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_0_48]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[LOAD_0_48]], ptr nonnull [[GEP_ARR_48]], i32 1, <vscale x 4 x i1> [[MASK]]) +; CHECK-NEXT: [[FADDOP0:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_ARR_16]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: [[FADDOP1:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_ARR_48]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: [[FADD:%.*]] = fadd <vscale x 4 x float> [[FADDOP0]], [[FADDOP1]] +; CHECK-NEXT: ret <vscale x 4 x float> [[FADD]] ; %arr = alloca [64 x i32], align 4 %mask = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 0, i32 4) @@ -99,9 +140,23 @@ define <vscale x 4 x float> @scalable_store_partial_overwrite(ptr %0) { define <vscale x 4 x float> @dead_scalable_store_small_mask(ptr %0) { ; CHECK-LABEL: define <vscale x 4 x float> @dead_scalable_store_small_mask( -; CHECK: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> %load.0.16, ptr nonnull %gep.arr.16, i32 1, <vscale x 4 x i1> %mask) -; CHECK-NOT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> %load.0.30, ptr nonnull %gep.arr.30, i32 1, <vscale x 4 x i1> %mask) -; CHECK: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> %load.0.46, ptr nonnull %gep.arr.46, i32 1, <vscale x 4 x i1> %mask) +; CHECK-SAME: ptr [[TMP0:%.*]]) { +; CHECK-NEXT: [[ARR:%.*]] = alloca [64 x i32], align 4 +; CHECK-NEXT: [[MASK:%.*]] = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 0, i32 4) +; CHECK-NEXT: [[GEP_0_16:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP0]], i64 16 +; CHECK-NEXT: [[GEP_0_46:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP0]], i64 46 +; CHECK-NEXT: [[GEP_ARR_16:%.*]] = getelementptr inbounds nuw i8, ptr [[ARR]], i64 16 +; CHECK-NEXT: [[GEP_ARR_46:%.*]] = getelementptr inbounds nuw i8, ptr [[ARR]], i64 46 +; CHECK-NEXT: [[LOAD_0_16:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_0_16]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[LOAD_0_16]], ptr nonnull [[GEP_ARR_16]], i32 1, <vscale x 4 x i1> [[MASK]]) +; CHECK-NEXT: [[LOAD_0_46:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_0_46]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[LOAD_0_46]], ptr nonnull [[GEP_ARR_46]], i32 1, <vscale x 4 x i1> [[MASK]]) +; CHECK-NEXT: [[SMALLMASK:%.*]] = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 0, i32 2) +; CHECK-NEXT: [[FADDOP0:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_ARR_16]], i32 1, <vscale x 4 x i1> [[SMALLMASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: [[FADDOP1:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[GEP_ARR_46]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: [[FADD:%.*]] = fadd <vscale x 4 x float> [[FADDOP0]], [[FADDOP1]] +; CHECK-NEXT: ret <vscale x 4 x float> [[FADD]] +; %arr = alloca [64 x i32], align 4 %mask = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 0, i32 4) @@ -131,7 +186,12 @@ define <vscale x 4 x float> @dead_scalable_store_small_mask(ptr %0) { define <vscale x 4 x float> @dead_scalar_store(ptr noalias %0, ptr %1) { ; CHECK-LABEL: define <vscale x 4 x float> @dead_scalar_store( -; CHECK-NOT: store i32 20, ptr %gep.1.12 +; CHECK-SAME: ptr noalias [[TMP0:%.*]], ptr [[TMP1:%.*]]) { +; CHECK-NEXT: [[MASK:%.*]] = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i128(i128 0, i128 4) +; CHECK-NEXT: [[LOAD_0:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[TMP0]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[LOAD_0]], ptr nonnull [[TMP1]], i32 1, <vscale x 4 x i1> [[MASK]]) +; CHECK-NEXT: [[RETVAL:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[TMP1]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: ret <vscale x 4 x float> [[RETVAL]] ; %mask = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i128(i128 0, i128 4) %gep.1.12 = getelementptr inbounds nuw i8, ptr %1, i64 12 @@ -144,10 +204,17 @@ define <vscale x 4 x float> @dead_scalar_store(ptr noalias %0, ptr %1) { } -; CHECK-LABEL: define <4 x float> @dead_scalable_store_fixed_large_mask( -; CHECK-NOT: store i32 20, ptr %1 -; CHECK: store i32 50, ptr %gep.5 define <4 x float> @dead_scalable_store_fixed_large_mask(ptr noalias %0, ptr %1) { +; CHECK-LABEL: define <4 x float> @dead_scalable_store_fixed_large_mask( +; CHECK-SAME: ptr noalias [[TMP0:%.*]], ptr [[TMP1:%.*]]) { +; CHECK-NEXT: [[MASK:%.*]] = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 0, i32 7) +; CHECK-NEXT: [[GEP_5:%.*]] = getelementptr inbounds nuw i32, ptr [[TMP1]], i64 5 +; CHECK-NEXT: store i32 50, ptr [[GEP_5]], align 4 +; CHECK-NEXT: [[LOAD_0:%.*]] = call <4 x float> @llvm.masked.load.v4f32.p0(ptr nonnull [[TMP0]], i32 1, <4 x i1> [[MASK]], <4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.v4f32.p0(<4 x float> [[LOAD_0]], ptr nonnull [[TMP1]], i32 1, <4 x i1> [[MASK]]) +; CHECK-NEXT: [[RETVAL:%.*]] = call <4 x float> @llvm.masked.load.v4f32.p0(ptr nonnull [[TMP1]], i32 1, <4 x i1> [[MASK]], <4 x float> zeroinitializer) +; CHECK-NEXT: ret <4 x float> [[RETVAL]] +; %mask = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 0, i32 7) store i32 20, ptr %1 @@ -164,8 +231,16 @@ define <4 x float> @dead_scalable_store_fixed_large_mask(ptr noalias %0, ptr %1) ; This get active lane mask may cover 4 or 8 integers define <vscale x 4 x float> @mask_gt_minimum_num_elts(ptr noalias %0, ptr %1) { ; CHECK-LABEL: define <vscale x 4 x float> @mask_gt_minimum_num_elts( -; CHECK: store i32 10, ptr %gep.1.12 -; CHECK: store i32 20, ptr %gep.1.28 +; CHECK-SAME: ptr noalias [[TMP0:%.*]], ptr [[TMP1:%.*]]) { +; CHECK-NEXT: [[MASK:%.*]] = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 0, i32 8) +; CHECK-NEXT: [[GEP_1_12:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP1]], i64 12 +; CHECK-NEXT: store i32 10, ptr [[GEP_1_12]], align 4 +; CHECK-NEXT: [[GEP_1_28:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP1]], i64 28 +; CHECK-NEXT: store i32 20, ptr [[GEP_1_28]], align 4 +; CHECK-NEXT: [[LOAD_0:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[TMP0]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[LOAD_0]], ptr nonnull [[TMP1]], i32 1, <vscale x 4 x i1> [[MASK]]) +; CHECK-NEXT: [[RETVAL:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[TMP1]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: ret <vscale x 4 x float> [[RETVAL]] ; %mask = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 0, i32 8) %gep.1.12 = getelementptr inbounds nuw i8, ptr %1, i64 12 @@ -182,7 +257,13 @@ define <vscale x 4 x float> @mask_gt_minimum_num_elts(ptr noalias %0, ptr %1) { ; Don't do anything if the mask's Op1 < Op0 define <vscale x 4 x float> @active_lane_mask_lt(ptr noalias %0, ptr %1) { ; CHECK-LABEL: define <vscale x 4 x float> @active_lane_mask_lt( -; CHECK: store i32 20, ptr %1 +; CHECK-SAME: ptr noalias [[TMP0:%.*]], ptr [[TMP1:%.*]]) { +; CHECK-NEXT: [[MASK:%.*]] = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 4, i32 2) +; CHECK-NEXT: store i32 20, ptr [[TMP1]], align 4 +; CHECK-NEXT: [[LOAD_0:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[TMP0]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[LOAD_0]], ptr nonnull [[TMP1]], i32 1, <vscale x 4 x i1> [[MASK]]) +; CHECK-NEXT: [[RETVAL:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[TMP1]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: ret <vscale x 4 x float> [[RETVAL]] ; %mask = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 4, i32 2) store i32 20, ptr %1 @@ -196,7 +277,13 @@ define <vscale x 4 x float> @active_lane_mask_lt(ptr noalias %0, ptr %1) { ; Don't do anything if the mask's Op1 == Op0 define <vscale x 4 x float> @active_lane_mask_eq(ptr noalias %0, ptr %1) { ; CHECK-LABEL: define <vscale x 4 x float> @active_lane_mask_eq( -; CHECK: store i32 20, ptr %1 +; CHECK-SAME: ptr noalias [[TMP0:%.*]], ptr [[TMP1:%.*]]) { +; CHECK-NEXT: [[MASK:%.*]] = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 2, i32 2) +; CHECK-NEXT: store i32 20, ptr [[TMP1]], align 4 +; CHECK-NEXT: [[LOAD_0:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[TMP0]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[LOAD_0]], ptr nonnull [[TMP1]], i32 1, <vscale x 4 x i1> [[MASK]]) +; CHECK-NEXT: [[RETVAL:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[TMP1]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: ret <vscale x 4 x float> [[RETVAL]] ; %mask = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 2, i32 2) store i32 20, ptr %1 @@ -209,8 +296,14 @@ define <vscale x 4 x float> @active_lane_mask_eq(ptr noalias %0, ptr %1) { define <vscale x 16 x i8> @scalar_stores_small_mask(ptr noalias %0, ptr %1) { ; CHECK-LABEL: define <vscale x 16 x i8> @scalar_stores_small_mask( -; CHECK-NOT: store i8 60, ptr %gep.1.6 -; CHECK: store i8 120, ptr %gep.1.8 +; CHECK-SAME: ptr noalias [[TMP0:%.*]], ptr [[TMP1:%.*]]) { +; CHECK-NEXT: [[MASK:%.*]] = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i8(i8 0, i8 7) +; CHECK-NEXT: [[GEP_1_8:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP1]], i64 8 +; CHECK-NEXT: store i8 120, ptr [[GEP_1_8]], align 1 +; CHECK-NEXT: [[LOAD_0:%.*]] = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr nonnull [[TMP0]], i32 1, <vscale x 16 x i1> [[MASK]], <vscale x 16 x i8> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[LOAD_0]], ptr [[TMP1]], i32 1, <vscale x 16 x i1> [[MASK]]) +; CHECK-NEXT: [[RETVAL:%.*]] = call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[TMP1]], i32 1, <vscale x 16 x i1> [[MASK]], <vscale x 16 x i8> zeroinitializer) +; CHECK-NEXT: ret <vscale x 16 x i8> [[RETVAL]] ; %mask = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i8.i8(i8 0, i8 7) %gep.1.6 = getelementptr inbounds nuw i8, ptr %1, i64 6 @@ -226,10 +319,14 @@ define <vscale x 16 x i8> @scalar_stores_small_mask(ptr noalias %0, ptr %1) { define <vscale x 4 x float> @dead_scalar_store_offset(ptr noalias %0, ptr %1) { ; CHECK-LABEL: define <vscale x 4 x float> @dead_scalar_store_offset( -; CHECK-NOT: store i32 10, ptr %gep.1.0 -; CHECK-NOT: store i32 20, ptr %gep.1.4 -; CHECK-NOT: store i32 30, ptr %gep.1.8 -; CHECK: store i32 40, ptr %gep.1.12 +; CHECK-SAME: ptr noalias [[TMP0:%.*]], ptr [[TMP1:%.*]]) { +; CHECK-NEXT: [[MASK:%.*]] = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 1, i32 4) +; CHECK-NEXT: [[GEP_1_12:%.*]] = getelementptr inbounds nuw i8, ptr [[TMP1]], i64 12 +; CHECK-NEXT: store i32 40, ptr [[GEP_1_12]], align 4 +; CHECK-NEXT: [[LOAD_0:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[TMP0]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[LOAD_0]], ptr nonnull [[TMP1]], i32 1, <vscale x 4 x i1> [[MASK]]) +; CHECK-NEXT: [[RETVAL:%.*]] = call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr nonnull [[TMP1]], i32 1, <vscale x 4 x i1> [[MASK]], <vscale x 4 x float> zeroinitializer) +; CHECK-NEXT: ret <vscale x 4 x float> [[RETVAL]] ; %mask = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 1, i32 4) %gep.1.0 = getelementptr inbounds nuw i8, ptr %1, i64 0 diff --git a/llvm/test/Analysis/CostModel/AArch64/cast.ll b/llvm/test/Analysis/CostModel/AArch64/cast.ll index 38bd98f..15d67489 100644 --- a/llvm/test/Analysis/CostModel/AArch64/cast.ll +++ b/llvm/test/Analysis/CostModel/AArch64/cast.ll @@ -7,708 +7,708 @@ target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128" define void @ext() { ; CHECK-LABEL: 'ext' -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r0 = sext i1 undef to i8 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r1 = zext i1 undef to i8 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r2 = sext i1 undef to i16 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r3 = zext i1 undef to i16 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r4 = sext i1 undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r5 = zext i1 undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r6 = sext i1 undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r7 = zext i1 undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r9 = sext i8 undef to i16 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r10 = zext i8 undef to i16 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r11 = sext i8 undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r12 = zext i8 undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r13 = sext i8 undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r14 = zext i8 undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r17 = sext i16 undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r18 = zext i16 undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r19 = sext i16 undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r20 = zext i16 undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r24 = sext i32 undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r25 = zext i32 undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i8i16 = sext <2 x i8> undef to <2 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i8i16 = zext <2 x i8> undef to <2 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i8i32 = sext <2 x i8> undef to <2 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i8i32 = zext <2 x i8> undef to <2 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = sext <2 x i8> undef to <2 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i8i64 = zext <2 x i8> undef to <2 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i16i32 = sext <2 x i16> undef to <2 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i16i32 = zext <2 x i16> undef to <2 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = sext <2 x i16> undef to <2 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i16i64 = zext <2 x i16> undef to <2 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = sext <2 x i32> undef to <2 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> undef to <2 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> undef to <4 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> undef to <4 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i64 = sext <4 x i8> undef to <4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z4i8i64 = zext <4 x i8> undef to <4 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> undef to <4 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %z4i16i32 = zext <4 x i16> undef to <4 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i16i64 = sext <4 x i16> undef to <4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z4i16i64 = zext <4 x i16> undef to <4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i32i64 = sext <4 x i32> undef to <4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i32i64 = zext <4 x i32> undef to <4 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = sext <8 x i8> undef to <8 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %z8i8i16 = zext <8 x i8> undef to <8 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i32 = sext <8 x i8> undef to <8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i32 = zext <8 x i8> undef to <8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i64 = sext <8 x i8> undef to <8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i64 = zext <8 x i8> undef to <8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i32 = sext <8 x i16> undef to <8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i32 = zext <8 x i16> undef to <8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i64 = sext <8 x i16> undef to <8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i64 = zext <8 x i16> undef to <8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i32i64 = sext <8 x i32> undef to <8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z8i32i64 = zext <8 x i32> undef to <8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i16 = sext <16 x i8> undef to <16 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i16 = zext <16 x i8> undef to <16 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i32 = sext <16 x i8> undef to <16 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i32 = zext <16 x i8> undef to <16 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i64 = sext <16 x i8> undef to <16 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i64 = zext <16 x i8> undef to <16 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i32 = sext <16 x i16> undef to <16 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i32 = zext <16 x i16> undef to <16 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i64 = sext <16 x i16> undef to <16 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i64 = zext <16 x i16> undef to <16 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %s16i32i64 = sext <16 x i32> undef to <16 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %z16i32i64 = zext <16 x i32> undef to <16 x i64> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r0 = sext i1 poison to i8 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r1 = zext i1 poison to i8 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r2 = sext i1 poison to i16 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r3 = zext i1 poison to i16 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r4 = sext i1 poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r5 = zext i1 poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r6 = sext i1 poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r7 = zext i1 poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r9 = sext i8 poison to i16 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r10 = zext i8 poison to i16 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r11 = sext i8 poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r12 = zext i8 poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r13 = sext i8 poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r14 = zext i8 poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r17 = sext i16 poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r18 = zext i16 poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r19 = sext i16 poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r20 = zext i16 poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r24 = sext i32 poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r25 = zext i32 poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i8i16 = sext <2 x i8> poison to <2 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i8i16 = zext <2 x i8> poison to <2 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i8i32 = sext <2 x i8> poison to <2 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i8i32 = zext <2 x i8> poison to <2 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = sext <2 x i8> poison to <2 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i8i64 = zext <2 x i8> poison to <2 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i16i32 = sext <2 x i16> poison to <2 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i16i32 = zext <2 x i16> poison to <2 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = sext <2 x i16> poison to <2 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i16i64 = zext <2 x i16> poison to <2 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = sext <2 x i32> poison to <2 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> poison to <2 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> poison to <4 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> poison to <4 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> poison to <4 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> poison to <4 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i64 = sext <4 x i8> poison to <4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z4i8i64 = zext <4 x i8> poison to <4 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> poison to <4 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %z4i16i32 = zext <4 x i16> poison to <4 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i16i64 = sext <4 x i16> poison to <4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z4i16i64 = zext <4 x i16> poison to <4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i32i64 = sext <4 x i32> poison to <4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i32i64 = zext <4 x i32> poison to <4 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = sext <8 x i8> poison to <8 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %z8i8i16 = zext <8 x i8> poison to <8 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i32 = sext <8 x i8> poison to <8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i32 = zext <8 x i8> poison to <8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i64 = sext <8 x i8> poison to <8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i64 = zext <8 x i8> poison to <8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i32 = sext <8 x i16> poison to <8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i32 = zext <8 x i16> poison to <8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i64 = sext <8 x i16> poison to <8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i64 = zext <8 x i16> poison to <8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i32i64 = sext <8 x i32> poison to <8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z8i32i64 = zext <8 x i32> poison to <8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i16 = sext <16 x i8> poison to <16 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i16 = zext <16 x i8> poison to <16 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i32 = sext <16 x i8> poison to <16 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i32 = zext <16 x i8> poison to <16 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i64 = sext <16 x i8> poison to <16 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i64 = zext <16 x i8> poison to <16 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i32 = sext <16 x i16> poison to <16 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i32 = zext <16 x i16> poison to <16 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i64 = sext <16 x i16> poison to <16 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i64 = zext <16 x i16> poison to <16 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %s16i32i64 = sext <16 x i32> poison to <16 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %z16i32i64 = zext <16 x i32> poison to <16 x i64> ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %r0 = sext i1 undef to i8 - %r1 = zext i1 undef to i8 - %r2 = sext i1 undef to i16 - %r3 = zext i1 undef to i16 - %r4 = sext i1 undef to i32 - %r5 = zext i1 undef to i32 - %r6 = sext i1 undef to i64 - %r7 = zext i1 undef to i64 - %r9 = sext i8 undef to i16 - %r10 = zext i8 undef to i16 - %r11 = sext i8 undef to i32 - %r12 = zext i8 undef to i32 - %r13 = sext i8 undef to i64 - %r14 = zext i8 undef to i64 - %r17 = sext i16 undef to i32 - %r18 = zext i16 undef to i32 - %r19 = sext i16 undef to i64 - %r20 = zext i16 undef to i64 - %r24 = sext i32 undef to i64 - %r25 = zext i32 undef to i64 + %r0 = sext i1 poison to i8 + %r1 = zext i1 poison to i8 + %r2 = sext i1 poison to i16 + %r3 = zext i1 poison to i16 + %r4 = sext i1 poison to i32 + %r5 = zext i1 poison to i32 + %r6 = sext i1 poison to i64 + %r7 = zext i1 poison to i64 + %r9 = sext i8 poison to i16 + %r10 = zext i8 poison to i16 + %r11 = sext i8 poison to i32 + %r12 = zext i8 poison to i32 + %r13 = sext i8 poison to i64 + %r14 = zext i8 poison to i64 + %r17 = sext i16 poison to i32 + %r18 = zext i16 poison to i32 + %r19 = sext i16 poison to i64 + %r20 = zext i16 poison to i64 + %r24 = sext i32 poison to i64 + %r25 = zext i32 poison to i64 - %s2i8i16 = sext <2 x i8> undef to <2 x i16> - %z2i8i16 = zext <2 x i8> undef to <2 x i16> - %s2i8i32 = sext <2 x i8> undef to <2 x i32> - %z2i8i32 = zext <2 x i8> undef to <2 x i32> - %s2i8i64 = sext <2 x i8> undef to <2 x i64> - %z2i8i64 = zext <2 x i8> undef to <2 x i64> - %s2i16i32 = sext <2 x i16> undef to <2 x i32> - %z2i16i32 = zext <2 x i16> undef to <2 x i32> - %s2i16i64 = sext <2 x i16> undef to <2 x i64> - %z2i16i64 = zext <2 x i16> undef to <2 x i64> - %s2i32i64 = sext <2 x i32> undef to <2 x i64> - %z2i32i64 = zext <2 x i32> undef to <2 x i64> + %s2i8i16 = sext <2 x i8> poison to <2 x i16> + %z2i8i16 = zext <2 x i8> poison to <2 x i16> + %s2i8i32 = sext <2 x i8> poison to <2 x i32> + %z2i8i32 = zext <2 x i8> poison to <2 x i32> + %s2i8i64 = sext <2 x i8> poison to <2 x i64> + %z2i8i64 = zext <2 x i8> poison to <2 x i64> + %s2i16i32 = sext <2 x i16> poison to <2 x i32> + %z2i16i32 = zext <2 x i16> poison to <2 x i32> + %s2i16i64 = sext <2 x i16> poison to <2 x i64> + %z2i16i64 = zext <2 x i16> poison to <2 x i64> + %s2i32i64 = sext <2 x i32> poison to <2 x i64> + %z2i32i64 = zext <2 x i32> poison to <2 x i64> - %s4i8i16 = sext <4 x i8> undef to <4 x i16> - %z4i8i16 = zext <4 x i8> undef to <4 x i16> - %s4i8i32 = sext <4 x i8> undef to <4 x i32> - %z4i8i32 = zext <4 x i8> undef to <4 x i32> - %s4i8i64 = sext <4 x i8> undef to <4 x i64> - %z4i8i64 = zext <4 x i8> undef to <4 x i64> - %s4i16i32 = sext <4 x i16> undef to <4 x i32> - %z4i16i32 = zext <4 x i16> undef to <4 x i32> - %s4i16i64 = sext <4 x i16> undef to <4 x i64> - %z4i16i64 = zext <4 x i16> undef to <4 x i64> - %s4i32i64 = sext <4 x i32> undef to <4 x i64> - %z4i32i64 = zext <4 x i32> undef to <4 x i64> + %s4i8i16 = sext <4 x i8> poison to <4 x i16> + %z4i8i16 = zext <4 x i8> poison to <4 x i16> + %s4i8i32 = sext <4 x i8> poison to <4 x i32> + %z4i8i32 = zext <4 x i8> poison to <4 x i32> + %s4i8i64 = sext <4 x i8> poison to <4 x i64> + %z4i8i64 = zext <4 x i8> poison to <4 x i64> + %s4i16i32 = sext <4 x i16> poison to <4 x i32> + %z4i16i32 = zext <4 x i16> poison to <4 x i32> + %s4i16i64 = sext <4 x i16> poison to <4 x i64> + %z4i16i64 = zext <4 x i16> poison to <4 x i64> + %s4i32i64 = sext <4 x i32> poison to <4 x i64> + %z4i32i64 = zext <4 x i32> poison to <4 x i64> - %s8i8i16 = sext <8 x i8> undef to <8 x i16> - %z8i8i16 = zext <8 x i8> undef to <8 x i16> - %s8i8i32 = sext <8 x i8> undef to <8 x i32> - %z8i8i32 = zext <8 x i8> undef to <8 x i32> - %s8i8i64 = sext <8 x i8> undef to <8 x i64> - %z8i8i64 = zext <8 x i8> undef to <8 x i64> - %s8i16i32 = sext <8 x i16> undef to <8 x i32> - %z8i16i32 = zext <8 x i16> undef to <8 x i32> - %s8i16i64 = sext <8 x i16> undef to <8 x i64> - %z8i16i64 = zext <8 x i16> undef to <8 x i64> - %s8i32i64 = sext <8 x i32> undef to <8 x i64> - %z8i32i64 = zext <8 x i32> undef to <8 x i64> + %s8i8i16 = sext <8 x i8> poison to <8 x i16> + %z8i8i16 = zext <8 x i8> poison to <8 x i16> + %s8i8i32 = sext <8 x i8> poison to <8 x i32> + %z8i8i32 = zext <8 x i8> poison to <8 x i32> + %s8i8i64 = sext <8 x i8> poison to <8 x i64> + %z8i8i64 = zext <8 x i8> poison to <8 x i64> + %s8i16i32 = sext <8 x i16> poison to <8 x i32> + %z8i16i32 = zext <8 x i16> poison to <8 x i32> + %s8i16i64 = sext <8 x i16> poison to <8 x i64> + %z8i16i64 = zext <8 x i16> poison to <8 x i64> + %s8i32i64 = sext <8 x i32> poison to <8 x i64> + %z8i32i64 = zext <8 x i32> poison to <8 x i64> - %s16i8i16 = sext <16 x i8> undef to <16 x i16> - %z16i8i16 = zext <16 x i8> undef to <16 x i16> - %s16i8i32 = sext <16 x i8> undef to <16 x i32> - %z16i8i32 = zext <16 x i8> undef to <16 x i32> - %s16i8i64 = sext <16 x i8> undef to <16 x i64> - %z16i8i64 = zext <16 x i8> undef to <16 x i64> - %s16i16i32 = sext <16 x i16> undef to <16 x i32> - %z16i16i32 = zext <16 x i16> undef to <16 x i32> - %s16i16i64 = sext <16 x i16> undef to <16 x i64> - %z16i16i64 = zext <16 x i16> undef to <16 x i64> - %s16i32i64 = sext <16 x i32> undef to <16 x i64> - %z16i32i64 = zext <16 x i32> undef to <16 x i64> + %s16i8i16 = sext <16 x i8> poison to <16 x i16> + %z16i8i16 = zext <16 x i8> poison to <16 x i16> + %s16i8i32 = sext <16 x i8> poison to <16 x i32> + %z16i8i32 = zext <16 x i8> poison to <16 x i32> + %s16i8i64 = sext <16 x i8> poison to <16 x i64> + %z16i8i64 = zext <16 x i8> poison to <16 x i64> + %s16i16i32 = sext <16 x i16> poison to <16 x i32> + %z16i16i32 = zext <16 x i16> poison to <16 x i32> + %s16i16i64 = sext <16 x i16> poison to <16 x i64> + %z16i16i64 = zext <16 x i16> poison to <16 x i64> + %s16i32i64 = sext <16 x i32> poison to <16 x i64> + %z16i32i64 = zext <16 x i32> poison to <16 x i64> ret void } define void @trunc() { ; CHECK-LABEL: 'trunc' -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r8 = trunc i8 undef to i1 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r15 = trunc i16 undef to i1 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r16 = trunc i16 undef to i8 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r21 = trunc i32 undef to i1 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r22 = trunc i32 undef to i8 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r23 = trunc i32 undef to i16 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r26 = trunc i64 undef to i1 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r27 = trunc i64 undef to i8 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r28 = trunc i64 undef to i16 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r29 = trunc i64 undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %s2i8i16 = trunc <2 x i16> undef to <2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %s2i8i32 = trunc <2 x i32> undef to <2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = trunc <2 x i64> undef to <2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %s2i16i32 = trunc <2 x i32> undef to <2 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = trunc <2 x i64> undef to <2 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = trunc <2 x i64> undef to <2 x i32> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %s4i8i16 = trunc <4 x i16> undef to <4 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = trunc <4 x i32> undef to <4 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i64 = trunc <4 x i64> undef to <4 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = trunc <4 x i32> undef to <4 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i16i64 = trunc <4 x i64> undef to <4 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i32i64 = trunc <4 x i64> undef to <4 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = trunc <8 x i16> undef to <8 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i32 = trunc <8 x i32> undef to <8 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i64 = trunc <8 x i64> undef to <8 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s8i16i32 = trunc <8 x i32> undef to <8 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i64 = trunc <8 x i64> undef to <8 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i32i64 = trunc <8 x i64> undef to <8 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %s16i8i16 = trunc <16 x i16> undef to <16 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i32 = trunc <16 x i32> undef to <16 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i64 = trunc <16 x i64> undef to <16 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i32 = trunc <16 x i32> undef to <16 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i64 = trunc <16 x i64> undef to <16 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i32i64 = trunc <16 x i64> undef to <16 x i32> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r8 = trunc i8 poison to i1 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r15 = trunc i16 poison to i1 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r16 = trunc i16 poison to i8 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r21 = trunc i32 poison to i1 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r22 = trunc i32 poison to i8 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r23 = trunc i32 poison to i16 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r26 = trunc i64 poison to i1 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r27 = trunc i64 poison to i8 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r28 = trunc i64 poison to i16 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r29 = trunc i64 poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %s2i8i16 = trunc <2 x i16> poison to <2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %s2i8i32 = trunc <2 x i32> poison to <2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = trunc <2 x i64> poison to <2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %s2i16i32 = trunc <2 x i32> poison to <2 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = trunc <2 x i64> poison to <2 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = trunc <2 x i64> poison to <2 x i32> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %s4i8i16 = trunc <4 x i16> poison to <4 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = trunc <4 x i32> poison to <4 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i64 = trunc <4 x i64> poison to <4 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = trunc <4 x i32> poison to <4 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i16i64 = trunc <4 x i64> poison to <4 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s4i32i64 = trunc <4 x i64> poison to <4 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = trunc <8 x i16> poison to <8 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i32 = trunc <8 x i32> poison to <8 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i64 = trunc <8 x i64> poison to <8 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s8i16i32 = trunc <8 x i32> poison to <8 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i64 = trunc <8 x i64> poison to <8 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i32i64 = trunc <8 x i64> poison to <8 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %s16i8i16 = trunc <16 x i16> poison to <16 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i32 = trunc <16 x i32> poison to <16 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i64 = trunc <16 x i64> poison to <16 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i32 = trunc <16 x i32> poison to <16 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i64 = trunc <16 x i64> poison to <16 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i32i64 = trunc <16 x i64> poison to <16 x i32> ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %r8 = trunc i8 undef to i1 - %r15 = trunc i16 undef to i1 - %r16 = trunc i16 undef to i8 - %r21 = trunc i32 undef to i1 - %r22 = trunc i32 undef to i8 - %r23 = trunc i32 undef to i16 - %r26 = trunc i64 undef to i1 - %r27 = trunc i64 undef to i8 - %r28 = trunc i64 undef to i16 - %r29 = trunc i64 undef to i32 + %r8 = trunc i8 poison to i1 + %r15 = trunc i16 poison to i1 + %r16 = trunc i16 poison to i8 + %r21 = trunc i32 poison to i1 + %r22 = trunc i32 poison to i8 + %r23 = trunc i32 poison to i16 + %r26 = trunc i64 poison to i1 + %r27 = trunc i64 poison to i8 + %r28 = trunc i64 poison to i16 + %r29 = trunc i64 poison to i32 - %s2i8i16 = trunc <2 x i16> undef to <2 x i8> - %s2i8i32 = trunc <2 x i32> undef to <2 x i8> - %s2i8i64 = trunc <2 x i64> undef to <2 x i8> - %s2i16i32 = trunc <2 x i32> undef to <2 x i16> - %s2i16i64 = trunc <2 x i64> undef to <2 x i16> - %s2i32i64 = trunc <2 x i64> undef to <2 x i32> + %s2i8i16 = trunc <2 x i16> poison to <2 x i8> + %s2i8i32 = trunc <2 x i32> poison to <2 x i8> + %s2i8i64 = trunc <2 x i64> poison to <2 x i8> + %s2i16i32 = trunc <2 x i32> poison to <2 x i16> + %s2i16i64 = trunc <2 x i64> poison to <2 x i16> + %s2i32i64 = trunc <2 x i64> poison to <2 x i32> - %s4i8i16 = trunc <4 x i16> undef to <4 x i8> - %s4i8i32 = trunc <4 x i32> undef to <4 x i8> - %s4i8i64 = trunc <4 x i64> undef to <4 x i8> - %s4i16i32 = trunc <4 x i32> undef to <4 x i16> - %s4i16i64 = trunc <4 x i64> undef to <4 x i16> - %s4i32i64 = trunc <4 x i64> undef to <4 x i32> + %s4i8i16 = trunc <4 x i16> poison to <4 x i8> + %s4i8i32 = trunc <4 x i32> poison to <4 x i8> + %s4i8i64 = trunc <4 x i64> poison to <4 x i8> + %s4i16i32 = trunc <4 x i32> poison to <4 x i16> + %s4i16i64 = trunc <4 x i64> poison to <4 x i16> + %s4i32i64 = trunc <4 x i64> poison to <4 x i32> - %s8i8i16 = trunc <8 x i16> undef to <8 x i8> - %s8i8i32 = trunc <8 x i32> undef to <8 x i8> - %s8i8i64 = trunc <8 x i64> undef to <8 x i8> - %s8i16i32 = trunc <8 x i32> undef to <8 x i16> - %s8i16i64 = trunc <8 x i64> undef to <8 x i16> - %s8i32i64 = trunc <8 x i64> undef to <8 x i32> + %s8i8i16 = trunc <8 x i16> poison to <8 x i8> + %s8i8i32 = trunc <8 x i32> poison to <8 x i8> + %s8i8i64 = trunc <8 x i64> poison to <8 x i8> + %s8i16i32 = trunc <8 x i32> poison to <8 x i16> + %s8i16i64 = trunc <8 x i64> poison to <8 x i16> + %s8i32i64 = trunc <8 x i64> poison to <8 x i32> - %s16i8i16 = trunc <16 x i16> undef to <16 x i8> - %s16i8i32 = trunc <16 x i32> undef to <16 x i8> - %s16i8i64 = trunc <16 x i64> undef to <16 x i8> - %s16i16i32 = trunc <16 x i32> undef to <16 x i16> - %s16i16i64 = trunc <16 x i64> undef to <16 x i16> - %s16i32i64 = trunc <16 x i64> undef to <16 x i32> + %s16i8i16 = trunc <16 x i16> poison to <16 x i8> + %s16i8i32 = trunc <16 x i32> poison to <16 x i8> + %s16i8i64 = trunc <16 x i64> poison to <16 x i8> + %s16i16i32 = trunc <16 x i32> poison to <16 x i16> + %s16i16i64 = trunc <16 x i64> poison to <16 x i16> + %s16i32i64 = trunc <16 x i64> poison to <16 x i32> ret void } define i32 @casts_no_users() { ; CHECK-LABEL: 'casts_no_users' -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui float undef to i1 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi float undef to i1 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui float undef to i8 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi float undef to i8 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui float undef to i16 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi float undef to i16 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui float undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi float undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui float undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi float undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r40 = fptoui double undef to i1 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r41 = fptosi double undef to i1 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r42 = fptoui double undef to i8 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r43 = fptosi double undef to i8 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r44 = fptoui double undef to i16 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r45 = fptosi double undef to i16 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r46 = fptoui double undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r47 = fptosi double undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r48 = fptoui double undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r49 = fptosi double undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r50 = sitofp i1 undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r51 = uitofp i1 undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r52 = sitofp i1 undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r53 = uitofp i1 undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r54 = sitofp i8 undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r55 = uitofp i8 undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r56 = sitofp i8 undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r57 = uitofp i8 undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r58 = sitofp i16 undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r59 = uitofp i16 undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r60 = sitofp i16 undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r61 = uitofp i16 undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r62 = sitofp i32 undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r63 = uitofp i32 undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r64 = sitofp i32 undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r65 = uitofp i32 undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r66 = sitofp i64 undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r67 = uitofp i64 undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r68 = sitofp i64 undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r69 = uitofp i64 undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r80 = fptrunc double undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r81 = fptrunc <2 x double> undef to <2 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r82 = fptrunc <4 x double> undef to <4 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r83 = fptrunc <8 x double> undef to <8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r84 = fptrunc <16 x double> undef to <16 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %truncf64f16 = fptrunc double undef to half -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %truncv2f64f16 = fptrunc <2 x double> undef to <2 x half> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %truncv4f64f16 = fptrunc <4 x double> undef to <4 x half> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %truncv8f64f16 = fptrunc <8 x double> undef to <8 x half> -; CHECK-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %truncv16f64f16 = fptrunc <16 x double> undef to <16 x half> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %truncv32f16 = fptrunc float undef to half -; CHECK-NEXT: Cost Model: Found costs of 1 for: %truncv2f32f16 = fptrunc <2 x float> undef to <2 x half> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %truncv4f32f16 = fptrunc <4 x float> undef to <4 x half> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %truncv8f32f16 = fptrunc <8 x float> undef to <8 x half> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %truncv16f32f16 = fptrunc <16 x float> undef to <16 x half> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r85 = fpext float undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r86 = fpext <2 x float> undef to <2 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r87 = fpext <4 x float> undef to <4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r88 = fpext <8 x float> undef to <8 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r89 = fpext <16 x float> undef to <16 x double> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %extf16f32 = fpext half undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %extv2f16f32 = fpext <2 x half> undef to <2 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %extv4f16f32 = fpext <4 x half> undef to <4 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extv8f16f32 = fpext <8 x half> undef to <8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %extv16f16f32 = fpext <16 x half> undef to <16 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %extf16f64 = fpext half undef to double -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extv2f16f64 = fpext <2 x half> undef to <2 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %extv4f16f64 = fpext <4 x half> undef to <4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %extv8f16f64 = fpext <8 x half> undef to <8 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %extv16f16f64 = fpext <16 x half> undef to <16 x double> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x float> undef to <2 x i1> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x float> undef to <2 x i1> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x float> undef to <2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x float> undef to <2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x float> undef to <2 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x float> undef to <2 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x float> undef to <2 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x float> undef to <2 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x float> undef to <2 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x float> undef to <2 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r100 = fptoui <2 x double> undef to <2 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r101 = fptosi <2 x double> undef to <2 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r102 = fptoui <2 x double> undef to <2 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r103 = fptosi <2 x double> undef to <2 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r104 = fptoui <2 x double> undef to <2 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r105 = fptosi <2 x double> undef to <2 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r106 = fptoui <2 x double> undef to <2 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r107 = fptosi <2 x double> undef to <2 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r108 = fptoui <2 x double> undef to <2 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r109 = fptosi <2 x double> undef to <2 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r110 = fptoui <4 x float> undef to <4 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r111 = fptosi <4 x float> undef to <4 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r112 = fptoui <4 x float> undef to <4 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r113 = fptosi <4 x float> undef to <4 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r114 = fptoui <4 x float> undef to <4 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r115 = fptosi <4 x float> undef to <4 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r116 = fptoui <4 x float> undef to <4 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r117 = fptosi <4 x float> undef to <4 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x float> undef to <4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x float> undef to <4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r120 = fptoui <4 x double> undef to <4 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r121 = fptosi <4 x double> undef to <4 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r122 = fptoui <4 x double> undef to <4 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r123 = fptosi <4 x double> undef to <4 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r124 = fptoui <4 x double> undef to <4 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r125 = fptosi <4 x double> undef to <4 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r126 = fptoui <4 x double> undef to <4 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r127 = fptosi <4 x double> undef to <4 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r128 = fptoui <4 x double> undef to <4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r129 = fptosi <4 x double> undef to <4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:41 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x float> undef to <8 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:41 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x float> undef to <8 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x float> undef to <8 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x float> undef to <8 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r134 = fptoui <8 x float> undef to <8 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r135 = fptosi <8 x float> undef to <8 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x float> undef to <8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x float> undef to <8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x float> undef to <8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x float> undef to <8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r140 = fptoui <8 x double> undef to <8 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r141 = fptosi <8 x double> undef to <8 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r142 = fptoui <8 x double> undef to <8 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r143 = fptosi <8 x double> undef to <8 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r144 = fptoui <8 x double> undef to <8 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r145 = fptosi <8 x double> undef to <8 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r146 = fptoui <8 x double> undef to <8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r147 = fptosi <8 x double> undef to <8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r148 = fptoui <8 x double> undef to <8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r149 = fptosi <8 x double> undef to <8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:83 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x float> undef to <16 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:83 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x float> undef to <16 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x float> undef to <16 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x float> undef to <16 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x float> undef to <16 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x float> undef to <16 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x float> undef to <16 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x float> undef to <16 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x float> undef to <16 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x float> undef to <16 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:87 CodeSize:1 Lat:1 SizeLat:1 for: %r160 = fptoui <16 x double> undef to <16 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:87 CodeSize:1 Lat:1 SizeLat:1 for: %r161 = fptosi <16 x double> undef to <16 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:23 CodeSize:1 Lat:1 SizeLat:1 for: %r162 = fptoui <16 x double> undef to <16 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:23 CodeSize:1 Lat:1 SizeLat:1 for: %r163 = fptosi <16 x double> undef to <16 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:22 CodeSize:1 Lat:1 SizeLat:1 for: %r164 = fptoui <16 x double> undef to <16 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:22 CodeSize:1 Lat:1 SizeLat:1 for: %r165 = fptosi <16 x double> undef to <16 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r166 = fptoui <16 x double> undef to <16 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r167 = fptosi <16 x double> undef to <16 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r168 = fptoui <16 x double> undef to <16 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r169 = fptosi <16 x double> undef to <16 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r170 = uitofp <2 x i1> undef to <2 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r171 = sitofp <2 x i1> undef to <2 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r172 = uitofp <2 x i8> undef to <2 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r173 = sitofp <2 x i8> undef to <2 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r174 = uitofp <2 x i16> undef to <2 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r175 = sitofp <2 x i16> undef to <2 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r176 = uitofp <2 x i32> undef to <2 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r177 = sitofp <2 x i32> undef to <2 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r178 = uitofp <2 x i64> undef to <2 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r179 = sitofp <2 x i64> undef to <2 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r180 = uitofp <2 x i1> undef to <2 x double> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r181 = sitofp <2 x i1> undef to <2 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r182 = uitofp <2 x i8> undef to <2 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r183 = sitofp <2 x i8> undef to <2 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r184 = uitofp <2 x i16> undef to <2 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r185 = sitofp <2 x i16> undef to <2 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r186 = uitofp <2 x i32> undef to <2 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r187 = sitofp <2 x i32> undef to <2 x double> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r188 = uitofp <2 x i64> undef to <2 x double> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r189 = sitofp <2 x i64> undef to <2 x double> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r190 = uitofp <4 x i1> undef to <4 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r191 = sitofp <4 x i1> undef to <4 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r192 = uitofp <4 x i8> undef to <4 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r193 = sitofp <4 x i8> undef to <4 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r194 = uitofp <4 x i16> undef to <4 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r195 = sitofp <4 x i16> undef to <4 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r196 = uitofp <4 x i32> undef to <4 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %r197 = sitofp <4 x i32> undef to <4 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:18 CodeSize:1 Lat:1 SizeLat:1 for: %r198 = uitofp <4 x i64> undef to <4 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:18 CodeSize:1 Lat:1 SizeLat:1 for: %r199 = sitofp <4 x i64> undef to <4 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r200 = uitofp <4 x i1> undef to <4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r201 = sitofp <4 x i1> undef to <4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r202 = uitofp <4 x i8> undef to <4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r203 = sitofp <4 x i8> undef to <4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r204 = uitofp <4 x i16> undef to <4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r205 = sitofp <4 x i16> undef to <4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r206 = uitofp <4 x i32> undef to <4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r207 = sitofp <4 x i32> undef to <4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r208 = uitofp <4 x i64> undef to <4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r209 = sitofp <4 x i64> undef to <4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r210 = uitofp <8 x i1> undef to <8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r211 = sitofp <8 x i1> undef to <8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r212 = uitofp <8 x i8> undef to <8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r213 = sitofp <8 x i8> undef to <8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r214 = uitofp <8 x i16> undef to <8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r215 = sitofp <8 x i16> undef to <8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r216 = uitofp <8 x i32> undef to <8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r217 = sitofp <8 x i32> undef to <8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:36 CodeSize:1 Lat:1 SizeLat:1 for: %r218 = uitofp <8 x i64> undef to <8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:36 CodeSize:1 Lat:1 SizeLat:1 for: %r219 = sitofp <8 x i64> undef to <8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r220 = uitofp <8 x i1> undef to <8 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r221 = sitofp <8 x i1> undef to <8 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r222 = uitofp <8 x i8> undef to <8 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r223 = sitofp <8 x i8> undef to <8 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r224 = uitofp <8 x i16> undef to <8 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r225 = sitofp <8 x i16> undef to <8 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r226 = uitofp <8 x i16> undef to <8 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r227 = sitofp <8 x i16> undef to <8 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r228 = uitofp <8 x i64> undef to <8 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r229 = sitofp <8 x i64> undef to <8 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r230 = uitofp <16 x i1> undef to <16 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r231 = sitofp <16 x i1> undef to <16 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r232 = uitofp <16 x i8> undef to <16 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r233 = sitofp <16 x i8> undef to <16 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r234 = uitofp <16 x i16> undef to <16 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r235 = sitofp <16 x i16> undef to <16 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r236 = uitofp <16 x i32> undef to <16 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r237 = sitofp <16 x i32> undef to <16 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:72 CodeSize:1 Lat:1 SizeLat:1 for: %r238 = uitofp <16 x i64> undef to <16 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:72 CodeSize:1 Lat:1 SizeLat:1 for: %r239 = sitofp <16 x i64> undef to <16 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %r240 = uitofp <16 x i1> undef to <16 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %r241 = sitofp <16 x i1> undef to <16 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:39 CodeSize:1 Lat:1 SizeLat:1 for: %r242 = uitofp <16 x i8> undef to <16 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:39 CodeSize:1 Lat:1 SizeLat:1 for: %r243 = sitofp <16 x i8> undef to <16 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %r244 = uitofp <16 x i16> undef to <16 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %r245 = sitofp <16 x i16> undef to <16 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %r246 = uitofp <16 x i16> undef to <16 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %r247 = sitofp <16 x i16> undef to <16 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r248 = uitofp <16 x i64> undef to <16 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r249 = sitofp <16 x i64> undef to <16 x double> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui float poison to i1 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi float poison to i1 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui float poison to i8 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi float poison to i8 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui float poison to i16 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi float poison to i16 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui float poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi float poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui float poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi float poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r40 = fptoui double poison to i1 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r41 = fptosi double poison to i1 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r42 = fptoui double poison to i8 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r43 = fptosi double poison to i8 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r44 = fptoui double poison to i16 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r45 = fptosi double poison to i16 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r46 = fptoui double poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r47 = fptosi double poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r48 = fptoui double poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r49 = fptosi double poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r50 = sitofp i1 poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r51 = uitofp i1 poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r52 = sitofp i1 poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r53 = uitofp i1 poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r54 = sitofp i8 poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r55 = uitofp i8 poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r56 = sitofp i8 poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r57 = uitofp i8 poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r58 = sitofp i16 poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r59 = uitofp i16 poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r60 = sitofp i16 poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r61 = uitofp i16 poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r62 = sitofp i32 poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r63 = uitofp i32 poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r64 = sitofp i32 poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r65 = uitofp i32 poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r66 = sitofp i64 poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r67 = uitofp i64 poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r68 = sitofp i64 poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r69 = uitofp i64 poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r80 = fptrunc double poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r81 = fptrunc <2 x double> poison to <2 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r82 = fptrunc <4 x double> poison to <4 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r83 = fptrunc <8 x double> poison to <8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r84 = fptrunc <16 x double> poison to <16 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %truncf64f16 = fptrunc double poison to half +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %truncv2f64f16 = fptrunc <2 x double> poison to <2 x half> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %truncv4f64f16 = fptrunc <4 x double> poison to <4 x half> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %truncv8f64f16 = fptrunc <8 x double> poison to <8 x half> +; CHECK-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %truncv16f64f16 = fptrunc <16 x double> poison to <16 x half> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %truncv32f16 = fptrunc float poison to half +; CHECK-NEXT: Cost Model: Found costs of 1 for: %truncv2f32f16 = fptrunc <2 x float> poison to <2 x half> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %truncv4f32f16 = fptrunc <4 x float> poison to <4 x half> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %truncv8f32f16 = fptrunc <8 x float> poison to <8 x half> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %truncv16f32f16 = fptrunc <16 x float> poison to <16 x half> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r85 = fpext float poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r86 = fpext <2 x float> poison to <2 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r87 = fpext <4 x float> poison to <4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r88 = fpext <8 x float> poison to <8 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r89 = fpext <16 x float> poison to <16 x double> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %extf16f32 = fpext half poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %extv2f16f32 = fpext <2 x half> poison to <2 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %extv4f16f32 = fpext <4 x half> poison to <4 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extv8f16f32 = fpext <8 x half> poison to <8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %extv16f16f32 = fpext <16 x half> poison to <16 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %extf16f64 = fpext half poison to double +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extv2f16f64 = fpext <2 x half> poison to <2 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %extv4f16f64 = fpext <4 x half> poison to <4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %extv8f16f64 = fpext <8 x half> poison to <8 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %extv16f16f64 = fpext <16 x half> poison to <16 x double> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x float> poison to <2 x i1> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x float> poison to <2 x i1> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x float> poison to <2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x float> poison to <2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x float> poison to <2 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x float> poison to <2 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x float> poison to <2 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x float> poison to <2 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x float> poison to <2 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x float> poison to <2 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r100 = fptoui <2 x double> poison to <2 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r101 = fptosi <2 x double> poison to <2 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r102 = fptoui <2 x double> poison to <2 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r103 = fptosi <2 x double> poison to <2 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r104 = fptoui <2 x double> poison to <2 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r105 = fptosi <2 x double> poison to <2 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r106 = fptoui <2 x double> poison to <2 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r107 = fptosi <2 x double> poison to <2 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r108 = fptoui <2 x double> poison to <2 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r109 = fptosi <2 x double> poison to <2 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r110 = fptoui <4 x float> poison to <4 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r111 = fptosi <4 x float> poison to <4 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r112 = fptoui <4 x float> poison to <4 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r113 = fptosi <4 x float> poison to <4 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r114 = fptoui <4 x float> poison to <4 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r115 = fptosi <4 x float> poison to <4 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r116 = fptoui <4 x float> poison to <4 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r117 = fptosi <4 x float> poison to <4 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x float> poison to <4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x float> poison to <4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r120 = fptoui <4 x double> poison to <4 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r121 = fptosi <4 x double> poison to <4 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r122 = fptoui <4 x double> poison to <4 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r123 = fptosi <4 x double> poison to <4 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r124 = fptoui <4 x double> poison to <4 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r125 = fptosi <4 x double> poison to <4 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r126 = fptoui <4 x double> poison to <4 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r127 = fptosi <4 x double> poison to <4 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r128 = fptoui <4 x double> poison to <4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r129 = fptosi <4 x double> poison to <4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:41 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x float> poison to <8 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:41 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x float> poison to <8 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x float> poison to <8 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x float> poison to <8 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r134 = fptoui <8 x float> poison to <8 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r135 = fptosi <8 x float> poison to <8 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x float> poison to <8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x float> poison to <8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x float> poison to <8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x float> poison to <8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r140 = fptoui <8 x double> poison to <8 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r141 = fptosi <8 x double> poison to <8 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r142 = fptoui <8 x double> poison to <8 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r143 = fptosi <8 x double> poison to <8 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r144 = fptoui <8 x double> poison to <8 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r145 = fptosi <8 x double> poison to <8 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r146 = fptoui <8 x double> poison to <8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r147 = fptosi <8 x double> poison to <8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r148 = fptoui <8 x double> poison to <8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r149 = fptosi <8 x double> poison to <8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:83 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x float> poison to <16 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:83 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x float> poison to <16 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x float> poison to <16 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x float> poison to <16 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x float> poison to <16 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x float> poison to <16 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x float> poison to <16 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x float> poison to <16 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x float> poison to <16 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x float> poison to <16 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:87 CodeSize:1 Lat:1 SizeLat:1 for: %r160 = fptoui <16 x double> poison to <16 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:87 CodeSize:1 Lat:1 SizeLat:1 for: %r161 = fptosi <16 x double> poison to <16 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:23 CodeSize:1 Lat:1 SizeLat:1 for: %r162 = fptoui <16 x double> poison to <16 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:23 CodeSize:1 Lat:1 SizeLat:1 for: %r163 = fptosi <16 x double> poison to <16 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:22 CodeSize:1 Lat:1 SizeLat:1 for: %r164 = fptoui <16 x double> poison to <16 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:22 CodeSize:1 Lat:1 SizeLat:1 for: %r165 = fptosi <16 x double> poison to <16 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r166 = fptoui <16 x double> poison to <16 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r167 = fptosi <16 x double> poison to <16 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r168 = fptoui <16 x double> poison to <16 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r169 = fptosi <16 x double> poison to <16 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r170 = uitofp <2 x i1> poison to <2 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r171 = sitofp <2 x i1> poison to <2 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r172 = uitofp <2 x i8> poison to <2 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r173 = sitofp <2 x i8> poison to <2 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r174 = uitofp <2 x i16> poison to <2 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r175 = sitofp <2 x i16> poison to <2 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r176 = uitofp <2 x i32> poison to <2 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r177 = sitofp <2 x i32> poison to <2 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r178 = uitofp <2 x i64> poison to <2 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r179 = sitofp <2 x i64> poison to <2 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r180 = uitofp <2 x i1> poison to <2 x double> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r181 = sitofp <2 x i1> poison to <2 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r182 = uitofp <2 x i8> poison to <2 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r183 = sitofp <2 x i8> poison to <2 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r184 = uitofp <2 x i16> poison to <2 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r185 = sitofp <2 x i16> poison to <2 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r186 = uitofp <2 x i32> poison to <2 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r187 = sitofp <2 x i32> poison to <2 x double> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r188 = uitofp <2 x i64> poison to <2 x double> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r189 = sitofp <2 x i64> poison to <2 x double> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r190 = uitofp <4 x i1> poison to <4 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r191 = sitofp <4 x i1> poison to <4 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r192 = uitofp <4 x i8> poison to <4 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r193 = sitofp <4 x i8> poison to <4 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r194 = uitofp <4 x i16> poison to <4 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r195 = sitofp <4 x i16> poison to <4 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r196 = uitofp <4 x i32> poison to <4 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %r197 = sitofp <4 x i32> poison to <4 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:18 CodeSize:1 Lat:1 SizeLat:1 for: %r198 = uitofp <4 x i64> poison to <4 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:18 CodeSize:1 Lat:1 SizeLat:1 for: %r199 = sitofp <4 x i64> poison to <4 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r200 = uitofp <4 x i1> poison to <4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r201 = sitofp <4 x i1> poison to <4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r202 = uitofp <4 x i8> poison to <4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r203 = sitofp <4 x i8> poison to <4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r204 = uitofp <4 x i16> poison to <4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r205 = sitofp <4 x i16> poison to <4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r206 = uitofp <4 x i32> poison to <4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r207 = sitofp <4 x i32> poison to <4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r208 = uitofp <4 x i64> poison to <4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r209 = sitofp <4 x i64> poison to <4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r210 = uitofp <8 x i1> poison to <8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r211 = sitofp <8 x i1> poison to <8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r212 = uitofp <8 x i8> poison to <8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r213 = sitofp <8 x i8> poison to <8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r214 = uitofp <8 x i16> poison to <8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r215 = sitofp <8 x i16> poison to <8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r216 = uitofp <8 x i32> poison to <8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r217 = sitofp <8 x i32> poison to <8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:36 CodeSize:1 Lat:1 SizeLat:1 for: %r218 = uitofp <8 x i64> poison to <8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:36 CodeSize:1 Lat:1 SizeLat:1 for: %r219 = sitofp <8 x i64> poison to <8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r220 = uitofp <8 x i1> poison to <8 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r221 = sitofp <8 x i1> poison to <8 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r222 = uitofp <8 x i8> poison to <8 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r223 = sitofp <8 x i8> poison to <8 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r224 = uitofp <8 x i16> poison to <8 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r225 = sitofp <8 x i16> poison to <8 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r226 = uitofp <8 x i16> poison to <8 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r227 = sitofp <8 x i16> poison to <8 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r228 = uitofp <8 x i64> poison to <8 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r229 = sitofp <8 x i64> poison to <8 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r230 = uitofp <16 x i1> poison to <16 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r231 = sitofp <16 x i1> poison to <16 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r232 = uitofp <16 x i8> poison to <16 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r233 = sitofp <16 x i8> poison to <16 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r234 = uitofp <16 x i16> poison to <16 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r235 = sitofp <16 x i16> poison to <16 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r236 = uitofp <16 x i32> poison to <16 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r237 = sitofp <16 x i32> poison to <16 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:72 CodeSize:1 Lat:1 SizeLat:1 for: %r238 = uitofp <16 x i64> poison to <16 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:72 CodeSize:1 Lat:1 SizeLat:1 for: %r239 = sitofp <16 x i64> poison to <16 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %r240 = uitofp <16 x i1> poison to <16 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %r241 = sitofp <16 x i1> poison to <16 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:39 CodeSize:1 Lat:1 SizeLat:1 for: %r242 = uitofp <16 x i8> poison to <16 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:39 CodeSize:1 Lat:1 SizeLat:1 for: %r243 = sitofp <16 x i8> poison to <16 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %r244 = uitofp <16 x i16> poison to <16 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %r245 = sitofp <16 x i16> poison to <16 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %r246 = uitofp <16 x i16> poison to <16 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %r247 = sitofp <16 x i16> poison to <16 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r248 = uitofp <16 x i64> poison to <16 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r249 = sitofp <16 x i64> poison to <16 x double> ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef ; - %r30 = fptoui float undef to i1 - %r31 = fptosi float undef to i1 - %r32 = fptoui float undef to i8 - %r33 = fptosi float undef to i8 - %r34 = fptoui float undef to i16 - %r35 = fptosi float undef to i16 - %r36 = fptoui float undef to i32 - %r37 = fptosi float undef to i32 - %r38 = fptoui float undef to i64 - %r39 = fptosi float undef to i64 - %r40 = fptoui double undef to i1 - %r41 = fptosi double undef to i1 - %r42 = fptoui double undef to i8 - %r43 = fptosi double undef to i8 - %r44 = fptoui double undef to i16 - %r45 = fptosi double undef to i16 - %r46 = fptoui double undef to i32 - %r47 = fptosi double undef to i32 - %r48 = fptoui double undef to i64 - %r49 = fptosi double undef to i64 - %r50 = sitofp i1 undef to float - %r51 = uitofp i1 undef to float - %r52 = sitofp i1 undef to double - %r53 = uitofp i1 undef to double - %r54 = sitofp i8 undef to float - %r55 = uitofp i8 undef to float - %r56 = sitofp i8 undef to double - %r57 = uitofp i8 undef to double - %r58 = sitofp i16 undef to float - %r59 = uitofp i16 undef to float - %r60 = sitofp i16 undef to double - %r61 = uitofp i16 undef to double - %r62 = sitofp i32 undef to float - %r63 = uitofp i32 undef to float - %r64 = sitofp i32 undef to double - %r65 = uitofp i32 undef to double - %r66 = sitofp i64 undef to float - %r67 = uitofp i64 undef to float - %r68 = sitofp i64 undef to double - %r69 = uitofp i64 undef to double - %r80 = fptrunc double undef to float - %r81 = fptrunc <2 x double> undef to <2 x float> - %r82 = fptrunc <4 x double> undef to <4 x float> - %r83 = fptrunc <8 x double> undef to <8 x float> - %r84 = fptrunc <16 x double> undef to <16 x float> - %truncf64f16 = fptrunc double undef to half - %truncv2f64f16 = fptrunc <2 x double> undef to <2 x half> - %truncv4f64f16 = fptrunc <4 x double> undef to <4 x half> - %truncv8f64f16 = fptrunc <8 x double> undef to <8 x half> - %truncv16f64f16 = fptrunc <16 x double> undef to <16 x half> - %truncv32f16 = fptrunc float undef to half - %truncv2f32f16 = fptrunc <2 x float> undef to <2 x half> - %truncv4f32f16 = fptrunc <4 x float> undef to <4 x half> - %truncv8f32f16 = fptrunc <8 x float> undef to <8 x half> - %truncv16f32f16 = fptrunc <16 x float> undef to <16 x half> - %r85 = fpext float undef to double - %r86 = fpext <2 x float> undef to <2 x double> - %r87 = fpext <4 x float> undef to <4 x double> - %r88 = fpext <8 x float> undef to <8 x double> - %r89 = fpext <16 x float> undef to <16 x double> - %extf16f32 = fpext half undef to float - %extv2f16f32 = fpext <2 x half> undef to <2 x float> - %extv4f16f32 = fpext <4 x half> undef to <4 x float> - %extv8f16f32 = fpext <8 x half> undef to <8 x float> - %extv16f16f32 = fpext <16 x half> undef to <16 x float> - %extf16f64 = fpext half undef to double - %extv2f16f64 = fpext <2 x half> undef to <2 x double> - %extv4f16f64 = fpext <4 x half> undef to <4 x double> - %extv8f16f64 = fpext <8 x half> undef to <8 x double> - %extv16f16f64 = fpext <16 x half> undef to <16 x double> - %r90 = fptoui <2 x float> undef to <2 x i1> - %r91 = fptosi <2 x float> undef to <2 x i1> - %r92 = fptoui <2 x float> undef to <2 x i8> - %r93 = fptosi <2 x float> undef to <2 x i8> - %r94 = fptoui <2 x float> undef to <2 x i16> - %r95 = fptosi <2 x float> undef to <2 x i16> - %r96 = fptoui <2 x float> undef to <2 x i32> - %r97 = fptosi <2 x float> undef to <2 x i32> - %r98 = fptoui <2 x float> undef to <2 x i64> - %r99 = fptosi <2 x float> undef to <2 x i64> - %r100 = fptoui <2 x double> undef to <2 x i1> - %r101 = fptosi <2 x double> undef to <2 x i1> - %r102 = fptoui <2 x double> undef to <2 x i8> - %r103 = fptosi <2 x double> undef to <2 x i8> - %r104 = fptoui <2 x double> undef to <2 x i16> - %r105 = fptosi <2 x double> undef to <2 x i16> - %r106 = fptoui <2 x double> undef to <2 x i32> - %r107 = fptosi <2 x double> undef to <2 x i32> - %r108 = fptoui <2 x double> undef to <2 x i64> - %r109 = fptosi <2 x double> undef to <2 x i64> + %r30 = fptoui float poison to i1 + %r31 = fptosi float poison to i1 + %r32 = fptoui float poison to i8 + %r33 = fptosi float poison to i8 + %r34 = fptoui float poison to i16 + %r35 = fptosi float poison to i16 + %r36 = fptoui float poison to i32 + %r37 = fptosi float poison to i32 + %r38 = fptoui float poison to i64 + %r39 = fptosi float poison to i64 + %r40 = fptoui double poison to i1 + %r41 = fptosi double poison to i1 + %r42 = fptoui double poison to i8 + %r43 = fptosi double poison to i8 + %r44 = fptoui double poison to i16 + %r45 = fptosi double poison to i16 + %r46 = fptoui double poison to i32 + %r47 = fptosi double poison to i32 + %r48 = fptoui double poison to i64 + %r49 = fptosi double poison to i64 + %r50 = sitofp i1 poison to float + %r51 = uitofp i1 poison to float + %r52 = sitofp i1 poison to double + %r53 = uitofp i1 poison to double + %r54 = sitofp i8 poison to float + %r55 = uitofp i8 poison to float + %r56 = sitofp i8 poison to double + %r57 = uitofp i8 poison to double + %r58 = sitofp i16 poison to float + %r59 = uitofp i16 poison to float + %r60 = sitofp i16 poison to double + %r61 = uitofp i16 poison to double + %r62 = sitofp i32 poison to float + %r63 = uitofp i32 poison to float + %r64 = sitofp i32 poison to double + %r65 = uitofp i32 poison to double + %r66 = sitofp i64 poison to float + %r67 = uitofp i64 poison to float + %r68 = sitofp i64 poison to double + %r69 = uitofp i64 poison to double + %r80 = fptrunc double poison to float + %r81 = fptrunc <2 x double> poison to <2 x float> + %r82 = fptrunc <4 x double> poison to <4 x float> + %r83 = fptrunc <8 x double> poison to <8 x float> + %r84 = fptrunc <16 x double> poison to <16 x float> + %truncf64f16 = fptrunc double poison to half + %truncv2f64f16 = fptrunc <2 x double> poison to <2 x half> + %truncv4f64f16 = fptrunc <4 x double> poison to <4 x half> + %truncv8f64f16 = fptrunc <8 x double> poison to <8 x half> + %truncv16f64f16 = fptrunc <16 x double> poison to <16 x half> + %truncv32f16 = fptrunc float poison to half + %truncv2f32f16 = fptrunc <2 x float> poison to <2 x half> + %truncv4f32f16 = fptrunc <4 x float> poison to <4 x half> + %truncv8f32f16 = fptrunc <8 x float> poison to <8 x half> + %truncv16f32f16 = fptrunc <16 x float> poison to <16 x half> + %r85 = fpext float poison to double + %r86 = fpext <2 x float> poison to <2 x double> + %r87 = fpext <4 x float> poison to <4 x double> + %r88 = fpext <8 x float> poison to <8 x double> + %r89 = fpext <16 x float> poison to <16 x double> + %extf16f32 = fpext half poison to float + %extv2f16f32 = fpext <2 x half> poison to <2 x float> + %extv4f16f32 = fpext <4 x half> poison to <4 x float> + %extv8f16f32 = fpext <8 x half> poison to <8 x float> + %extv16f16f32 = fpext <16 x half> poison to <16 x float> + %extf16f64 = fpext half poison to double + %extv2f16f64 = fpext <2 x half> poison to <2 x double> + %extv4f16f64 = fpext <4 x half> poison to <4 x double> + %extv8f16f64 = fpext <8 x half> poison to <8 x double> + %extv16f16f64 = fpext <16 x half> poison to <16 x double> + %r90 = fptoui <2 x float> poison to <2 x i1> + %r91 = fptosi <2 x float> poison to <2 x i1> + %r92 = fptoui <2 x float> poison to <2 x i8> + %r93 = fptosi <2 x float> poison to <2 x i8> + %r94 = fptoui <2 x float> poison to <2 x i16> + %r95 = fptosi <2 x float> poison to <2 x i16> + %r96 = fptoui <2 x float> poison to <2 x i32> + %r97 = fptosi <2 x float> poison to <2 x i32> + %r98 = fptoui <2 x float> poison to <2 x i64> + %r99 = fptosi <2 x float> poison to <2 x i64> + %r100 = fptoui <2 x double> poison to <2 x i1> + %r101 = fptosi <2 x double> poison to <2 x i1> + %r102 = fptoui <2 x double> poison to <2 x i8> + %r103 = fptosi <2 x double> poison to <2 x i8> + %r104 = fptoui <2 x double> poison to <2 x i16> + %r105 = fptosi <2 x double> poison to <2 x i16> + %r106 = fptoui <2 x double> poison to <2 x i32> + %r107 = fptosi <2 x double> poison to <2 x i32> + %r108 = fptoui <2 x double> poison to <2 x i64> + %r109 = fptosi <2 x double> poison to <2 x i64> - %r110 = fptoui <4 x float> undef to <4 x i1> - %r111 = fptosi <4 x float> undef to <4 x i1> - %r112 = fptoui <4 x float> undef to <4 x i8> - %r113 = fptosi <4 x float> undef to <4 x i8> - %r114 = fptoui <4 x float> undef to <4 x i16> - %r115 = fptosi <4 x float> undef to <4 x i16> - %r116 = fptoui <4 x float> undef to <4 x i32> - %r117 = fptosi <4 x float> undef to <4 x i32> - %r118 = fptoui <4 x float> undef to <4 x i64> - %r119 = fptosi <4 x float> undef to <4 x i64> + %r110 = fptoui <4 x float> poison to <4 x i1> + %r111 = fptosi <4 x float> poison to <4 x i1> + %r112 = fptoui <4 x float> poison to <4 x i8> + %r113 = fptosi <4 x float> poison to <4 x i8> + %r114 = fptoui <4 x float> poison to <4 x i16> + %r115 = fptosi <4 x float> poison to <4 x i16> + %r116 = fptoui <4 x float> poison to <4 x i32> + %r117 = fptosi <4 x float> poison to <4 x i32> + %r118 = fptoui <4 x float> poison to <4 x i64> + %r119 = fptosi <4 x float> poison to <4 x i64> - %r120 = fptoui <4 x double> undef to <4 x i1> - %r121 = fptosi <4 x double> undef to <4 x i1> - %r122 = fptoui <4 x double> undef to <4 x i8> - %r123 = fptosi <4 x double> undef to <4 x i8> - %r124 = fptoui <4 x double> undef to <4 x i16> - %r125 = fptosi <4 x double> undef to <4 x i16> - %r126 = fptoui <4 x double> undef to <4 x i32> - %r127 = fptosi <4 x double> undef to <4 x i32> - %r128 = fptoui <4 x double> undef to <4 x i64> - %r129 = fptosi <4 x double> undef to <4 x i64> + %r120 = fptoui <4 x double> poison to <4 x i1> + %r121 = fptosi <4 x double> poison to <4 x i1> + %r122 = fptoui <4 x double> poison to <4 x i8> + %r123 = fptosi <4 x double> poison to <4 x i8> + %r124 = fptoui <4 x double> poison to <4 x i16> + %r125 = fptosi <4 x double> poison to <4 x i16> + %r126 = fptoui <4 x double> poison to <4 x i32> + %r127 = fptosi <4 x double> poison to <4 x i32> + %r128 = fptoui <4 x double> poison to <4 x i64> + %r129 = fptosi <4 x double> poison to <4 x i64> - %r130 = fptoui <8 x float> undef to <8 x i1> - %r131 = fptosi <8 x float> undef to <8 x i1> - %r132 = fptoui <8 x float> undef to <8 x i8> - %r133 = fptosi <8 x float> undef to <8 x i8> - %r134 = fptoui <8 x float> undef to <8 x i16> - %r135 = fptosi <8 x float> undef to <8 x i16> - %r136 = fptoui <8 x float> undef to <8 x i32> - %r137 = fptosi <8 x float> undef to <8 x i32> - %r138 = fptoui <8 x float> undef to <8 x i64> - %r139 = fptosi <8 x float> undef to <8 x i64> + %r130 = fptoui <8 x float> poison to <8 x i1> + %r131 = fptosi <8 x float> poison to <8 x i1> + %r132 = fptoui <8 x float> poison to <8 x i8> + %r133 = fptosi <8 x float> poison to <8 x i8> + %r134 = fptoui <8 x float> poison to <8 x i16> + %r135 = fptosi <8 x float> poison to <8 x i16> + %r136 = fptoui <8 x float> poison to <8 x i32> + %r137 = fptosi <8 x float> poison to <8 x i32> + %r138 = fptoui <8 x float> poison to <8 x i64> + %r139 = fptosi <8 x float> poison to <8 x i64> - %r140 = fptoui <8 x double> undef to <8 x i1> - %r141 = fptosi <8 x double> undef to <8 x i1> - %r142 = fptoui <8 x double> undef to <8 x i8> - %r143 = fptosi <8 x double> undef to <8 x i8> - %r144 = fptoui <8 x double> undef to <8 x i16> - %r145 = fptosi <8 x double> undef to <8 x i16> - %r146 = fptoui <8 x double> undef to <8 x i32> - %r147 = fptosi <8 x double> undef to <8 x i32> - %r148 = fptoui <8 x double> undef to <8 x i64> - %r149 = fptosi <8 x double> undef to <8 x i64> + %r140 = fptoui <8 x double> poison to <8 x i1> + %r141 = fptosi <8 x double> poison to <8 x i1> + %r142 = fptoui <8 x double> poison to <8 x i8> + %r143 = fptosi <8 x double> poison to <8 x i8> + %r144 = fptoui <8 x double> poison to <8 x i16> + %r145 = fptosi <8 x double> poison to <8 x i16> + %r146 = fptoui <8 x double> poison to <8 x i32> + %r147 = fptosi <8 x double> poison to <8 x i32> + %r148 = fptoui <8 x double> poison to <8 x i64> + %r149 = fptosi <8 x double> poison to <8 x i64> - %r150 = fptoui <16 x float> undef to <16 x i1> - %r151 = fptosi <16 x float> undef to <16 x i1> - %r152 = fptoui <16 x float> undef to <16 x i8> - %r153 = fptosi <16 x float> undef to <16 x i8> - %r154 = fptoui <16 x float> undef to <16 x i16> - %r155 = fptosi <16 x float> undef to <16 x i16> - %r156 = fptoui <16 x float> undef to <16 x i32> - %r157 = fptosi <16 x float> undef to <16 x i32> - %r158 = fptoui <16 x float> undef to <16 x i64> - %r159 = fptosi <16 x float> undef to <16 x i64> + %r150 = fptoui <16 x float> poison to <16 x i1> + %r151 = fptosi <16 x float> poison to <16 x i1> + %r152 = fptoui <16 x float> poison to <16 x i8> + %r153 = fptosi <16 x float> poison to <16 x i8> + %r154 = fptoui <16 x float> poison to <16 x i16> + %r155 = fptosi <16 x float> poison to <16 x i16> + %r156 = fptoui <16 x float> poison to <16 x i32> + %r157 = fptosi <16 x float> poison to <16 x i32> + %r158 = fptoui <16 x float> poison to <16 x i64> + %r159 = fptosi <16 x float> poison to <16 x i64> - %r160 = fptoui <16 x double> undef to <16 x i1> - %r161 = fptosi <16 x double> undef to <16 x i1> - %r162 = fptoui <16 x double> undef to <16 x i8> - %r163 = fptosi <16 x double> undef to <16 x i8> - %r164 = fptoui <16 x double> undef to <16 x i16> - %r165 = fptosi <16 x double> undef to <16 x i16> - %r166 = fptoui <16 x double> undef to <16 x i32> - %r167 = fptosi <16 x double> undef to <16 x i32> - %r168 = fptoui <16 x double> undef to <16 x i64> - %r169 = fptosi <16 x double> undef to <16 x i64> + %r160 = fptoui <16 x double> poison to <16 x i1> + %r161 = fptosi <16 x double> poison to <16 x i1> + %r162 = fptoui <16 x double> poison to <16 x i8> + %r163 = fptosi <16 x double> poison to <16 x i8> + %r164 = fptoui <16 x double> poison to <16 x i16> + %r165 = fptosi <16 x double> poison to <16 x i16> + %r166 = fptoui <16 x double> poison to <16 x i32> + %r167 = fptosi <16 x double> poison to <16 x i32> + %r168 = fptoui <16 x double> poison to <16 x i64> + %r169 = fptosi <16 x double> poison to <16 x i64> - %r170 = uitofp <2 x i1> undef to <2 x float> - %r171 = sitofp <2 x i1> undef to <2 x float> - %r172 = uitofp <2 x i8> undef to <2 x float> - %r173 = sitofp <2 x i8> undef to <2 x float> - %r174 = uitofp <2 x i16> undef to <2 x float> - %r175 = sitofp <2 x i16> undef to <2 x float> - %r176 = uitofp <2 x i32> undef to <2 x float> - %r177 = sitofp <2 x i32> undef to <2 x float> - %r178 = uitofp <2 x i64> undef to <2 x float> - %r179 = sitofp <2 x i64> undef to <2 x float> + %r170 = uitofp <2 x i1> poison to <2 x float> + %r171 = sitofp <2 x i1> poison to <2 x float> + %r172 = uitofp <2 x i8> poison to <2 x float> + %r173 = sitofp <2 x i8> poison to <2 x float> + %r174 = uitofp <2 x i16> poison to <2 x float> + %r175 = sitofp <2 x i16> poison to <2 x float> + %r176 = uitofp <2 x i32> poison to <2 x float> + %r177 = sitofp <2 x i32> poison to <2 x float> + %r178 = uitofp <2 x i64> poison to <2 x float> + %r179 = sitofp <2 x i64> poison to <2 x float> - %r180 = uitofp <2 x i1> undef to <2 x double> - %r181 = sitofp <2 x i1> undef to <2 x double> - %r182 = uitofp <2 x i8> undef to <2 x double> - %r183 = sitofp <2 x i8> undef to <2 x double> - %r184 = uitofp <2 x i16> undef to <2 x double> - %r185 = sitofp <2 x i16> undef to <2 x double> - %r186 = uitofp <2 x i32> undef to <2 x double> - %r187 = sitofp <2 x i32> undef to <2 x double> - %r188 = uitofp <2 x i64> undef to <2 x double> - %r189 = sitofp <2 x i64> undef to <2 x double> + %r180 = uitofp <2 x i1> poison to <2 x double> + %r181 = sitofp <2 x i1> poison to <2 x double> + %r182 = uitofp <2 x i8> poison to <2 x double> + %r183 = sitofp <2 x i8> poison to <2 x double> + %r184 = uitofp <2 x i16> poison to <2 x double> + %r185 = sitofp <2 x i16> poison to <2 x double> + %r186 = uitofp <2 x i32> poison to <2 x double> + %r187 = sitofp <2 x i32> poison to <2 x double> + %r188 = uitofp <2 x i64> poison to <2 x double> + %r189 = sitofp <2 x i64> poison to <2 x double> - %r190 = uitofp <4 x i1> undef to <4 x float> - %r191 = sitofp <4 x i1> undef to <4 x float> - %r192 = uitofp <4 x i8> undef to <4 x float> - %r193 = sitofp <4 x i8> undef to <4 x float> - %r194 = uitofp <4 x i16> undef to <4 x float> - %r195 = sitofp <4 x i16> undef to <4 x float> - %r196 = uitofp <4 x i32> undef to <4 x float> - %r197 = sitofp <4 x i32> undef to <4 x float> - %r198 = uitofp <4 x i64> undef to <4 x float> - %r199 = sitofp <4 x i64> undef to <4 x float> + %r190 = uitofp <4 x i1> poison to <4 x float> + %r191 = sitofp <4 x i1> poison to <4 x float> + %r192 = uitofp <4 x i8> poison to <4 x float> + %r193 = sitofp <4 x i8> poison to <4 x float> + %r194 = uitofp <4 x i16> poison to <4 x float> + %r195 = sitofp <4 x i16> poison to <4 x float> + %r196 = uitofp <4 x i32> poison to <4 x float> + %r197 = sitofp <4 x i32> poison to <4 x float> + %r198 = uitofp <4 x i64> poison to <4 x float> + %r199 = sitofp <4 x i64> poison to <4 x float> - %r200 = uitofp <4 x i1> undef to <4 x double> - %r201 = sitofp <4 x i1> undef to <4 x double> - %r202 = uitofp <4 x i8> undef to <4 x double> - %r203 = sitofp <4 x i8> undef to <4 x double> - %r204 = uitofp <4 x i16> undef to <4 x double> - %r205 = sitofp <4 x i16> undef to <4 x double> - %r206 = uitofp <4 x i32> undef to <4 x double> - %r207 = sitofp <4 x i32> undef to <4 x double> - %r208 = uitofp <4 x i64> undef to <4 x double> - %r209 = sitofp <4 x i64> undef to <4 x double> + %r200 = uitofp <4 x i1> poison to <4 x double> + %r201 = sitofp <4 x i1> poison to <4 x double> + %r202 = uitofp <4 x i8> poison to <4 x double> + %r203 = sitofp <4 x i8> poison to <4 x double> + %r204 = uitofp <4 x i16> poison to <4 x double> + %r205 = sitofp <4 x i16> poison to <4 x double> + %r206 = uitofp <4 x i32> poison to <4 x double> + %r207 = sitofp <4 x i32> poison to <4 x double> + %r208 = uitofp <4 x i64> poison to <4 x double> + %r209 = sitofp <4 x i64> poison to <4 x double> - %r210 = uitofp <8 x i1> undef to <8 x float> - %r211 = sitofp <8 x i1> undef to <8 x float> - %r212 = uitofp <8 x i8> undef to <8 x float> - %r213 = sitofp <8 x i8> undef to <8 x float> - %r214 = uitofp <8 x i16> undef to <8 x float> - %r215 = sitofp <8 x i16> undef to <8 x float> - %r216 = uitofp <8 x i32> undef to <8 x float> - %r217 = sitofp <8 x i32> undef to <8 x float> - %r218 = uitofp <8 x i64> undef to <8 x float> - %r219 = sitofp <8 x i64> undef to <8 x float> + %r210 = uitofp <8 x i1> poison to <8 x float> + %r211 = sitofp <8 x i1> poison to <8 x float> + %r212 = uitofp <8 x i8> poison to <8 x float> + %r213 = sitofp <8 x i8> poison to <8 x float> + %r214 = uitofp <8 x i16> poison to <8 x float> + %r215 = sitofp <8 x i16> poison to <8 x float> + %r216 = uitofp <8 x i32> poison to <8 x float> + %r217 = sitofp <8 x i32> poison to <8 x float> + %r218 = uitofp <8 x i64> poison to <8 x float> + %r219 = sitofp <8 x i64> poison to <8 x float> - %r220 = uitofp <8 x i1> undef to <8 x double> - %r221 = sitofp <8 x i1> undef to <8 x double> - %r222 = uitofp <8 x i8> undef to <8 x double> - %r223 = sitofp <8 x i8> undef to <8 x double> - %r224 = uitofp <8 x i16> undef to <8 x double> - %r225 = sitofp <8 x i16> undef to <8 x double> - %r226 = uitofp <8 x i16> undef to <8 x double> - %r227 = sitofp <8 x i16> undef to <8 x double> - %r228 = uitofp <8 x i64> undef to <8 x double> - %r229 = sitofp <8 x i64> undef to <8 x double> + %r220 = uitofp <8 x i1> poison to <8 x double> + %r221 = sitofp <8 x i1> poison to <8 x double> + %r222 = uitofp <8 x i8> poison to <8 x double> + %r223 = sitofp <8 x i8> poison to <8 x double> + %r224 = uitofp <8 x i16> poison to <8 x double> + %r225 = sitofp <8 x i16> poison to <8 x double> + %r226 = uitofp <8 x i16> poison to <8 x double> + %r227 = sitofp <8 x i16> poison to <8 x double> + %r228 = uitofp <8 x i64> poison to <8 x double> + %r229 = sitofp <8 x i64> poison to <8 x double> - %r230 = uitofp <16 x i1> undef to <16 x float> - %r231 = sitofp <16 x i1> undef to <16 x float> - %r232 = uitofp <16 x i8> undef to <16 x float> - %r233 = sitofp <16 x i8> undef to <16 x float> - %r234 = uitofp <16 x i16> undef to <16 x float> - %r235 = sitofp <16 x i16> undef to <16 x float> - %r236 = uitofp <16 x i32> undef to <16 x float> - %r237 = sitofp <16 x i32> undef to <16 x float> - %r238 = uitofp <16 x i64> undef to <16 x float> - %r239 = sitofp <16 x i64> undef to <16 x float> + %r230 = uitofp <16 x i1> poison to <16 x float> + %r231 = sitofp <16 x i1> poison to <16 x float> + %r232 = uitofp <16 x i8> poison to <16 x float> + %r233 = sitofp <16 x i8> poison to <16 x float> + %r234 = uitofp <16 x i16> poison to <16 x float> + %r235 = sitofp <16 x i16> poison to <16 x float> + %r236 = uitofp <16 x i32> poison to <16 x float> + %r237 = sitofp <16 x i32> poison to <16 x float> + %r238 = uitofp <16 x i64> poison to <16 x float> + %r239 = sitofp <16 x i64> poison to <16 x float> - %r240 = uitofp <16 x i1> undef to <16 x double> - %r241 = sitofp <16 x i1> undef to <16 x double> - %r242 = uitofp <16 x i8> undef to <16 x double> - %r243 = sitofp <16 x i8> undef to <16 x double> - %r244 = uitofp <16 x i16> undef to <16 x double> - %r245 = sitofp <16 x i16> undef to <16 x double> - %r246 = uitofp <16 x i16> undef to <16 x double> - %r247 = sitofp <16 x i16> undef to <16 x double> - %r248 = uitofp <16 x i64> undef to <16 x double> - %r249 = sitofp <16 x i64> undef to <16 x double> + %r240 = uitofp <16 x i1> poison to <16 x double> + %r241 = sitofp <16 x i1> poison to <16 x double> + %r242 = uitofp <16 x i8> poison to <16 x double> + %r243 = sitofp <16 x i8> poison to <16 x double> + %r244 = uitofp <16 x i16> poison to <16 x double> + %r245 = sitofp <16 x i16> poison to <16 x double> + %r246 = uitofp <16 x i16> poison to <16 x double> + %r247 = sitofp <16 x i16> poison to <16 x double> + %r248 = uitofp <16 x i64> poison to <16 x double> + %r249 = sitofp <16 x i64> poison to <16 x double> ret i32 undef } @@ -836,24 +836,24 @@ define i32 @casts_with_users(i8 %a, i16 %b, i32 %c, i64 %d, i1 %e) { define i32 @bitcasts() { ; CHECK-LABEL: 'bitcasts' -; CHECK-NEXT: Cost Model: Found costs of 0 for: %a = bitcast i32 undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %b = bitcast float undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %c = bitcast i32 undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %d = bitcast float undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %e = bitcast i64 undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %f = bitcast double undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %g = bitcast half undef to i16 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %h = bitcast i16 undef to half +; CHECK-NEXT: Cost Model: Found costs of 0 for: %a = bitcast i32 poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %b = bitcast float poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %c = bitcast i32 poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %d = bitcast float poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %e = bitcast i64 poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %f = bitcast double poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %g = bitcast half poison to i16 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %h = bitcast i16 poison to half ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef ; - %a = bitcast i32 undef to i32 - %b = bitcast float undef to float - %c = bitcast i32 undef to float - %d = bitcast float undef to i32 - %e = bitcast i64 undef to double - %f = bitcast double undef to i64 - %g = bitcast half undef to i16 - %h = bitcast i16 undef to half + %a = bitcast i32 poison to i32 + %b = bitcast float poison to float + %c = bitcast i32 poison to float + %d = bitcast float poison to i32 + %e = bitcast i64 poison to double + %f = bitcast double poison to i64 + %g = bitcast half poison to i16 + %h = bitcast i16 poison to half ret i32 undef } @@ -941,31 +941,31 @@ define i32 @load_extends() { define i32 @store_truncs() { ; CHECK-LABEL: 'store_truncs' -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r0 = trunc i64 undef to i8 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r0 = trunc i64 poison to i8 ; CHECK-NEXT: Cost Model: Found costs of 1 for: store i8 %r0, ptr undef, align 1 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r1 = trunc i64 undef to i16 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r1 = trunc i64 poison to i16 ; CHECK-NEXT: Cost Model: Found costs of 1 for: store i16 %r1, ptr undef, align 2 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r2 = trunc i64 undef to i32 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r2 = trunc i64 poison to i32 ; CHECK-NEXT: Cost Model: Found costs of 1 for: store i32 %r2, ptr undef, align 4 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r3 = trunc i32 undef to i8 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r3 = trunc i32 poison to i8 ; CHECK-NEXT: Cost Model: Found costs of 1 for: store i8 %r3, ptr undef, align 1 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r4 = trunc i32 undef to i16 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r4 = trunc i32 poison to i16 ; CHECK-NEXT: Cost Model: Found costs of 1 for: store i16 %r4, ptr undef, align 2 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r5 = trunc i16 undef to i8 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r5 = trunc i16 poison to i8 ; CHECK-NEXT: Cost Model: Found costs of 1 for: store i8 %r5, ptr undef, align 1 ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef ; - %r0 = trunc i64 undef to i8 + %r0 = trunc i64 poison to i8 store i8 %r0, ptr undef - %r1 = trunc i64 undef to i16 + %r1 = trunc i64 poison to i16 store i16 %r1, ptr undef - %r2 = trunc i64 undef to i32 + %r2 = trunc i64 poison to i32 store i32 %r2, ptr undef - %r3 = trunc i32 undef to i8 + %r3 = trunc i32 poison to i8 store i8 %r3, ptr undef - %r4 = trunc i32 undef to i16 + %r4 = trunc i32 poison to i16 store i16 %r4, ptr undef - %r5 = trunc i16 undef to i8 + %r5 = trunc i16 poison to i8 store i8 %r5, ptr undef ret i32 undef } @@ -1013,296 +1013,296 @@ declare void @use(i16, i16, i32, i32, i64, i64, i32, i32, i64, i64, i64, i64) define void @fp16cast() { ; CHECK-NOFP16-LABEL: 'fp16cast' -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui half undef to i1 -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi half undef to i1 -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui half undef to i8 -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi half undef to i8 -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui half undef to i16 -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi half undef to i16 -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui half undef to i32 -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi half undef to i32 -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui half undef to i64 -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi half undef to i64 -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x half> undef to <2 x i1> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x half> undef to <2 x i1> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x half> undef to <2 x i8> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x half> undef to <2 x i8> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x half> undef to <2 x i16> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x half> undef to <2 x i16> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x half> undef to <2 x i32> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x half> undef to <2 x i32> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x half> undef to <2 x i64> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x half> undef to <2 x i64> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x half> undef to <4 x i1> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x half> undef to <4 x i1> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x half> undef to <4 x i8> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x half> undef to <4 x i8> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x half> undef to <4 x i16> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x half> undef to <4 x i16> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r116 = fptoui <4 x half> undef to <4 x i32> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r117 = fptosi <4 x half> undef to <4 x i32> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x half> undef to <4 x i64> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x half> undef to <4 x i64> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x half> undef to <8 x i1> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x half> undef to <8 x i1> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x half> undef to <8 x i8> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x half> undef to <8 x i8> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x half> undef to <8 x i16> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x half> undef to <8 x i16> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:41 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x half> undef to <8 x i32> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:41 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x half> undef to <8 x i32> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x half> undef to <8 x i64> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x half> undef to <8 x i64> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x half> undef to <16 x i1> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x half> undef to <16 x i1> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x half> undef to <16 x i8> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x half> undef to <16 x i8> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x half> undef to <16 x i16> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x half> undef to <16 x i16> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:82 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x half> undef to <16 x i32> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:82 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x half> undef to <16 x i32> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:86 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x half> undef to <16 x i64> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:86 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x half> undef to <16 x i64> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r250 = uitofp <8 x i1> undef to <8 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r251 = sitofp <8 x i1> undef to <8 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r252 = uitofp <8 x i8> undef to <8 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r253 = sitofp <8 x i8> undef to <8 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r254 = uitofp <8 x i16> undef to <8 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r255 = sitofp <8 x i16> undef to <8 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r256 = uitofp <8 x i32> undef to <8 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r257 = sitofp <8 x i32> undef to <8 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r258 = uitofp <8 x i64> undef to <8 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r259 = sitofp <8 x i64> undef to <8 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r260 = uitofp <16 x i1> undef to <16 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r261 = sitofp <16 x i1> undef to <16 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r262 = uitofp <16 x i8> undef to <16 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r263 = sitofp <16 x i8> undef to <16 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r264 = uitofp <16 x i16> undef to <16 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r265 = sitofp <16 x i16> undef to <16 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %r266 = uitofp <16 x i32> undef to <16 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %r267 = sitofp <16 x i32> undef to <16 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %r268 = uitofp <16 x i64> undef to <16 x half> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %r269 = sitofp <16 x i64> undef to <16 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui half poison to i1 +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi half poison to i1 +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui half poison to i8 +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi half poison to i8 +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui half poison to i16 +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi half poison to i16 +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui half poison to i32 +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi half poison to i32 +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui half poison to i64 +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi half poison to i64 +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x half> poison to <2 x i1> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x half> poison to <2 x i1> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x half> poison to <2 x i8> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x half> poison to <2 x i8> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x half> poison to <2 x i16> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x half> poison to <2 x i16> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x half> poison to <2 x i32> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x half> poison to <2 x i32> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x half> poison to <2 x i64> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x half> poison to <2 x i64> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x half> poison to <4 x i1> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x half> poison to <4 x i1> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x half> poison to <4 x i8> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x half> poison to <4 x i8> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x half> poison to <4 x i16> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x half> poison to <4 x i16> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r116 = fptoui <4 x half> poison to <4 x i32> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r117 = fptosi <4 x half> poison to <4 x i32> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x half> poison to <4 x i64> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x half> poison to <4 x i64> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x half> poison to <8 x i1> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x half> poison to <8 x i1> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x half> poison to <8 x i8> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x half> poison to <8 x i8> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x half> poison to <8 x i16> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x half> poison to <8 x i16> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:41 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x half> poison to <8 x i32> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:41 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x half> poison to <8 x i32> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x half> poison to <8 x i64> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x half> poison to <8 x i64> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x half> poison to <16 x i1> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x half> poison to <16 x i1> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x half> poison to <16 x i8> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x half> poison to <16 x i8> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x half> poison to <16 x i16> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x half> poison to <16 x i16> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:82 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x half> poison to <16 x i32> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:82 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x half> poison to <16 x i32> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:86 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x half> poison to <16 x i64> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:86 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x half> poison to <16 x i64> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r250 = uitofp <8 x i1> poison to <8 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r251 = sitofp <8 x i1> poison to <8 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r252 = uitofp <8 x i8> poison to <8 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r253 = sitofp <8 x i8> poison to <8 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r254 = uitofp <8 x i16> poison to <8 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %r255 = sitofp <8 x i16> poison to <8 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r256 = uitofp <8 x i32> poison to <8 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r257 = sitofp <8 x i32> poison to <8 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r258 = uitofp <8 x i64> poison to <8 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r259 = sitofp <8 x i64> poison to <8 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r260 = uitofp <16 x i1> poison to <16 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r261 = sitofp <16 x i1> poison to <16 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r262 = uitofp <16 x i8> poison to <16 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r263 = sitofp <16 x i8> poison to <16 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r264 = uitofp <16 x i16> poison to <16 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r265 = sitofp <16 x i16> poison to <16 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %r266 = uitofp <16 x i32> poison to <16 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %r267 = sitofp <16 x i32> poison to <16 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %r268 = uitofp <16 x i64> poison to <16 x half> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %r269 = sitofp <16 x i64> poison to <16 x half> ; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-FP16-LABEL: 'fp16cast' -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui half undef to i1 -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi half undef to i1 -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui half undef to i8 -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi half undef to i8 -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui half undef to i16 -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi half undef to i16 -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui half undef to i32 -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi half undef to i32 -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui half undef to i64 -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi half undef to i64 -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x half> undef to <2 x i1> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x half> undef to <2 x i1> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x half> undef to <2 x i8> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x half> undef to <2 x i8> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x half> undef to <2 x i16> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x half> undef to <2 x i16> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x half> undef to <2 x i32> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x half> undef to <2 x i32> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x half> undef to <2 x i64> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x half> undef to <2 x i64> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x half> undef to <4 x i1> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x half> undef to <4 x i1> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x half> undef to <4 x i8> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x half> undef to <4 x i8> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x half> undef to <4 x i16> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x half> undef to <4 x i16> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r116 = fptoui <4 x half> undef to <4 x i32> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r117 = fptosi <4 x half> undef to <4 x i32> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x half> undef to <4 x i64> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x half> undef to <4 x i64> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x half> undef to <8 x i1> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x half> undef to <8 x i1> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x half> undef to <8 x i8> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x half> undef to <8 x i8> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x half> undef to <8 x i16> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x half> undef to <8 x i16> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x half> undef to <8 x i32> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x half> undef to <8 x i32> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x half> undef to <8 x i64> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x half> undef to <8 x i64> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x half> undef to <16 x i1> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x half> undef to <16 x i1> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x half> undef to <16 x i8> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x half> undef to <16 x i8> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x half> undef to <16 x i16> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x half> undef to <16 x i16> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x half> undef to <16 x i32> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x half> undef to <16 x i32> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:86 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x half> undef to <16 x i64> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:86 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x half> undef to <16 x i64> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r250 = uitofp <8 x i1> undef to <8 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r251 = sitofp <8 x i1> undef to <8 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r252 = uitofp <8 x i8> undef to <8 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r253 = sitofp <8 x i8> undef to <8 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r254 = uitofp <8 x i16> undef to <8 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r255 = sitofp <8 x i16> undef to <8 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r256 = uitofp <8 x i32> undef to <8 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r257 = sitofp <8 x i32> undef to <8 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r258 = uitofp <8 x i64> undef to <8 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r259 = sitofp <8 x i64> undef to <8 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r260 = uitofp <16 x i1> undef to <16 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r261 = sitofp <16 x i1> undef to <16 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r262 = uitofp <16 x i8> undef to <16 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r263 = sitofp <16 x i8> undef to <16 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r264 = uitofp <16 x i16> undef to <16 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r265 = sitofp <16 x i16> undef to <16 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %r266 = uitofp <16 x i32> undef to <16 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %r267 = sitofp <16 x i32> undef to <16 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %r268 = uitofp <16 x i64> undef to <16 x half> -; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %r269 = sitofp <16 x i64> undef to <16 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui half poison to i1 +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi half poison to i1 +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui half poison to i8 +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi half poison to i8 +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui half poison to i16 +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi half poison to i16 +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui half poison to i32 +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi half poison to i32 +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui half poison to i64 +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi half poison to i64 +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x half> poison to <2 x i1> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x half> poison to <2 x i1> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x half> poison to <2 x i8> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x half> poison to <2 x i8> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x half> poison to <2 x i16> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x half> poison to <2 x i16> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x half> poison to <2 x i32> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x half> poison to <2 x i32> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x half> poison to <2 x i64> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x half> poison to <2 x i64> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x half> poison to <4 x i1> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x half> poison to <4 x i1> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x half> poison to <4 x i8> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x half> poison to <4 x i8> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x half> poison to <4 x i16> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x half> poison to <4 x i16> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r116 = fptoui <4 x half> poison to <4 x i32> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r117 = fptosi <4 x half> poison to <4 x i32> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x half> poison to <4 x i64> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x half> poison to <4 x i64> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x half> poison to <8 x i1> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x half> poison to <8 x i1> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x half> poison to <8 x i8> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x half> poison to <8 x i8> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x half> poison to <8 x i16> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x half> poison to <8 x i16> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x half> poison to <8 x i32> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x half> poison to <8 x i32> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x half> poison to <8 x i64> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x half> poison to <8 x i64> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x half> poison to <16 x i1> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x half> poison to <16 x i1> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x half> poison to <16 x i8> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x half> poison to <16 x i8> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x half> poison to <16 x i16> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x half> poison to <16 x i16> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x half> poison to <16 x i32> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x half> poison to <16 x i32> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:86 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x half> poison to <16 x i64> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:86 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x half> poison to <16 x i64> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r250 = uitofp <8 x i1> poison to <8 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r251 = sitofp <8 x i1> poison to <8 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r252 = uitofp <8 x i8> poison to <8 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r253 = sitofp <8 x i8> poison to <8 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r254 = uitofp <8 x i16> poison to <8 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of 1 for: %r255 = sitofp <8 x i16> poison to <8 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r256 = uitofp <8 x i32> poison to <8 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r257 = sitofp <8 x i32> poison to <8 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r258 = uitofp <8 x i64> poison to <8 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r259 = sitofp <8 x i64> poison to <8 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r260 = uitofp <16 x i1> poison to <16 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r261 = sitofp <16 x i1> poison to <16 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r262 = uitofp <16 x i8> poison to <16 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r263 = sitofp <16 x i8> poison to <16 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r264 = uitofp <16 x i16> poison to <16 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r265 = sitofp <16 x i16> poison to <16 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %r266 = uitofp <16 x i32> poison to <16 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %r267 = sitofp <16 x i32> poison to <16 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %r268 = uitofp <16 x i64> poison to <16 x half> +; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %r269 = sitofp <16 x i64> poison to <16 x half> ; CHECK-FP16-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %r30 = fptoui half undef to i1 - %r31 = fptosi half undef to i1 - %r32 = fptoui half undef to i8 - %r33 = fptosi half undef to i8 - %r34 = fptoui half undef to i16 - %r35 = fptosi half undef to i16 - %r36 = fptoui half undef to i32 - %r37 = fptosi half undef to i32 - %r38 = fptoui half undef to i64 - %r39 = fptosi half undef to i64 + %r30 = fptoui half poison to i1 + %r31 = fptosi half poison to i1 + %r32 = fptoui half poison to i8 + %r33 = fptosi half poison to i8 + %r34 = fptoui half poison to i16 + %r35 = fptosi half poison to i16 + %r36 = fptoui half poison to i32 + %r37 = fptosi half poison to i32 + %r38 = fptoui half poison to i64 + %r39 = fptosi half poison to i64 - %r90 = fptoui <2 x half> undef to <2 x i1> - %r91 = fptosi <2 x half> undef to <2 x i1> - %r92 = fptoui <2 x half> undef to <2 x i8> - %r93 = fptosi <2 x half> undef to <2 x i8> - %r94 = fptoui <2 x half> undef to <2 x i16> - %r95 = fptosi <2 x half> undef to <2 x i16> - %r96 = fptoui <2 x half> undef to <2 x i32> - %r97 = fptosi <2 x half> undef to <2 x i32> - %r98 = fptoui <2 x half> undef to <2 x i64> - %r99 = fptosi <2 x half> undef to <2 x i64> + %r90 = fptoui <2 x half> poison to <2 x i1> + %r91 = fptosi <2 x half> poison to <2 x i1> + %r92 = fptoui <2 x half> poison to <2 x i8> + %r93 = fptosi <2 x half> poison to <2 x i8> + %r94 = fptoui <2 x half> poison to <2 x i16> + %r95 = fptosi <2 x half> poison to <2 x i16> + %r96 = fptoui <2 x half> poison to <2 x i32> + %r97 = fptosi <2 x half> poison to <2 x i32> + %r98 = fptoui <2 x half> poison to <2 x i64> + %r99 = fptosi <2 x half> poison to <2 x i64> - %r110 = fptoui <4 x half> undef to <4 x i1> - %r111 = fptosi <4 x half> undef to <4 x i1> - %r112 = fptoui <4 x half> undef to <4 x i8> - %r113 = fptosi <4 x half> undef to <4 x i8> - %r114 = fptoui <4 x half> undef to <4 x i16> - %r115 = fptosi <4 x half> undef to <4 x i16> - %r116 = fptoui <4 x half> undef to <4 x i32> - %r117 = fptosi <4 x half> undef to <4 x i32> - %r118 = fptoui <4 x half> undef to <4 x i64> - %r119 = fptosi <4 x half> undef to <4 x i64> + %r110 = fptoui <4 x half> poison to <4 x i1> + %r111 = fptosi <4 x half> poison to <4 x i1> + %r112 = fptoui <4 x half> poison to <4 x i8> + %r113 = fptosi <4 x half> poison to <4 x i8> + %r114 = fptoui <4 x half> poison to <4 x i16> + %r115 = fptosi <4 x half> poison to <4 x i16> + %r116 = fptoui <4 x half> poison to <4 x i32> + %r117 = fptosi <4 x half> poison to <4 x i32> + %r118 = fptoui <4 x half> poison to <4 x i64> + %r119 = fptosi <4 x half> poison to <4 x i64> - %r130 = fptoui <8 x half> undef to <8 x i1> - %r131 = fptosi <8 x half> undef to <8 x i1> - %r132 = fptoui <8 x half> undef to <8 x i8> - %r133 = fptosi <8 x half> undef to <8 x i8> - %r134 = fptoui <8 x half> undef to <8 x i16> - %r135 = fptosi <8 x half> undef to <8 x i16> - %r136 = fptoui <8 x half> undef to <8 x i32> - %r137 = fptosi <8 x half> undef to <8 x i32> - %r138 = fptoui <8 x half> undef to <8 x i64> - %r139 = fptosi <8 x half> undef to <8 x i64> + %r130 = fptoui <8 x half> poison to <8 x i1> + %r131 = fptosi <8 x half> poison to <8 x i1> + %r132 = fptoui <8 x half> poison to <8 x i8> + %r133 = fptosi <8 x half> poison to <8 x i8> + %r134 = fptoui <8 x half> poison to <8 x i16> + %r135 = fptosi <8 x half> poison to <8 x i16> + %r136 = fptoui <8 x half> poison to <8 x i32> + %r137 = fptosi <8 x half> poison to <8 x i32> + %r138 = fptoui <8 x half> poison to <8 x i64> + %r139 = fptosi <8 x half> poison to <8 x i64> - %r150 = fptoui <16 x half> undef to <16 x i1> - %r151 = fptosi <16 x half> undef to <16 x i1> - %r152 = fptoui <16 x half> undef to <16 x i8> - %r153 = fptosi <16 x half> undef to <16 x i8> - %r154 = fptoui <16 x half> undef to <16 x i16> - %r155 = fptosi <16 x half> undef to <16 x i16> - %r156 = fptoui <16 x half> undef to <16 x i32> - %r157 = fptosi <16 x half> undef to <16 x i32> - %r158 = fptoui <16 x half> undef to <16 x i64> - %r159 = fptosi <16 x half> undef to <16 x i64> + %r150 = fptoui <16 x half> poison to <16 x i1> + %r151 = fptosi <16 x half> poison to <16 x i1> + %r152 = fptoui <16 x half> poison to <16 x i8> + %r153 = fptosi <16 x half> poison to <16 x i8> + %r154 = fptoui <16 x half> poison to <16 x i16> + %r155 = fptosi <16 x half> poison to <16 x i16> + %r156 = fptoui <16 x half> poison to <16 x i32> + %r157 = fptosi <16 x half> poison to <16 x i32> + %r158 = fptoui <16 x half> poison to <16 x i64> + %r159 = fptosi <16 x half> poison to <16 x i64> - %r250 = uitofp <8 x i1> undef to <8 x half> - %r251 = sitofp <8 x i1> undef to <8 x half> - %r252 = uitofp <8 x i8> undef to <8 x half> - %r253 = sitofp <8 x i8> undef to <8 x half> - %r254 = uitofp <8 x i16> undef to <8 x half> - %r255 = sitofp <8 x i16> undef to <8 x half> - %r256 = uitofp <8 x i32> undef to <8 x half> - %r257 = sitofp <8 x i32> undef to <8 x half> - %r258 = uitofp <8 x i64> undef to <8 x half> - %r259 = sitofp <8 x i64> undef to <8 x half> + %r250 = uitofp <8 x i1> poison to <8 x half> + %r251 = sitofp <8 x i1> poison to <8 x half> + %r252 = uitofp <8 x i8> poison to <8 x half> + %r253 = sitofp <8 x i8> poison to <8 x half> + %r254 = uitofp <8 x i16> poison to <8 x half> + %r255 = sitofp <8 x i16> poison to <8 x half> + %r256 = uitofp <8 x i32> poison to <8 x half> + %r257 = sitofp <8 x i32> poison to <8 x half> + %r258 = uitofp <8 x i64> poison to <8 x half> + %r259 = sitofp <8 x i64> poison to <8 x half> - %r260 = uitofp <16 x i1> undef to <16 x half> - %r261 = sitofp <16 x i1> undef to <16 x half> - %r262 = uitofp <16 x i8> undef to <16 x half> - %r263 = sitofp <16 x i8> undef to <16 x half> - %r264 = uitofp <16 x i16> undef to <16 x half> - %r265 = sitofp <16 x i16> undef to <16 x half> - %r266 = uitofp <16 x i32> undef to <16 x half> - %r267 = sitofp <16 x i32> undef to <16 x half> - %r268 = uitofp <16 x i64> undef to <16 x half> - %r269 = sitofp <16 x i64> undef to <16 x half> + %r260 = uitofp <16 x i1> poison to <16 x half> + %r261 = sitofp <16 x i1> poison to <16 x half> + %r262 = uitofp <16 x i8> poison to <16 x half> + %r263 = sitofp <16 x i8> poison to <16 x half> + %r264 = uitofp <16 x i16> poison to <16 x half> + %r265 = sitofp <16 x i16> poison to <16 x half> + %r266 = uitofp <16 x i32> poison to <16 x half> + %r267 = sitofp <16 x i32> poison to <16 x half> + %r268 = uitofp <16 x i64> poison to <16 x half> + %r269 = sitofp <16 x i64> poison to <16 x half> ret void } define void @bf16cast() { ; CHECK-NOFP16-LABEL: 'bf16cast' -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %extf16f32 = fpext bfloat undef to float -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %extv2f16f32 = fpext <2 x bfloat> undef to <2 x float> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %extv4f16f32 = fpext <4 x bfloat> undef to <4 x float> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extv8f16f32 = fpext <8 x bfloat> undef to <8 x float> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %extv16f16f32 = fpext <16 x bfloat> undef to <16 x float> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extf16f64 = fpext bfloat undef to double -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extv2f16f64 = fpext <2 x bfloat> undef to <2 x double> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %extv4f16f64 = fpext <4 x bfloat> undef to <4 x double> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %extv8f16f64 = fpext <8 x bfloat> undef to <8 x double> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %extv16f16f64 = fpext <16 x bfloat> undef to <16 x double> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %truncf16f32 = fptrunc float undef to bfloat -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %truncv2f16f32 = fptrunc <2 x float> undef to <2 x bfloat> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %truncv4f16f32 = fptrunc <4 x float> undef to <4 x bfloat> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %truncv8f16f32 = fptrunc <8 x float> undef to <8 x bfloat> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:30 CodeSize:1 Lat:1 SizeLat:1 for: %truncv16f16f32 = fptrunc <16 x float> undef to <16 x bfloat> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %truncf16f64 = fptrunc double undef to bfloat -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %truncv2f16f64 = fptrunc <2 x double> undef to <2 x bfloat> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %truncv4f16f64 = fptrunc <4 x double> undef to <4 x bfloat> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %truncv8f16f64 = fptrunc <8 x double> undef to <8 x bfloat> -; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %truncv16f16f64 = fptrunc <16 x double> undef to <16 x bfloat> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %extf16f32 = fpext bfloat poison to float +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %extv2f16f32 = fpext <2 x bfloat> poison to <2 x float> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of 1 for: %extv4f16f32 = fpext <4 x bfloat> poison to <4 x float> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extv8f16f32 = fpext <8 x bfloat> poison to <8 x float> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %extv16f16f32 = fpext <16 x bfloat> poison to <16 x float> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extf16f64 = fpext bfloat poison to double +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extv2f16f64 = fpext <2 x bfloat> poison to <2 x double> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %extv4f16f64 = fpext <4 x bfloat> poison to <4 x double> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %extv8f16f64 = fpext <8 x bfloat> poison to <8 x double> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %extv16f16f64 = fpext <16 x bfloat> poison to <16 x double> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %truncf16f32 = fptrunc float poison to bfloat +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %truncv2f16f32 = fptrunc <2 x float> poison to <2 x bfloat> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %truncv4f16f32 = fptrunc <4 x float> poison to <4 x bfloat> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %truncv8f16f32 = fptrunc <8 x float> poison to <8 x bfloat> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:30 CodeSize:1 Lat:1 SizeLat:1 for: %truncv16f16f32 = fptrunc <16 x float> poison to <16 x bfloat> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %truncf16f64 = fptrunc double poison to bfloat +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %truncv2f16f64 = fptrunc <2 x double> poison to <2 x bfloat> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %truncv4f16f64 = fptrunc <4 x double> poison to <4 x bfloat> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %truncv8f16f64 = fptrunc <8 x double> poison to <8 x bfloat> +; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %truncv16f16f64 = fptrunc <16 x double> poison to <16 x bfloat> ; CHECK-NOFP16-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-BF16-LABEL: 'bf16cast' -; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %extf16f32 = fpext bfloat undef to float -; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %extv2f16f32 = fpext <2 x bfloat> undef to <2 x float> -; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %extv4f16f32 = fpext <4 x bfloat> undef to <4 x float> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extv8f16f32 = fpext <8 x bfloat> undef to <8 x float> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %extv16f16f32 = fpext <16 x bfloat> undef to <16 x float> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extf16f64 = fpext bfloat undef to double -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extv2f16f64 = fpext <2 x bfloat> undef to <2 x double> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %extv4f16f64 = fpext <4 x bfloat> undef to <4 x double> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %extv8f16f64 = fpext <8 x bfloat> undef to <8 x double> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %extv16f16f64 = fpext <16 x bfloat> undef to <16 x double> -; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %truncf16f32 = fptrunc float undef to bfloat -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %truncv2f16f32 = fptrunc <2 x float> undef to <2 x bfloat> -; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %truncv4f16f32 = fptrunc <4 x float> undef to <4 x bfloat> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %truncv8f16f32 = fptrunc <8 x float> undef to <8 x bfloat> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %truncv16f16f32 = fptrunc <16 x float> undef to <16 x bfloat> -; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %truncf16f64 = fptrunc double undef to bfloat -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %truncv2f16f64 = fptrunc <2 x double> undef to <2 x bfloat> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %truncv4f16f64 = fptrunc <4 x double> undef to <4 x bfloat> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %truncv8f16f64 = fptrunc <8 x double> undef to <8 x bfloat> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %truncv16f16f64 = fptrunc <16 x double> undef to <16 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %extf16f32 = fpext bfloat poison to float +; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %extv2f16f32 = fpext <2 x bfloat> poison to <2 x float> +; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %extv4f16f32 = fpext <4 x bfloat> poison to <4 x float> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extv8f16f32 = fpext <8 x bfloat> poison to <8 x float> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %extv16f16f32 = fpext <16 x bfloat> poison to <16 x float> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extf16f64 = fpext bfloat poison to double +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %extv2f16f64 = fpext <2 x bfloat> poison to <2 x double> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %extv4f16f64 = fpext <4 x bfloat> poison to <4 x double> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %extv8f16f64 = fpext <8 x bfloat> poison to <8 x double> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %extv16f16f64 = fpext <16 x bfloat> poison to <16 x double> +; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %truncf16f32 = fptrunc float poison to bfloat +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %truncv2f16f32 = fptrunc <2 x float> poison to <2 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %truncv4f16f32 = fptrunc <4 x float> poison to <4 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %truncv8f16f32 = fptrunc <8 x float> poison to <8 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %truncv16f16f32 = fptrunc <16 x float> poison to <16 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %truncf16f64 = fptrunc double poison to bfloat +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %truncv2f16f64 = fptrunc <2 x double> poison to <2 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %truncv4f16f64 = fptrunc <4 x double> poison to <4 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %truncv8f16f64 = fptrunc <8 x double> poison to <8 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %truncv16f16f64 = fptrunc <16 x double> poison to <16 x bfloat> ; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %extf16f32 = fpext bfloat undef to float - %extv2f16f32 = fpext <2 x bfloat> undef to <2 x float> - %extv4f16f32 = fpext <4 x bfloat> undef to <4 x float> - %extv8f16f32 = fpext <8 x bfloat> undef to <8 x float> - %extv16f16f32 = fpext <16 x bfloat> undef to <16 x float> - %extf16f64 = fpext bfloat undef to double - %extv2f16f64 = fpext <2 x bfloat> undef to <2 x double> - %extv4f16f64 = fpext <4 x bfloat> undef to <4 x double> - %extv8f16f64 = fpext <8 x bfloat> undef to <8 x double> - %extv16f16f64 = fpext <16 x bfloat> undef to <16 x double> - %truncf16f32 = fptrunc float undef to bfloat - %truncv2f16f32 = fptrunc <2 x float> undef to <2 x bfloat> - %truncv4f16f32 = fptrunc <4 x float> undef to <4 x bfloat> - %truncv8f16f32 = fptrunc <8 x float> undef to <8 x bfloat> - %truncv16f16f32 = fptrunc <16 x float> undef to <16 x bfloat> - %truncf16f64 = fptrunc double undef to bfloat - %truncv2f16f64 = fptrunc <2 x double> undef to <2 x bfloat> - %truncv4f16f64 = fptrunc <4 x double> undef to <4 x bfloat> - %truncv8f16f64 = fptrunc <8 x double> undef to <8 x bfloat> - %truncv16f16f64 = fptrunc <16 x double> undef to <16 x bfloat> + %extf16f32 = fpext bfloat poison to float + %extv2f16f32 = fpext <2 x bfloat> poison to <2 x float> + %extv4f16f32 = fpext <4 x bfloat> poison to <4 x float> + %extv8f16f32 = fpext <8 x bfloat> poison to <8 x float> + %extv16f16f32 = fpext <16 x bfloat> poison to <16 x float> + %extf16f64 = fpext bfloat poison to double + %extv2f16f64 = fpext <2 x bfloat> poison to <2 x double> + %extv4f16f64 = fpext <4 x bfloat> poison to <4 x double> + %extv8f16f64 = fpext <8 x bfloat> poison to <8 x double> + %extv16f16f64 = fpext <16 x bfloat> poison to <16 x double> + %truncf16f32 = fptrunc float poison to bfloat + %truncv2f16f32 = fptrunc <2 x float> poison to <2 x bfloat> + %truncv4f16f32 = fptrunc <4 x float> poison to <4 x bfloat> + %truncv8f16f32 = fptrunc <8 x float> poison to <8 x bfloat> + %truncv16f16f32 = fptrunc <16 x float> poison to <16 x bfloat> + %truncf16f64 = fptrunc double poison to bfloat + %truncv2f16f64 = fptrunc <2 x double> poison to <2 x bfloat> + %truncv4f16f64 = fptrunc <4 x double> poison to <4 x bfloat> + %truncv8f16f64 = fptrunc <8 x double> poison to <8 x bfloat> + %truncv16f16f64 = fptrunc <16 x double> poison to <16 x bfloat> ret void } diff --git a/llvm/test/Analysis/CostModel/AArch64/no-sve-no-neon.ll b/llvm/test/Analysis/CostModel/AArch64/no-sve-no-neon.ll index 20b83be..9aea58e 100644 --- a/llvm/test/Analysis/CostModel/AArch64/no-sve-no-neon.ll +++ b/llvm/test/Analysis/CostModel/AArch64/no-sve-no-neon.ll @@ -7,13 +7,13 @@ target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128" define void @uitofp() { ; CHECK-NONEON-LABEL: 'uitofp' -; CHECK-NONEON-NEXT: Cost Model: Found costs of RThru:48 CodeSize:1 Lat:1 SizeLat:1 for: %conv = uitofp <16 x i64> undef to <16 x float> +; CHECK-NONEON-NEXT: Cost Model: Found costs of RThru:48 CodeSize:1 Lat:1 SizeLat:1 for: %conv = uitofp <16 x i64> poison to <16 x float> ; CHECK-NONEON-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-WITHSVE-LABEL: 'uitofp' -; CHECK-WITHSVE-NEXT: Cost Model: Found costs of RThru:16 CodeSize:1 Lat:1 SizeLat:1 for: %conv = uitofp <16 x i64> undef to <16 x float> +; CHECK-WITHSVE-NEXT: Cost Model: Found costs of RThru:16 CodeSize:1 Lat:1 SizeLat:1 for: %conv = uitofp <16 x i64> poison to <16 x float> ; CHECK-WITHSVE-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %conv = uitofp <16 x i64> undef to <16 x float> + %conv = uitofp <16 x i64> poison to <16 x float> ret void } diff --git a/llvm/test/Analysis/CostModel/AArch64/sve-cast.ll b/llvm/test/Analysis/CostModel/AArch64/sve-cast.ll index cfb130e..ecb4e14 100644 --- a/llvm/test/Analysis/CostModel/AArch64/sve-cast.ll +++ b/llvm/test/Analysis/CostModel/AArch64/sve-cast.ll @@ -8,1631 +8,1631 @@ target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128" define void @ext() { ; CHECK-SVE-LABEL: 'ext' -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r0 = sext i1 undef to i8 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r1 = zext i1 undef to i8 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r2 = sext i1 undef to i16 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r3 = zext i1 undef to i16 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r4 = sext i1 undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r5 = zext i1 undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r6 = sext i1 undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r7 = zext i1 undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r9 = sext i8 undef to i16 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r10 = zext i8 undef to i16 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r11 = sext i8 undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r12 = zext i8 undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r13 = sext i8 undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r14 = zext i8 undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r17 = sext i16 undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r18 = zext i16 undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r19 = sext i16 undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r20 = zext i16 undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r24 = sext i32 undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r25 = zext i32 undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i8i16 = sext <2 x i8> undef to <2 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i8i16 = zext <2 x i8> undef to <2 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i8i32 = sext <2 x i8> undef to <2 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i8i32 = zext <2 x i8> undef to <2 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = sext <2 x i8> undef to <2 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i8i64 = zext <2 x i8> undef to <2 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i16i32 = sext <2 x i16> undef to <2 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i16i32 = zext <2 x i16> undef to <2 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = sext <2 x i16> undef to <2 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i16i64 = zext <2 x i16> undef to <2 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = sext <2 x i32> undef to <2 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> undef to <2 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> undef to <4 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> undef to <4 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i64 = sext <4 x i8> undef to <4 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z4i8i64 = zext <4 x i8> undef to <4 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> undef to <4 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z4i16i32 = zext <4 x i16> undef to <4 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i16i64 = sext <4 x i16> undef to <4 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z4i16i64 = zext <4 x i16> undef to <4 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i32i64 = sext <4 x i32> undef to <4 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i32i64 = zext <4 x i32> undef to <4 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = sext <8 x i8> undef to <8 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z8i8i16 = zext <8 x i8> undef to <8 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i32 = sext <8 x i8> undef to <8 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i32 = zext <8 x i8> undef to <8 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i64 = sext <8 x i8> undef to <8 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i64 = zext <8 x i8> undef to <8 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i32 = sext <8 x i16> undef to <8 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i32 = zext <8 x i16> undef to <8 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i64 = sext <8 x i16> undef to <8 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i64 = zext <8 x i16> undef to <8 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i32i64 = sext <8 x i32> undef to <8 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z8i32i64 = zext <8 x i32> undef to <8 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i16 = sext <16 x i8> undef to <16 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i16 = zext <16 x i8> undef to <16 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i32 = sext <16 x i8> undef to <16 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i32 = zext <16 x i8> undef to <16 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i64 = sext <16 x i8> undef to <16 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i64 = zext <16 x i8> undef to <16 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i32 = sext <16 x i16> undef to <16 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i32 = zext <16 x i16> undef to <16 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i64 = sext <16 x i16> undef to <16 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i64 = zext <16 x i16> undef to <16 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %s16i32i64 = sext <16 x i32> undef to <16 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %z16i32i64 = zext <16 x i32> undef to <16 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r0 = sext i1 poison to i8 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r1 = zext i1 poison to i8 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r2 = sext i1 poison to i16 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r3 = zext i1 poison to i16 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r4 = sext i1 poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r5 = zext i1 poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r6 = sext i1 poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r7 = zext i1 poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r9 = sext i8 poison to i16 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r10 = zext i8 poison to i16 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r11 = sext i8 poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r12 = zext i8 poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r13 = sext i8 poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r14 = zext i8 poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r17 = sext i16 poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r18 = zext i16 poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r19 = sext i16 poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r20 = zext i16 poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r24 = sext i32 poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r25 = zext i32 poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i8i16 = sext <2 x i8> poison to <2 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i8i16 = zext <2 x i8> poison to <2 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i8i32 = sext <2 x i8> poison to <2 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i8i32 = zext <2 x i8> poison to <2 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = sext <2 x i8> poison to <2 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i8i64 = zext <2 x i8> poison to <2 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i16i32 = sext <2 x i16> poison to <2 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i16i32 = zext <2 x i16> poison to <2 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = sext <2 x i16> poison to <2 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i16i64 = zext <2 x i16> poison to <2 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = sext <2 x i32> poison to <2 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> poison to <2 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> poison to <4 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> poison to <4 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> poison to <4 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> poison to <4 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i64 = sext <4 x i8> poison to <4 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z4i8i64 = zext <4 x i8> poison to <4 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> poison to <4 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z4i16i32 = zext <4 x i16> poison to <4 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i16i64 = sext <4 x i16> poison to <4 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z4i16i64 = zext <4 x i16> poison to <4 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i32i64 = sext <4 x i32> poison to <4 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i32i64 = zext <4 x i32> poison to <4 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = sext <8 x i8> poison to <8 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %z8i8i16 = zext <8 x i8> poison to <8 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i32 = sext <8 x i8> poison to <8 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i32 = zext <8 x i8> poison to <8 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i64 = sext <8 x i8> poison to <8 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i64 = zext <8 x i8> poison to <8 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i32 = sext <8 x i16> poison to <8 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i32 = zext <8 x i16> poison to <8 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i64 = sext <8 x i16> poison to <8 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i64 = zext <8 x i16> poison to <8 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i32i64 = sext <8 x i32> poison to <8 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z8i32i64 = zext <8 x i32> poison to <8 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i16 = sext <16 x i8> poison to <16 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i16 = zext <16 x i8> poison to <16 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i32 = sext <16 x i8> poison to <16 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i32 = zext <16 x i8> poison to <16 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i64 = sext <16 x i8> poison to <16 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i64 = zext <16 x i8> poison to <16 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i32 = sext <16 x i16> poison to <16 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i32 = zext <16 x i16> poison to <16 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i64 = sext <16 x i16> poison to <16 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:12 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i64 = zext <16 x i16> poison to <16 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %s16i32i64 = sext <16 x i32> poison to <16 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %z16i32i64 = zext <16 x i32> poison to <16 x i64> ; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; SVE128-NO-NEON-LABEL: 'ext' -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r0 = sext i1 undef to i8 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r1 = zext i1 undef to i8 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r2 = sext i1 undef to i16 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r3 = zext i1 undef to i16 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r4 = sext i1 undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r5 = zext i1 undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r6 = sext i1 undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r7 = zext i1 undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r9 = sext i8 undef to i16 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r10 = zext i8 undef to i16 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r11 = sext i8 undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r12 = zext i8 undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r13 = sext i8 undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r14 = zext i8 undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r17 = sext i16 undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r18 = zext i16 undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r19 = sext i16 undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r20 = zext i16 undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r24 = sext i32 undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r25 = zext i32 undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s2i8i16 = sext <2 x i8> undef to <2 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z2i8i16 = zext <2 x i8> undef to <2 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s2i8i32 = sext <2 x i8> undef to <2 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z2i8i32 = zext <2 x i8> undef to <2 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = sext <2 x i8> undef to <2 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z2i8i64 = zext <2 x i8> undef to <2 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s2i16i32 = sext <2 x i16> undef to <2 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z2i16i32 = zext <2 x i16> undef to <2 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = sext <2 x i16> undef to <2 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z2i16i64 = zext <2 x i16> undef to <2 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = sext <2 x i32> undef to <2 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> undef to <2 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> undef to <4 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> undef to <4 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i64 = sext <4 x i8> undef to <4 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i8i64 = zext <4 x i8> undef to <4 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> undef to <4 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z4i16i32 = zext <4 x i16> undef to <4 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i16i64 = sext <4 x i16> undef to <4 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i16i64 = zext <4 x i16> undef to <4 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i32i64 = sext <4 x i32> undef to <4 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i32i64 = zext <4 x i32> undef to <4 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = sext <8 x i8> undef to <8 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z8i8i16 = zext <8 x i8> undef to <8 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i32 = sext <8 x i8> undef to <8 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i32 = zext <8 x i8> undef to <8 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i64 = sext <8 x i8> undef to <8 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i64 = zext <8 x i8> undef to <8 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i32 = sext <8 x i16> undef to <8 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i32 = zext <8 x i16> undef to <8 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i64 = sext <8 x i16> undef to <8 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i64 = zext <8 x i16> undef to <8 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i32i64 = sext <8 x i32> undef to <8 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z8i32i64 = zext <8 x i32> undef to <8 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i16 = sext <16 x i8> undef to <16 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i16 = zext <16 x i8> undef to <16 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i32 = sext <16 x i8> undef to <16 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i32 = zext <16 x i8> undef to <16 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i64 = sext <16 x i8> undef to <16 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i64 = zext <16 x i8> undef to <16 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i32 = sext <16 x i16> undef to <16 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i32 = zext <16 x i16> undef to <16 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i64 = sext <16 x i16> undef to <16 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i64 = zext <16 x i16> undef to <16 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %s16i32i64 = sext <16 x i32> undef to <16 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %z16i32i64 = zext <16 x i32> undef to <16 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r0 = sext i1 poison to i8 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r1 = zext i1 poison to i8 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r2 = sext i1 poison to i16 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r3 = zext i1 poison to i16 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r4 = sext i1 poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r5 = zext i1 poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r6 = sext i1 poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r7 = zext i1 poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r9 = sext i8 poison to i16 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r10 = zext i8 poison to i16 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r11 = sext i8 poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r12 = zext i8 poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r13 = sext i8 poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r14 = zext i8 poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r17 = sext i16 poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r18 = zext i16 poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r19 = sext i16 poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r20 = zext i16 poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r24 = sext i32 poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r25 = zext i32 poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s2i8i16 = sext <2 x i8> poison to <2 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z2i8i16 = zext <2 x i8> poison to <2 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s2i8i32 = sext <2 x i8> poison to <2 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z2i8i32 = zext <2 x i8> poison to <2 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = sext <2 x i8> poison to <2 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z2i8i64 = zext <2 x i8> poison to <2 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s2i16i32 = sext <2 x i16> poison to <2 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z2i16i32 = zext <2 x i16> poison to <2 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = sext <2 x i16> poison to <2 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z2i16i64 = zext <2 x i16> poison to <2 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = sext <2 x i32> poison to <2 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> poison to <2 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> poison to <4 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> poison to <4 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> poison to <4 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> poison to <4 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i64 = sext <4 x i8> poison to <4 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i8i64 = zext <4 x i8> poison to <4 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> poison to <4 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z4i16i32 = zext <4 x i16> poison to <4 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i16i64 = sext <4 x i16> poison to <4 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i16i64 = zext <4 x i16> poison to <4 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i32i64 = sext <4 x i32> poison to <4 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z4i32i64 = zext <4 x i32> poison to <4 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = sext <8 x i8> poison to <8 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %z8i8i16 = zext <8 x i8> poison to <8 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i32 = sext <8 x i8> poison to <8 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i32 = zext <8 x i8> poison to <8 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i64 = sext <8 x i8> poison to <8 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i64 = zext <8 x i8> poison to <8 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i32 = sext <8 x i16> poison to <8 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i32 = zext <8 x i16> poison to <8 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i64 = sext <8 x i16> poison to <8 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i64 = zext <8 x i16> poison to <8 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i32i64 = sext <8 x i32> poison to <8 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z8i32i64 = zext <8 x i32> poison to <8 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i16 = sext <16 x i8> poison to <16 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i16 = zext <16 x i8> poison to <16 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i32 = sext <16 x i8> poison to <16 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i32 = zext <16 x i8> poison to <16 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i64 = sext <16 x i8> poison to <16 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i64 = zext <16 x i8> poison to <16 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i32 = sext <16 x i16> poison to <16 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i32 = zext <16 x i16> poison to <16 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i64 = sext <16 x i16> poison to <16 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i64 = zext <16 x i16> poison to <16 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %s16i32i64 = sext <16 x i32> poison to <16 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %z16i32i64 = zext <16 x i32> poison to <16 x i64> ; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; FIXED-MIN-256-LABEL: 'ext' -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r0 = sext i1 undef to i8 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r1 = zext i1 undef to i8 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r2 = sext i1 undef to i16 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r3 = zext i1 undef to i16 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r4 = sext i1 undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r5 = zext i1 undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r6 = sext i1 undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r7 = zext i1 undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r9 = sext i8 undef to i16 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r10 = zext i8 undef to i16 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r11 = sext i8 undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r12 = zext i8 undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r13 = sext i8 undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r14 = zext i8 undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r17 = sext i16 undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r18 = zext i16 undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r19 = sext i16 undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r20 = zext i16 undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r24 = sext i32 undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r25 = zext i32 undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i8i16 = sext <2 x i8> undef to <2 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i8i16 = zext <2 x i8> undef to <2 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i8i32 = sext <2 x i8> undef to <2 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i8i32 = zext <2 x i8> undef to <2 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = sext <2 x i8> undef to <2 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i8i64 = zext <2 x i8> undef to <2 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i16i32 = sext <2 x i16> undef to <2 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i16i32 = zext <2 x i16> undef to <2 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = sext <2 x i16> undef to <2 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i16i64 = zext <2 x i16> undef to <2 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = sext <2 x i32> undef to <2 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> undef to <2 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> undef to <4 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> undef to <4 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i8i64 = sext <4 x i8> undef to <4 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i8i64 = zext <4 x i8> undef to <4 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> undef to <4 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i16i32 = zext <4 x i16> undef to <4 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i16i64 = sext <4 x i16> undef to <4 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i16i64 = zext <4 x i16> undef to <4 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i32i64 = sext <4 x i32> undef to <4 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i32i64 = zext <4 x i32> undef to <4 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = sext <8 x i8> undef to <8 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z8i8i16 = zext <8 x i8> undef to <8 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s8i8i32 = sext <8 x i8> undef to <8 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z8i8i32 = zext <8 x i8> undef to <8 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i64 = sext <8 x i8> undef to <8 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i64 = zext <8 x i8> undef to <8 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s8i16i32 = sext <8 x i16> undef to <8 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z8i16i32 = zext <8 x i16> undef to <8 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i64 = sext <8 x i16> undef to <8 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i64 = zext <8 x i16> undef to <8 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i32i64 = sext <8 x i32> undef to <8 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i32i64 = zext <8 x i32> undef to <8 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s16i8i16 = sext <16 x i8> undef to <16 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z16i8i16 = zext <16 x i8> undef to <16 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i32 = sext <16 x i8> undef to <16 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i32 = zext <16 x i8> undef to <16 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i64 = sext <16 x i8> undef to <16 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i64 = zext <16 x i8> undef to <16 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i32 = sext <16 x i16> undef to <16 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i32 = zext <16 x i16> undef to <16 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i64 = sext <16 x i16> undef to <16 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i64 = zext <16 x i16> undef to <16 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i32i64 = sext <16 x i32> undef to <16 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i32i64 = zext <16 x i32> undef to <16 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r0 = sext i1 poison to i8 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r1 = zext i1 poison to i8 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r2 = sext i1 poison to i16 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r3 = zext i1 poison to i16 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r4 = sext i1 poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r5 = zext i1 poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r6 = sext i1 poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r7 = zext i1 poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r9 = sext i8 poison to i16 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r10 = zext i8 poison to i16 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r11 = sext i8 poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r12 = zext i8 poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r13 = sext i8 poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r14 = zext i8 poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r17 = sext i16 poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r18 = zext i16 poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r19 = sext i16 poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r20 = zext i16 poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r24 = sext i32 poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r25 = zext i32 poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i8i16 = sext <2 x i8> poison to <2 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i8i16 = zext <2 x i8> poison to <2 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i8i32 = sext <2 x i8> poison to <2 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i8i32 = zext <2 x i8> poison to <2 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = sext <2 x i8> poison to <2 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i8i64 = zext <2 x i8> poison to <2 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i16i32 = sext <2 x i16> poison to <2 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i16i32 = zext <2 x i16> poison to <2 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = sext <2 x i16> poison to <2 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i16i64 = zext <2 x i16> poison to <2 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = sext <2 x i32> poison to <2 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> poison to <2 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> poison to <4 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> poison to <4 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> poison to <4 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> poison to <4 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i8i64 = sext <4 x i8> poison to <4 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i8i64 = zext <4 x i8> poison to <4 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> poison to <4 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i16i32 = zext <4 x i16> poison to <4 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i16i64 = sext <4 x i16> poison to <4 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i16i64 = zext <4 x i16> poison to <4 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i32i64 = sext <4 x i32> poison to <4 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z4i32i64 = zext <4 x i32> poison to <4 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = sext <8 x i8> poison to <8 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z8i8i16 = zext <8 x i8> poison to <8 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s8i8i32 = sext <8 x i8> poison to <8 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z8i8i32 = zext <8 x i8> poison to <8 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i64 = sext <8 x i8> poison to <8 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i8i64 = zext <8 x i8> poison to <8 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s8i16i32 = sext <8 x i16> poison to <8 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z8i16i32 = zext <8 x i16> poison to <8 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i64 = sext <8 x i16> poison to <8 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i16i64 = zext <8 x i16> poison to <8 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i32i64 = sext <8 x i32> poison to <8 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z8i32i64 = zext <8 x i32> poison to <8 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s16i8i16 = sext <16 x i8> poison to <16 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %z16i8i16 = zext <16 x i8> poison to <16 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i32 = sext <16 x i8> poison to <16 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i32 = zext <16 x i8> poison to <16 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i64 = sext <16 x i8> poison to <16 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i8i64 = zext <16 x i8> poison to <16 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i32 = sext <16 x i16> poison to <16 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i32 = zext <16 x i16> poison to <16 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i64 = sext <16 x i16> poison to <16 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i16i64 = zext <16 x i16> poison to <16 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i32i64 = sext <16 x i32> poison to <16 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %z16i32i64 = zext <16 x i32> poison to <16 x i64> ; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; FIXED-MIN-2048-LABEL: 'ext' -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r0 = sext i1 undef to i8 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r1 = zext i1 undef to i8 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r2 = sext i1 undef to i16 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r3 = zext i1 undef to i16 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r4 = sext i1 undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r5 = zext i1 undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r6 = sext i1 undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r7 = zext i1 undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r9 = sext i8 undef to i16 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r10 = zext i8 undef to i16 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r11 = sext i8 undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r12 = zext i8 undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r13 = sext i8 undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r14 = zext i8 undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r17 = sext i16 undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r18 = zext i16 undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r19 = sext i16 undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r20 = zext i16 undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r24 = sext i32 undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r25 = zext i32 undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i8i16 = sext <2 x i8> undef to <2 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i8i16 = zext <2 x i8> undef to <2 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i8i32 = sext <2 x i8> undef to <2 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i8i32 = zext <2 x i8> undef to <2 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = sext <2 x i8> undef to <2 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i8i64 = zext <2 x i8> undef to <2 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i16i32 = sext <2 x i16> undef to <2 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i16i32 = zext <2 x i16> undef to <2 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = sext <2 x i16> undef to <2 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i16i64 = zext <2 x i16> undef to <2 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = sext <2 x i32> undef to <2 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> undef to <2 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> undef to <4 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> undef to <4 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> undef to <4 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> undef to <4 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i8i64 = sext <4 x i8> undef to <4 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i8i64 = zext <4 x i8> undef to <4 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> undef to <4 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i16i32 = zext <4 x i16> undef to <4 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i16i64 = sext <4 x i16> undef to <4 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i16i64 = zext <4 x i16> undef to <4 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i32i64 = sext <4 x i32> undef to <4 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i32i64 = zext <4 x i32> undef to <4 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = sext <8 x i8> undef to <8 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z8i8i16 = zext <8 x i8> undef to <8 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i8i32 = sext <8 x i8> undef to <8 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z8i8i32 = zext <8 x i8> undef to <8 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i8i64 = sext <8 x i8> undef to <8 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z8i8i64 = zext <8 x i8> undef to <8 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i16i32 = sext <8 x i16> undef to <8 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z8i16i32 = zext <8 x i16> undef to <8 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i16i64 = sext <8 x i16> undef to <8 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z8i16i64 = zext <8 x i16> undef to <8 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i32i64 = sext <8 x i32> undef to <8 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z8i32i64 = zext <8 x i32> undef to <8 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s16i8i16 = sext <16 x i8> undef to <16 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z16i8i16 = zext <16 x i8> undef to <16 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s16i8i32 = sext <16 x i8> undef to <16 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z16i8i32 = zext <16 x i8> undef to <16 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s16i8i64 = sext <16 x i8> undef to <16 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z16i8i64 = zext <16 x i8> undef to <16 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s16i16i32 = sext <16 x i16> undef to <16 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z16i16i32 = zext <16 x i16> undef to <16 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s16i16i64 = sext <16 x i16> undef to <16 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z16i16i64 = zext <16 x i16> undef to <16 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s16i32i64 = sext <16 x i32> undef to <16 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z16i32i64 = zext <16 x i32> undef to <16 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r0 = sext i1 poison to i8 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r1 = zext i1 poison to i8 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r2 = sext i1 poison to i16 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r3 = zext i1 poison to i16 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r4 = sext i1 poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r5 = zext i1 poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r6 = sext i1 poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r7 = zext i1 poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r9 = sext i8 poison to i16 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r10 = zext i8 poison to i16 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r11 = sext i8 poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r12 = zext i8 poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r13 = sext i8 poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r14 = zext i8 poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r17 = sext i16 poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r18 = zext i16 poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r19 = sext i16 poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r20 = zext i16 poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r24 = sext i32 poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r25 = zext i32 poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i8i16 = sext <2 x i8> poison to <2 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i8i16 = zext <2 x i8> poison to <2 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i8i32 = sext <2 x i8> poison to <2 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i8i32 = zext <2 x i8> poison to <2 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = sext <2 x i8> poison to <2 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i8i64 = zext <2 x i8> poison to <2 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i16i32 = sext <2 x i16> poison to <2 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i16i32 = zext <2 x i16> poison to <2 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = sext <2 x i16> poison to <2 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i16i64 = zext <2 x i16> poison to <2 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = sext <2 x i32> poison to <2 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z2i32i64 = zext <2 x i32> poison to <2 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i8i16 = sext <4 x i8> poison to <4 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i8i16 = zext <4 x i8> poison to <4 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = sext <4 x i8> poison to <4 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i8i32 = zext <4 x i8> poison to <4 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i8i64 = sext <4 x i8> poison to <4 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i8i64 = zext <4 x i8> poison to <4 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = sext <4 x i16> poison to <4 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i16i32 = zext <4 x i16> poison to <4 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i16i64 = sext <4 x i16> poison to <4 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i16i64 = zext <4 x i16> poison to <4 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i32i64 = sext <4 x i32> poison to <4 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z4i32i64 = zext <4 x i32> poison to <4 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = sext <8 x i8> poison to <8 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z8i8i16 = zext <8 x i8> poison to <8 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i8i32 = sext <8 x i8> poison to <8 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z8i8i32 = zext <8 x i8> poison to <8 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i8i64 = sext <8 x i8> poison to <8 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z8i8i64 = zext <8 x i8> poison to <8 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i16i32 = sext <8 x i16> poison to <8 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z8i16i32 = zext <8 x i16> poison to <8 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i16i64 = sext <8 x i16> poison to <8 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z8i16i64 = zext <8 x i16> poison to <8 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i32i64 = sext <8 x i32> poison to <8 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z8i32i64 = zext <8 x i32> poison to <8 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s16i8i16 = sext <16 x i8> poison to <16 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z16i8i16 = zext <16 x i8> poison to <16 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s16i8i32 = sext <16 x i8> poison to <16 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z16i8i32 = zext <16 x i8> poison to <16 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s16i8i64 = sext <16 x i8> poison to <16 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z16i8i64 = zext <16 x i8> poison to <16 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s16i16i32 = sext <16 x i16> poison to <16 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z16i16i32 = zext <16 x i16> poison to <16 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s16i16i64 = sext <16 x i16> poison to <16 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z16i16i64 = zext <16 x i16> poison to <16 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s16i32i64 = sext <16 x i32> poison to <16 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %z16i32i64 = zext <16 x i32> poison to <16 x i64> ; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %r0 = sext i1 undef to i8 - %r1 = zext i1 undef to i8 - %r2 = sext i1 undef to i16 - %r3 = zext i1 undef to i16 - %r4 = sext i1 undef to i32 - %r5 = zext i1 undef to i32 - %r6 = sext i1 undef to i64 - %r7 = zext i1 undef to i64 - %r9 = sext i8 undef to i16 - %r10 = zext i8 undef to i16 - %r11 = sext i8 undef to i32 - %r12 = zext i8 undef to i32 - %r13 = sext i8 undef to i64 - %r14 = zext i8 undef to i64 - %r17 = sext i16 undef to i32 - %r18 = zext i16 undef to i32 - %r19 = sext i16 undef to i64 - %r20 = zext i16 undef to i64 - %r24 = sext i32 undef to i64 - %r25 = zext i32 undef to i64 + %r0 = sext i1 poison to i8 + %r1 = zext i1 poison to i8 + %r2 = sext i1 poison to i16 + %r3 = zext i1 poison to i16 + %r4 = sext i1 poison to i32 + %r5 = zext i1 poison to i32 + %r6 = sext i1 poison to i64 + %r7 = zext i1 poison to i64 + %r9 = sext i8 poison to i16 + %r10 = zext i8 poison to i16 + %r11 = sext i8 poison to i32 + %r12 = zext i8 poison to i32 + %r13 = sext i8 poison to i64 + %r14 = zext i8 poison to i64 + %r17 = sext i16 poison to i32 + %r18 = zext i16 poison to i32 + %r19 = sext i16 poison to i64 + %r20 = zext i16 poison to i64 + %r24 = sext i32 poison to i64 + %r25 = zext i32 poison to i64 - %s2i8i16 = sext <2 x i8> undef to <2 x i16> - %z2i8i16 = zext <2 x i8> undef to <2 x i16> - %s2i8i32 = sext <2 x i8> undef to <2 x i32> - %z2i8i32 = zext <2 x i8> undef to <2 x i32> - %s2i8i64 = sext <2 x i8> undef to <2 x i64> - %z2i8i64 = zext <2 x i8> undef to <2 x i64> - %s2i16i32 = sext <2 x i16> undef to <2 x i32> - %z2i16i32 = zext <2 x i16> undef to <2 x i32> - %s2i16i64 = sext <2 x i16> undef to <2 x i64> - %z2i16i64 = zext <2 x i16> undef to <2 x i64> - %s2i32i64 = sext <2 x i32> undef to <2 x i64> - %z2i32i64 = zext <2 x i32> undef to <2 x i64> + %s2i8i16 = sext <2 x i8> poison to <2 x i16> + %z2i8i16 = zext <2 x i8> poison to <2 x i16> + %s2i8i32 = sext <2 x i8> poison to <2 x i32> + %z2i8i32 = zext <2 x i8> poison to <2 x i32> + %s2i8i64 = sext <2 x i8> poison to <2 x i64> + %z2i8i64 = zext <2 x i8> poison to <2 x i64> + %s2i16i32 = sext <2 x i16> poison to <2 x i32> + %z2i16i32 = zext <2 x i16> poison to <2 x i32> + %s2i16i64 = sext <2 x i16> poison to <2 x i64> + %z2i16i64 = zext <2 x i16> poison to <2 x i64> + %s2i32i64 = sext <2 x i32> poison to <2 x i64> + %z2i32i64 = zext <2 x i32> poison to <2 x i64> - %s4i8i16 = sext <4 x i8> undef to <4 x i16> - %z4i8i16 = zext <4 x i8> undef to <4 x i16> - %s4i8i32 = sext <4 x i8> undef to <4 x i32> - %z4i8i32 = zext <4 x i8> undef to <4 x i32> - %s4i8i64 = sext <4 x i8> undef to <4 x i64> - %z4i8i64 = zext <4 x i8> undef to <4 x i64> - %s4i16i32 = sext <4 x i16> undef to <4 x i32> - %z4i16i32 = zext <4 x i16> undef to <4 x i32> - %s4i16i64 = sext <4 x i16> undef to <4 x i64> - %z4i16i64 = zext <4 x i16> undef to <4 x i64> - %s4i32i64 = sext <4 x i32> undef to <4 x i64> - %z4i32i64 = zext <4 x i32> undef to <4 x i64> + %s4i8i16 = sext <4 x i8> poison to <4 x i16> + %z4i8i16 = zext <4 x i8> poison to <4 x i16> + %s4i8i32 = sext <4 x i8> poison to <4 x i32> + %z4i8i32 = zext <4 x i8> poison to <4 x i32> + %s4i8i64 = sext <4 x i8> poison to <4 x i64> + %z4i8i64 = zext <4 x i8> poison to <4 x i64> + %s4i16i32 = sext <4 x i16> poison to <4 x i32> + %z4i16i32 = zext <4 x i16> poison to <4 x i32> + %s4i16i64 = sext <4 x i16> poison to <4 x i64> + %z4i16i64 = zext <4 x i16> poison to <4 x i64> + %s4i32i64 = sext <4 x i32> poison to <4 x i64> + %z4i32i64 = zext <4 x i32> poison to <4 x i64> - %s8i8i16 = sext <8 x i8> undef to <8 x i16> - %z8i8i16 = zext <8 x i8> undef to <8 x i16> - %s8i8i32 = sext <8 x i8> undef to <8 x i32> - %z8i8i32 = zext <8 x i8> undef to <8 x i32> - %s8i8i64 = sext <8 x i8> undef to <8 x i64> - %z8i8i64 = zext <8 x i8> undef to <8 x i64> - %s8i16i32 = sext <8 x i16> undef to <8 x i32> - %z8i16i32 = zext <8 x i16> undef to <8 x i32> - %s8i16i64 = sext <8 x i16> undef to <8 x i64> - %z8i16i64 = zext <8 x i16> undef to <8 x i64> - %s8i32i64 = sext <8 x i32> undef to <8 x i64> - %z8i32i64 = zext <8 x i32> undef to <8 x i64> + %s8i8i16 = sext <8 x i8> poison to <8 x i16> + %z8i8i16 = zext <8 x i8> poison to <8 x i16> + %s8i8i32 = sext <8 x i8> poison to <8 x i32> + %z8i8i32 = zext <8 x i8> poison to <8 x i32> + %s8i8i64 = sext <8 x i8> poison to <8 x i64> + %z8i8i64 = zext <8 x i8> poison to <8 x i64> + %s8i16i32 = sext <8 x i16> poison to <8 x i32> + %z8i16i32 = zext <8 x i16> poison to <8 x i32> + %s8i16i64 = sext <8 x i16> poison to <8 x i64> + %z8i16i64 = zext <8 x i16> poison to <8 x i64> + %s8i32i64 = sext <8 x i32> poison to <8 x i64> + %z8i32i64 = zext <8 x i32> poison to <8 x i64> - %s16i8i16 = sext <16 x i8> undef to <16 x i16> - %z16i8i16 = zext <16 x i8> undef to <16 x i16> - %s16i8i32 = sext <16 x i8> undef to <16 x i32> - %z16i8i32 = zext <16 x i8> undef to <16 x i32> - %s16i8i64 = sext <16 x i8> undef to <16 x i64> - %z16i8i64 = zext <16 x i8> undef to <16 x i64> - %s16i16i32 = sext <16 x i16> undef to <16 x i32> - %z16i16i32 = zext <16 x i16> undef to <16 x i32> - %s16i16i64 = sext <16 x i16> undef to <16 x i64> - %z16i16i64 = zext <16 x i16> undef to <16 x i64> - %s16i32i64 = sext <16 x i32> undef to <16 x i64> - %z16i32i64 = zext <16 x i32> undef to <16 x i64> + %s16i8i16 = sext <16 x i8> poison to <16 x i16> + %z16i8i16 = zext <16 x i8> poison to <16 x i16> + %s16i8i32 = sext <16 x i8> poison to <16 x i32> + %z16i8i32 = zext <16 x i8> poison to <16 x i32> + %s16i8i64 = sext <16 x i8> poison to <16 x i64> + %z16i8i64 = zext <16 x i8> poison to <16 x i64> + %s16i16i32 = sext <16 x i16> poison to <16 x i32> + %z16i16i32 = zext <16 x i16> poison to <16 x i32> + %s16i16i64 = sext <16 x i16> poison to <16 x i64> + %z16i16i64 = zext <16 x i16> poison to <16 x i64> + %s16i32i64 = sext <16 x i32> poison to <16 x i64> + %z16i32i64 = zext <16 x i32> poison to <16 x i64> ret void } define void @trunc() { ; CHECK-SVE-LABEL: 'trunc' -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r8 = trunc i8 undef to i1 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r15 = trunc i16 undef to i1 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r16 = trunc i16 undef to i8 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r21 = trunc i32 undef to i1 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r22 = trunc i32 undef to i8 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r23 = trunc i32 undef to i16 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r26 = trunc i64 undef to i1 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r27 = trunc i64 undef to i8 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r28 = trunc i64 undef to i16 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r29 = trunc i64 undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %s2i8i16 = trunc <2 x i16> undef to <2 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %s2i8i32 = trunc <2 x i32> undef to <2 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = trunc <2 x i64> undef to <2 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %s2i16i32 = trunc <2 x i32> undef to <2 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = trunc <2 x i64> undef to <2 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = trunc <2 x i64> undef to <2 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %s4i8i16 = trunc <4 x i16> undef to <4 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = trunc <4 x i32> undef to <4 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i64 = trunc <4 x i64> undef to <4 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = trunc <4 x i32> undef to <4 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i16i64 = trunc <4 x i64> undef to <4 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i32i64 = trunc <4 x i64> undef to <4 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = trunc <8 x i16> undef to <8 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i32 = trunc <8 x i32> undef to <8 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i64 = trunc <8 x i64> undef to <8 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s8i16i32 = trunc <8 x i32> undef to <8 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i64 = trunc <8 x i64> undef to <8 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i32i64 = trunc <8 x i64> undef to <8 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s16i8i16 = trunc <16 x i16> undef to <16 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i32 = trunc <16 x i32> undef to <16 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i64 = trunc <16 x i64> undef to <16 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i32 = trunc <16 x i32> undef to <16 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i64 = trunc <16 x i64> undef to <16 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i32i64 = trunc <16 x i64> undef to <16 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r8 = trunc i8 poison to i1 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r15 = trunc i16 poison to i1 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r16 = trunc i16 poison to i8 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r21 = trunc i32 poison to i1 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r22 = trunc i32 poison to i8 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r23 = trunc i32 poison to i16 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r26 = trunc i64 poison to i1 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r27 = trunc i64 poison to i8 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r28 = trunc i64 poison to i16 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %r29 = trunc i64 poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %s2i8i16 = trunc <2 x i16> poison to <2 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %s2i8i32 = trunc <2 x i32> poison to <2 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = trunc <2 x i64> poison to <2 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %s2i16i32 = trunc <2 x i32> poison to <2 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = trunc <2 x i64> poison to <2 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = trunc <2 x i64> poison to <2 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 0 for: %s4i8i16 = trunc <4 x i16> poison to <4 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = trunc <4 x i32> poison to <4 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s4i8i64 = trunc <4 x i64> poison to <4 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = trunc <4 x i32> poison to <4 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s4i16i64 = trunc <4 x i64> poison to <4 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s4i32i64 = trunc <4 x i64> poison to <4 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = trunc <8 x i16> poison to <8 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i32 = trunc <8 x i32> poison to <8 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s8i8i64 = trunc <8 x i64> poison to <8 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s8i16i32 = trunc <8 x i32> poison to <8 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s8i16i64 = trunc <8 x i64> poison to <8 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s8i32i64 = trunc <8 x i64> poison to <8 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %s16i8i16 = trunc <16 x i16> poison to <16 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i32 = trunc <16 x i32> poison to <16 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %s16i8i64 = trunc <16 x i64> poison to <16 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i32 = trunc <16 x i32> poison to <16 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %s16i16i64 = trunc <16 x i64> poison to <16 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %s16i32i64 = trunc <16 x i64> poison to <16 x i32> ; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; SVE128-NO-NEON-LABEL: 'trunc' -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r8 = trunc i8 undef to i1 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r15 = trunc i16 undef to i1 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r16 = trunc i16 undef to i8 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r21 = trunc i32 undef to i1 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r22 = trunc i32 undef to i8 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r23 = trunc i32 undef to i16 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r26 = trunc i64 undef to i1 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r27 = trunc i64 undef to i8 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r28 = trunc i64 undef to i16 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r29 = trunc i64 undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s2i8i16 = trunc <2 x i16> undef to <2 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s2i8i32 = trunc <2 x i32> undef to <2 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s2i8i64 = trunc <2 x i64> undef to <2 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s2i16i32 = trunc <2 x i32> undef to <2 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s2i16i64 = trunc <2 x i64> undef to <2 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s2i32i64 = trunc <2 x i64> undef to <2 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s4i8i16 = trunc <4 x i16> undef to <4 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s4i8i32 = trunc <4 x i32> undef to <4 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s4i8i64 = trunc <4 x i64> undef to <4 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s4i16i32 = trunc <4 x i32> undef to <4 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s4i16i64 = trunc <4 x i64> undef to <4 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s4i32i64 = trunc <4 x i64> undef to <4 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s8i8i16 = trunc <8 x i16> undef to <8 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s8i8i32 = trunc <8 x i32> undef to <8 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s8i8i64 = trunc <8 x i64> undef to <8 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s8i16i32 = trunc <8 x i32> undef to <8 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s8i16i64 = trunc <8 x i64> undef to <8 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s8i32i64 = trunc <8 x i64> undef to <8 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s16i8i16 = trunc <16 x i16> undef to <16 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s16i8i32 = trunc <16 x i32> undef to <16 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s16i8i64 = trunc <16 x i64> undef to <16 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s16i16i32 = trunc <16 x i32> undef to <16 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s16i16i64 = trunc <16 x i64> undef to <16 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s16i32i64 = trunc <16 x i64> undef to <16 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r8 = trunc i8 poison to i1 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r15 = trunc i16 poison to i1 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r16 = trunc i16 poison to i8 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r21 = trunc i32 poison to i1 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r22 = trunc i32 poison to i8 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r23 = trunc i32 poison to i16 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r26 = trunc i64 poison to i1 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r27 = trunc i64 poison to i8 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r28 = trunc i64 poison to i16 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %r29 = trunc i64 poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s2i8i16 = trunc <2 x i16> poison to <2 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s2i8i32 = trunc <2 x i32> poison to <2 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s2i8i64 = trunc <2 x i64> poison to <2 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s2i16i32 = trunc <2 x i32> poison to <2 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s2i16i64 = trunc <2 x i64> poison to <2 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s2i32i64 = trunc <2 x i64> poison to <2 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s4i8i16 = trunc <4 x i16> poison to <4 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s4i8i32 = trunc <4 x i32> poison to <4 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s4i8i64 = trunc <4 x i64> poison to <4 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s4i16i32 = trunc <4 x i32> poison to <4 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s4i16i64 = trunc <4 x i64> poison to <4 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s4i32i64 = trunc <4 x i64> poison to <4 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s8i8i16 = trunc <8 x i16> poison to <8 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s8i8i32 = trunc <8 x i32> poison to <8 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s8i8i64 = trunc <8 x i64> poison to <8 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s8i16i32 = trunc <8 x i32> poison to <8 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s8i16i64 = trunc <8 x i64> poison to <8 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s8i32i64 = trunc <8 x i64> poison to <8 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s16i8i16 = trunc <16 x i16> poison to <16 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s16i8i32 = trunc <16 x i32> poison to <16 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s16i8i64 = trunc <16 x i64> poison to <16 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s16i16i32 = trunc <16 x i32> poison to <16 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s16i16i64 = trunc <16 x i64> poison to <16 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 0 for: %s16i32i64 = trunc <16 x i64> poison to <16 x i32> ; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; FIXED-MIN-256-LABEL: 'trunc' -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r8 = trunc i8 undef to i1 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r15 = trunc i16 undef to i1 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r16 = trunc i16 undef to i8 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r21 = trunc i32 undef to i1 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r22 = trunc i32 undef to i8 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r23 = trunc i32 undef to i16 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r26 = trunc i64 undef to i1 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r27 = trunc i64 undef to i8 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r28 = trunc i64 undef to i16 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r29 = trunc i64 undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s2i8i16 = trunc <2 x i16> undef to <2 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s2i8i32 = trunc <2 x i32> undef to <2 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = trunc <2 x i64> undef to <2 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s2i16i32 = trunc <2 x i32> undef to <2 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = trunc <2 x i64> undef to <2 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = trunc <2 x i64> undef to <2 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s4i8i16 = trunc <4 x i16> undef to <4 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = trunc <4 x i32> undef to <4 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s4i8i64 = trunc <4 x i64> undef to <4 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = trunc <4 x i32> undef to <4 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s4i16i64 = trunc <4 x i64> undef to <4 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s4i32i64 = trunc <4 x i64> undef to <4 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = trunc <8 x i16> undef to <8 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s8i8i32 = trunc <8 x i32> undef to <8 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s8i8i64 = trunc <8 x i64> undef to <8 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s8i16i32 = trunc <8 x i32> undef to <8 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s8i16i64 = trunc <8 x i64> undef to <8 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s8i32i64 = trunc <8 x i64> undef to <8 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s16i8i16 = trunc <16 x i16> undef to <16 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s16i8i32 = trunc <16 x i32> undef to <16 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s16i8i64 = trunc <16 x i64> undef to <16 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s16i16i32 = trunc <16 x i32> undef to <16 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s16i16i64 = trunc <16 x i64> undef to <16 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s16i32i64 = trunc <16 x i64> undef to <16 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r8 = trunc i8 poison to i1 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r15 = trunc i16 poison to i1 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r16 = trunc i16 poison to i8 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r21 = trunc i32 poison to i1 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r22 = trunc i32 poison to i8 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r23 = trunc i32 poison to i16 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r26 = trunc i64 poison to i1 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r27 = trunc i64 poison to i8 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r28 = trunc i64 poison to i16 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %r29 = trunc i64 poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s2i8i16 = trunc <2 x i16> poison to <2 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s2i8i32 = trunc <2 x i32> poison to <2 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = trunc <2 x i64> poison to <2 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s2i16i32 = trunc <2 x i32> poison to <2 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = trunc <2 x i64> poison to <2 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = trunc <2 x i64> poison to <2 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s4i8i16 = trunc <4 x i16> poison to <4 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = trunc <4 x i32> poison to <4 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s4i8i64 = trunc <4 x i64> poison to <4 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = trunc <4 x i32> poison to <4 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s4i16i64 = trunc <4 x i64> poison to <4 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s4i32i64 = trunc <4 x i64> poison to <4 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = trunc <8 x i16> poison to <8 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s8i8i32 = trunc <8 x i32> poison to <8 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s8i8i64 = trunc <8 x i64> poison to <8 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s8i16i32 = trunc <8 x i32> poison to <8 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s8i16i64 = trunc <8 x i64> poison to <8 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s8i32i64 = trunc <8 x i64> poison to <8 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s16i8i16 = trunc <16 x i16> poison to <16 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s16i8i32 = trunc <16 x i32> poison to <16 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s16i8i64 = trunc <16 x i64> poison to <16 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s16i16i32 = trunc <16 x i32> poison to <16 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s16i16i64 = trunc <16 x i64> poison to <16 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 0 for: %s16i32i64 = trunc <16 x i64> poison to <16 x i32> ; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; FIXED-MIN-2048-LABEL: 'trunc' -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r8 = trunc i8 undef to i1 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r15 = trunc i16 undef to i1 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r16 = trunc i16 undef to i8 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r21 = trunc i32 undef to i1 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r22 = trunc i32 undef to i8 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r23 = trunc i32 undef to i16 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r26 = trunc i64 undef to i1 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r27 = trunc i64 undef to i8 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r28 = trunc i64 undef to i16 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r29 = trunc i64 undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s2i8i16 = trunc <2 x i16> undef to <2 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s2i8i32 = trunc <2 x i32> undef to <2 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = trunc <2 x i64> undef to <2 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s2i16i32 = trunc <2 x i32> undef to <2 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = trunc <2 x i64> undef to <2 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = trunc <2 x i64> undef to <2 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s4i8i16 = trunc <4 x i16> undef to <4 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = trunc <4 x i32> undef to <4 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s4i8i64 = trunc <4 x i64> undef to <4 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = trunc <4 x i32> undef to <4 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s4i16i64 = trunc <4 x i64> undef to <4 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s4i32i64 = trunc <4 x i64> undef to <4 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = trunc <8 x i16> undef to <8 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s8i8i32 = trunc <8 x i32> undef to <8 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s8i8i64 = trunc <8 x i64> undef to <8 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s8i16i32 = trunc <8 x i32> undef to <8 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s8i16i64 = trunc <8 x i64> undef to <8 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s8i32i64 = trunc <8 x i64> undef to <8 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s16i8i16 = trunc <16 x i16> undef to <16 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s16i8i32 = trunc <16 x i32> undef to <16 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s16i8i64 = trunc <16 x i64> undef to <16 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s16i16i32 = trunc <16 x i32> undef to <16 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s16i16i64 = trunc <16 x i64> undef to <16 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s16i32i64 = trunc <16 x i64> undef to <16 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r8 = trunc i8 poison to i1 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r15 = trunc i16 poison to i1 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r16 = trunc i16 poison to i8 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r21 = trunc i32 poison to i1 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r22 = trunc i32 poison to i8 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r23 = trunc i32 poison to i16 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r26 = trunc i64 poison to i1 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r27 = trunc i64 poison to i8 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r28 = trunc i64 poison to i16 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %r29 = trunc i64 poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s2i8i16 = trunc <2 x i16> poison to <2 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s2i8i32 = trunc <2 x i32> poison to <2 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i8i64 = trunc <2 x i64> poison to <2 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s2i16i32 = trunc <2 x i32> poison to <2 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i16i64 = trunc <2 x i64> poison to <2 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s2i32i64 = trunc <2 x i64> poison to <2 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s4i8i16 = trunc <4 x i16> poison to <4 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i8i32 = trunc <4 x i32> poison to <4 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s4i8i64 = trunc <4 x i64> poison to <4 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s4i16i32 = trunc <4 x i32> poison to <4 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s4i16i64 = trunc <4 x i64> poison to <4 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s4i32i64 = trunc <4 x i64> poison to <4 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %s8i8i16 = trunc <8 x i16> poison to <8 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s8i8i32 = trunc <8 x i32> poison to <8 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s8i8i64 = trunc <8 x i64> poison to <8 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s8i16i32 = trunc <8 x i32> poison to <8 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s8i16i64 = trunc <8 x i64> poison to <8 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s8i32i64 = trunc <8 x i64> poison to <8 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s16i8i16 = trunc <16 x i16> poison to <16 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s16i8i32 = trunc <16 x i32> poison to <16 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s16i8i64 = trunc <16 x i64> poison to <16 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s16i16i32 = trunc <16 x i32> poison to <16 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s16i16i64 = trunc <16 x i64> poison to <16 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 0 for: %s16i32i64 = trunc <16 x i64> poison to <16 x i32> ; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %r8 = trunc i8 undef to i1 - %r15 = trunc i16 undef to i1 - %r16 = trunc i16 undef to i8 - %r21 = trunc i32 undef to i1 - %r22 = trunc i32 undef to i8 - %r23 = trunc i32 undef to i16 - %r26 = trunc i64 undef to i1 - %r27 = trunc i64 undef to i8 - %r28 = trunc i64 undef to i16 - %r29 = trunc i64 undef to i32 + %r8 = trunc i8 poison to i1 + %r15 = trunc i16 poison to i1 + %r16 = trunc i16 poison to i8 + %r21 = trunc i32 poison to i1 + %r22 = trunc i32 poison to i8 + %r23 = trunc i32 poison to i16 + %r26 = trunc i64 poison to i1 + %r27 = trunc i64 poison to i8 + %r28 = trunc i64 poison to i16 + %r29 = trunc i64 poison to i32 - %s2i8i16 = trunc <2 x i16> undef to <2 x i8> - %s2i8i32 = trunc <2 x i32> undef to <2 x i8> - %s2i8i64 = trunc <2 x i64> undef to <2 x i8> - %s2i16i32 = trunc <2 x i32> undef to <2 x i16> - %s2i16i64 = trunc <2 x i64> undef to <2 x i16> - %s2i32i64 = trunc <2 x i64> undef to <2 x i32> + %s2i8i16 = trunc <2 x i16> poison to <2 x i8> + %s2i8i32 = trunc <2 x i32> poison to <2 x i8> + %s2i8i64 = trunc <2 x i64> poison to <2 x i8> + %s2i16i32 = trunc <2 x i32> poison to <2 x i16> + %s2i16i64 = trunc <2 x i64> poison to <2 x i16> + %s2i32i64 = trunc <2 x i64> poison to <2 x i32> - %s4i8i16 = trunc <4 x i16> undef to <4 x i8> - %s4i8i32 = trunc <4 x i32> undef to <4 x i8> - %s4i8i64 = trunc <4 x i64> undef to <4 x i8> - %s4i16i32 = trunc <4 x i32> undef to <4 x i16> - %s4i16i64 = trunc <4 x i64> undef to <4 x i16> - %s4i32i64 = trunc <4 x i64> undef to <4 x i32> + %s4i8i16 = trunc <4 x i16> poison to <4 x i8> + %s4i8i32 = trunc <4 x i32> poison to <4 x i8> + %s4i8i64 = trunc <4 x i64> poison to <4 x i8> + %s4i16i32 = trunc <4 x i32> poison to <4 x i16> + %s4i16i64 = trunc <4 x i64> poison to <4 x i16> + %s4i32i64 = trunc <4 x i64> poison to <4 x i32> - %s8i8i16 = trunc <8 x i16> undef to <8 x i8> - %s8i8i32 = trunc <8 x i32> undef to <8 x i8> - %s8i8i64 = trunc <8 x i64> undef to <8 x i8> - %s8i16i32 = trunc <8 x i32> undef to <8 x i16> - %s8i16i64 = trunc <8 x i64> undef to <8 x i16> - %s8i32i64 = trunc <8 x i64> undef to <8 x i32> + %s8i8i16 = trunc <8 x i16> poison to <8 x i8> + %s8i8i32 = trunc <8 x i32> poison to <8 x i8> + %s8i8i64 = trunc <8 x i64> poison to <8 x i8> + %s8i16i32 = trunc <8 x i32> poison to <8 x i16> + %s8i16i64 = trunc <8 x i64> poison to <8 x i16> + %s8i32i64 = trunc <8 x i64> poison to <8 x i32> - %s16i8i16 = trunc <16 x i16> undef to <16 x i8> - %s16i8i32 = trunc <16 x i32> undef to <16 x i8> - %s16i8i64 = trunc <16 x i64> undef to <16 x i8> - %s16i16i32 = trunc <16 x i32> undef to <16 x i16> - %s16i16i64 = trunc <16 x i64> undef to <16 x i16> - %s16i32i64 = trunc <16 x i64> undef to <16 x i32> + %s16i8i16 = trunc <16 x i16> poison to <16 x i8> + %s16i8i32 = trunc <16 x i32> poison to <16 x i8> + %s16i8i64 = trunc <16 x i64> poison to <16 x i8> + %s16i16i32 = trunc <16 x i32> poison to <16 x i16> + %s16i16i64 = trunc <16 x i64> poison to <16 x i16> + %s16i32i64 = trunc <16 x i64> poison to <16 x i32> ret void } define i32 @casts_no_users() { ; CHECK-SVE-LABEL: 'casts_no_users' -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui float undef to i1 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi float undef to i1 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui float undef to i8 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi float undef to i8 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui float undef to i16 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi float undef to i16 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui float undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi float undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui float undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi float undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r40 = fptoui double undef to i1 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r41 = fptosi double undef to i1 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r42 = fptoui double undef to i8 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r43 = fptosi double undef to i8 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r44 = fptoui double undef to i16 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r45 = fptosi double undef to i16 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r46 = fptoui double undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r47 = fptosi double undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r48 = fptoui double undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r49 = fptosi double undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r50 = sitofp i1 undef to float -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r51 = uitofp i1 undef to float -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r52 = sitofp i1 undef to double -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r53 = uitofp i1 undef to double -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r54 = sitofp i8 undef to float -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r55 = uitofp i8 undef to float -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r56 = sitofp i8 undef to double -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r57 = uitofp i8 undef to double -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r58 = sitofp i16 undef to float -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r59 = uitofp i16 undef to float -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r60 = sitofp i16 undef to double -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r61 = uitofp i16 undef to double -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r62 = sitofp i32 undef to float -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r63 = uitofp i32 undef to float -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r64 = sitofp i32 undef to double -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r65 = uitofp i32 undef to double -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r66 = sitofp i64 undef to float -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r67 = uitofp i64 undef to float -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r68 = sitofp i64 undef to double -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r69 = uitofp i64 undef to double -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r80 = fptrunc double undef to float -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r81 = fptrunc <2 x double> undef to <2 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r82 = fptrunc <4 x double> undef to <4 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r83 = fptrunc <8 x double> undef to <8 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r84 = fptrunc <16 x double> undef to <16 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r85 = fpext float undef to double -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r86 = fpext <2 x float> undef to <2 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r87 = fpext <4 x float> undef to <4 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r88 = fpext <8 x float> undef to <8 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r89 = fpext <16 x float> undef to <16 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x float> undef to <2 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x float> undef to <2 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x float> undef to <2 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x float> undef to <2 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x float> undef to <2 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x float> undef to <2 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x float> undef to <2 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x float> undef to <2 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x float> undef to <2 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x float> undef to <2 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r100 = fptoui <2 x double> undef to <2 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r101 = fptosi <2 x double> undef to <2 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r102 = fptoui <2 x double> undef to <2 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r103 = fptosi <2 x double> undef to <2 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r104 = fptoui <2 x double> undef to <2 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r105 = fptosi <2 x double> undef to <2 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r106 = fptoui <2 x double> undef to <2 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r107 = fptosi <2 x double> undef to <2 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r108 = fptoui <2 x double> undef to <2 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r109 = fptosi <2 x double> undef to <2 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r110 = fptoui <4 x float> undef to <4 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r111 = fptosi <4 x float> undef to <4 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r112 = fptoui <4 x float> undef to <4 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r113 = fptosi <4 x float> undef to <4 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r114 = fptoui <4 x float> undef to <4 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r115 = fptosi <4 x float> undef to <4 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r116 = fptoui <4 x float> undef to <4 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r117 = fptosi <4 x float> undef to <4 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x float> undef to <4 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x float> undef to <4 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r120 = fptoui <4 x double> undef to <4 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r121 = fptosi <4 x double> undef to <4 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r122 = fptoui <4 x double> undef to <4 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r123 = fptosi <4 x double> undef to <4 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r124 = fptoui <4 x double> undef to <4 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r125 = fptosi <4 x double> undef to <4 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r126 = fptoui <4 x double> undef to <4 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r127 = fptosi <4 x double> undef to <4 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r128 = fptoui <4 x double> undef to <4 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r129 = fptosi <4 x double> undef to <4 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:41 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x float> undef to <8 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:41 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x float> undef to <8 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x float> undef to <8 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x float> undef to <8 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r134 = fptoui <8 x float> undef to <8 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r135 = fptosi <8 x float> undef to <8 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x float> undef to <8 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x float> undef to <8 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x float> undef to <8 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x float> undef to <8 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r140 = fptoui <8 x double> undef to <8 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r141 = fptosi <8 x double> undef to <8 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r142 = fptoui <8 x double> undef to <8 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r143 = fptosi <8 x double> undef to <8 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r144 = fptoui <8 x double> undef to <8 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r145 = fptosi <8 x double> undef to <8 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r146 = fptoui <8 x double> undef to <8 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r147 = fptosi <8 x double> undef to <8 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r148 = fptoui <8 x double> undef to <8 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r149 = fptosi <8 x double> undef to <8 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:83 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x float> undef to <16 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:83 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x float> undef to <16 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x float> undef to <16 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x float> undef to <16 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x float> undef to <16 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x float> undef to <16 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x float> undef to <16 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x float> undef to <16 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x float> undef to <16 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x float> undef to <16 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:87 CodeSize:1 Lat:1 SizeLat:1 for: %r160 = fptoui <16 x double> undef to <16 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:87 CodeSize:1 Lat:1 SizeLat:1 for: %r161 = fptosi <16 x double> undef to <16 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:23 CodeSize:1 Lat:1 SizeLat:1 for: %r162 = fptoui <16 x double> undef to <16 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:23 CodeSize:1 Lat:1 SizeLat:1 for: %r163 = fptosi <16 x double> undef to <16 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:22 CodeSize:1 Lat:1 SizeLat:1 for: %r164 = fptoui <16 x double> undef to <16 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:22 CodeSize:1 Lat:1 SizeLat:1 for: %r165 = fptosi <16 x double> undef to <16 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r166 = fptoui <16 x double> undef to <16 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r167 = fptosi <16 x double> undef to <16 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r168 = fptoui <16 x double> undef to <16 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r169 = fptosi <16 x double> undef to <16 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r170 = uitofp <2 x i1> undef to <2 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r171 = sitofp <2 x i1> undef to <2 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r172 = uitofp <2 x i8> undef to <2 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r173 = sitofp <2 x i8> undef to <2 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r174 = uitofp <2 x i16> undef to <2 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r175 = sitofp <2 x i16> undef to <2 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r176 = uitofp <2 x i32> undef to <2 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r177 = sitofp <2 x i32> undef to <2 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r178 = uitofp <2 x i64> undef to <2 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r179 = sitofp <2 x i64> undef to <2 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r180 = uitofp <2 x i1> undef to <2 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r181 = sitofp <2 x i1> undef to <2 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r182 = uitofp <2 x i8> undef to <2 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r183 = sitofp <2 x i8> undef to <2 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r184 = uitofp <2 x i16> undef to <2 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r185 = sitofp <2 x i16> undef to <2 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r186 = uitofp <2 x i32> undef to <2 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r187 = sitofp <2 x i32> undef to <2 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r188 = uitofp <2 x i64> undef to <2 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r189 = sitofp <2 x i64> undef to <2 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r190 = uitofp <4 x i1> undef to <4 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r191 = sitofp <4 x i1> undef to <4 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r192 = uitofp <4 x i8> undef to <4 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r193 = sitofp <4 x i8> undef to <4 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r194 = uitofp <4 x i16> undef to <4 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r195 = sitofp <4 x i16> undef to <4 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r196 = uitofp <4 x i32> undef to <4 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r197 = sitofp <4 x i32> undef to <4 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:18 CodeSize:1 Lat:1 SizeLat:1 for: %r198 = uitofp <4 x i64> undef to <4 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:18 CodeSize:1 Lat:1 SizeLat:1 for: %r199 = sitofp <4 x i64> undef to <4 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r200 = uitofp <4 x i1> undef to <4 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r201 = sitofp <4 x i1> undef to <4 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r202 = uitofp <4 x i8> undef to <4 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r203 = sitofp <4 x i8> undef to <4 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r204 = uitofp <4 x i16> undef to <4 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r205 = sitofp <4 x i16> undef to <4 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r206 = uitofp <4 x i32> undef to <4 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r207 = sitofp <4 x i32> undef to <4 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r208 = uitofp <4 x i64> undef to <4 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r209 = sitofp <4 x i64> undef to <4 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r210 = uitofp <8 x i1> undef to <8 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r211 = sitofp <8 x i1> undef to <8 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r212 = uitofp <8 x i8> undef to <8 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r213 = sitofp <8 x i8> undef to <8 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r214 = uitofp <8 x i16> undef to <8 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r215 = sitofp <8 x i16> undef to <8 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r216 = uitofp <8 x i32> undef to <8 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r217 = sitofp <8 x i32> undef to <8 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:36 CodeSize:1 Lat:1 SizeLat:1 for: %r218 = uitofp <8 x i64> undef to <8 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:36 CodeSize:1 Lat:1 SizeLat:1 for: %r219 = sitofp <8 x i64> undef to <8 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r220 = uitofp <8 x i1> undef to <8 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r221 = sitofp <8 x i1> undef to <8 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r222 = uitofp <8 x i8> undef to <8 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r223 = sitofp <8 x i8> undef to <8 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r224 = uitofp <8 x i16> undef to <8 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r225 = sitofp <8 x i16> undef to <8 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r226 = uitofp <8 x i32> undef to <8 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r227 = sitofp <8 x i32> undef to <8 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r228 = uitofp <8 x i64> undef to <8 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r229 = sitofp <8 x i64> undef to <8 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r230 = uitofp <16 x i1> undef to <16 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r231 = sitofp <16 x i1> undef to <16 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r232 = uitofp <16 x i8> undef to <16 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r233 = sitofp <16 x i8> undef to <16 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r234 = uitofp <16 x i16> undef to <16 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r235 = sitofp <16 x i16> undef to <16 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r236 = uitofp <16 x i32> undef to <16 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r237 = sitofp <16 x i32> undef to <16 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:72 CodeSize:1 Lat:1 SizeLat:1 for: %r238 = uitofp <16 x i64> undef to <16 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:72 CodeSize:1 Lat:1 SizeLat:1 for: %r239 = sitofp <16 x i64> undef to <16 x float> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %r240 = uitofp <16 x i1> undef to <16 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %r241 = sitofp <16 x i1> undef to <16 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:39 CodeSize:1 Lat:1 SizeLat:1 for: %r242 = uitofp <16 x i8> undef to <16 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:39 CodeSize:1 Lat:1 SizeLat:1 for: %r243 = sitofp <16 x i8> undef to <16 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %r244 = uitofp <16 x i16> undef to <16 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %r245 = sitofp <16 x i16> undef to <16 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:16 CodeSize:1 Lat:1 SizeLat:1 for: %r246 = uitofp <16 x i32> undef to <16 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:16 CodeSize:1 Lat:1 SizeLat:1 for: %r247 = sitofp <16 x i32> undef to <16 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r248 = uitofp <16 x i64> undef to <16 x double> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r249 = sitofp <16 x i64> undef to <16 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui float poison to i1 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi float poison to i1 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui float poison to i8 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi float poison to i8 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui float poison to i16 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi float poison to i16 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui float poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi float poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui float poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi float poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r40 = fptoui double poison to i1 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r41 = fptosi double poison to i1 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r42 = fptoui double poison to i8 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r43 = fptosi double poison to i8 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r44 = fptoui double poison to i16 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r45 = fptosi double poison to i16 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r46 = fptoui double poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r47 = fptosi double poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r48 = fptoui double poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r49 = fptosi double poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r50 = sitofp i1 poison to float +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r51 = uitofp i1 poison to float +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r52 = sitofp i1 poison to double +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r53 = uitofp i1 poison to double +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r54 = sitofp i8 poison to float +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r55 = uitofp i8 poison to float +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r56 = sitofp i8 poison to double +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r57 = uitofp i8 poison to double +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r58 = sitofp i16 poison to float +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r59 = uitofp i16 poison to float +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r60 = sitofp i16 poison to double +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r61 = uitofp i16 poison to double +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r62 = sitofp i32 poison to float +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r63 = uitofp i32 poison to float +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r64 = sitofp i32 poison to double +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r65 = uitofp i32 poison to double +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r66 = sitofp i64 poison to float +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r67 = uitofp i64 poison to float +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r68 = sitofp i64 poison to double +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r69 = uitofp i64 poison to double +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r80 = fptrunc double poison to float +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r81 = fptrunc <2 x double> poison to <2 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r82 = fptrunc <4 x double> poison to <4 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r83 = fptrunc <8 x double> poison to <8 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r84 = fptrunc <16 x double> poison to <16 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r85 = fpext float poison to double +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r86 = fpext <2 x float> poison to <2 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r87 = fpext <4 x float> poison to <4 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r88 = fpext <8 x float> poison to <8 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r89 = fpext <16 x float> poison to <16 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x float> poison to <2 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x float> poison to <2 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x float> poison to <2 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x float> poison to <2 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x float> poison to <2 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x float> poison to <2 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x float> poison to <2 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x float> poison to <2 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x float> poison to <2 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x float> poison to <2 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r100 = fptoui <2 x double> poison to <2 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r101 = fptosi <2 x double> poison to <2 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r102 = fptoui <2 x double> poison to <2 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r103 = fptosi <2 x double> poison to <2 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r104 = fptoui <2 x double> poison to <2 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r105 = fptosi <2 x double> poison to <2 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r106 = fptoui <2 x double> poison to <2 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r107 = fptosi <2 x double> poison to <2 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r108 = fptoui <2 x double> poison to <2 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r109 = fptosi <2 x double> poison to <2 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r110 = fptoui <4 x float> poison to <4 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r111 = fptosi <4 x float> poison to <4 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r112 = fptoui <4 x float> poison to <4 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r113 = fptosi <4 x float> poison to <4 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r114 = fptoui <4 x float> poison to <4 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r115 = fptosi <4 x float> poison to <4 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r116 = fptoui <4 x float> poison to <4 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r117 = fptosi <4 x float> poison to <4 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x float> poison to <4 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x float> poison to <4 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r120 = fptoui <4 x double> poison to <4 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r121 = fptosi <4 x double> poison to <4 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r122 = fptoui <4 x double> poison to <4 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r123 = fptosi <4 x double> poison to <4 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r124 = fptoui <4 x double> poison to <4 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r125 = fptosi <4 x double> poison to <4 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r126 = fptoui <4 x double> poison to <4 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r127 = fptosi <4 x double> poison to <4 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r128 = fptoui <4 x double> poison to <4 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r129 = fptosi <4 x double> poison to <4 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:41 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x float> poison to <8 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:41 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x float> poison to <8 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x float> poison to <8 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x float> poison to <8 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r134 = fptoui <8 x float> poison to <8 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %r135 = fptosi <8 x float> poison to <8 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x float> poison to <8 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x float> poison to <8 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x float> poison to <8 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x float> poison to <8 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r140 = fptoui <8 x double> poison to <8 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r141 = fptosi <8 x double> poison to <8 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r142 = fptoui <8 x double> poison to <8 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r143 = fptosi <8 x double> poison to <8 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r144 = fptoui <8 x double> poison to <8 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r145 = fptosi <8 x double> poison to <8 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r146 = fptoui <8 x double> poison to <8 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r147 = fptosi <8 x double> poison to <8 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r148 = fptoui <8 x double> poison to <8 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r149 = fptosi <8 x double> poison to <8 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:83 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x float> poison to <16 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:83 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x float> poison to <16 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x float> poison to <16 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x float> poison to <16 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x float> poison to <16 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x float> poison to <16 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x float> poison to <16 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x float> poison to <16 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x float> poison to <16 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x float> poison to <16 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:87 CodeSize:1 Lat:1 SizeLat:1 for: %r160 = fptoui <16 x double> poison to <16 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:87 CodeSize:1 Lat:1 SizeLat:1 for: %r161 = fptosi <16 x double> poison to <16 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:23 CodeSize:1 Lat:1 SizeLat:1 for: %r162 = fptoui <16 x double> poison to <16 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:23 CodeSize:1 Lat:1 SizeLat:1 for: %r163 = fptosi <16 x double> poison to <16 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:22 CodeSize:1 Lat:1 SizeLat:1 for: %r164 = fptoui <16 x double> poison to <16 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:22 CodeSize:1 Lat:1 SizeLat:1 for: %r165 = fptosi <16 x double> poison to <16 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r166 = fptoui <16 x double> poison to <16 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r167 = fptosi <16 x double> poison to <16 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r168 = fptoui <16 x double> poison to <16 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r169 = fptosi <16 x double> poison to <16 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r170 = uitofp <2 x i1> poison to <2 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r171 = sitofp <2 x i1> poison to <2 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r172 = uitofp <2 x i8> poison to <2 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r173 = sitofp <2 x i8> poison to <2 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r174 = uitofp <2 x i16> poison to <2 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r175 = sitofp <2 x i16> poison to <2 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r176 = uitofp <2 x i32> poison to <2 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r177 = sitofp <2 x i32> poison to <2 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r178 = uitofp <2 x i64> poison to <2 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r179 = sitofp <2 x i64> poison to <2 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r180 = uitofp <2 x i1> poison to <2 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r181 = sitofp <2 x i1> poison to <2 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r182 = uitofp <2 x i8> poison to <2 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r183 = sitofp <2 x i8> poison to <2 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r184 = uitofp <2 x i16> poison to <2 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r185 = sitofp <2 x i16> poison to <2 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r186 = uitofp <2 x i32> poison to <2 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r187 = sitofp <2 x i32> poison to <2 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r188 = uitofp <2 x i64> poison to <2 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r189 = sitofp <2 x i64> poison to <2 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r190 = uitofp <4 x i1> poison to <4 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r191 = sitofp <4 x i1> poison to <4 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r192 = uitofp <4 x i8> poison to <4 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r193 = sitofp <4 x i8> poison to <4 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r194 = uitofp <4 x i16> poison to <4 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r195 = sitofp <4 x i16> poison to <4 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r196 = uitofp <4 x i32> poison to <4 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r197 = sitofp <4 x i32> poison to <4 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:18 CodeSize:1 Lat:1 SizeLat:1 for: %r198 = uitofp <4 x i64> poison to <4 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:18 CodeSize:1 Lat:1 SizeLat:1 for: %r199 = sitofp <4 x i64> poison to <4 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r200 = uitofp <4 x i1> poison to <4 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r201 = sitofp <4 x i1> poison to <4 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r202 = uitofp <4 x i8> poison to <4 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r203 = sitofp <4 x i8> poison to <4 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r204 = uitofp <4 x i16> poison to <4 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %r205 = sitofp <4 x i16> poison to <4 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r206 = uitofp <4 x i32> poison to <4 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r207 = sitofp <4 x i32> poison to <4 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r208 = uitofp <4 x i64> poison to <4 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r209 = sitofp <4 x i64> poison to <4 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r210 = uitofp <8 x i1> poison to <8 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r211 = sitofp <8 x i1> poison to <8 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r212 = uitofp <8 x i8> poison to <8 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r213 = sitofp <8 x i8> poison to <8 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r214 = uitofp <8 x i16> poison to <8 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r215 = sitofp <8 x i16> poison to <8 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r216 = uitofp <8 x i32> poison to <8 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r217 = sitofp <8 x i32> poison to <8 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:36 CodeSize:1 Lat:1 SizeLat:1 for: %r218 = uitofp <8 x i64> poison to <8 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:36 CodeSize:1 Lat:1 SizeLat:1 for: %r219 = sitofp <8 x i64> poison to <8 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r220 = uitofp <8 x i1> poison to <8 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r221 = sitofp <8 x i1> poison to <8 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r222 = uitofp <8 x i8> poison to <8 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r223 = sitofp <8 x i8> poison to <8 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r224 = uitofp <8 x i16> poison to <8 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %r225 = sitofp <8 x i16> poison to <8 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r226 = uitofp <8 x i32> poison to <8 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r227 = sitofp <8 x i32> poison to <8 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r228 = uitofp <8 x i64> poison to <8 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r229 = sitofp <8 x i64> poison to <8 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r230 = uitofp <16 x i1> poison to <16 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r231 = sitofp <16 x i1> poison to <16 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r232 = uitofp <16 x i8> poison to <16 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r233 = sitofp <16 x i8> poison to <16 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r234 = uitofp <16 x i16> poison to <16 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r235 = sitofp <16 x i16> poison to <16 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r236 = uitofp <16 x i32> poison to <16 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r237 = sitofp <16 x i32> poison to <16 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:72 CodeSize:1 Lat:1 SizeLat:1 for: %r238 = uitofp <16 x i64> poison to <16 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:72 CodeSize:1 Lat:1 SizeLat:1 for: %r239 = sitofp <16 x i64> poison to <16 x float> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %r240 = uitofp <16 x i1> poison to <16 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:15 CodeSize:1 Lat:1 SizeLat:1 for: %r241 = sitofp <16 x i1> poison to <16 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:39 CodeSize:1 Lat:1 SizeLat:1 for: %r242 = uitofp <16 x i8> poison to <16 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:39 CodeSize:1 Lat:1 SizeLat:1 for: %r243 = sitofp <16 x i8> poison to <16 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %r244 = uitofp <16 x i16> poison to <16 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:38 CodeSize:1 Lat:1 SizeLat:1 for: %r245 = sitofp <16 x i16> poison to <16 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:16 CodeSize:1 Lat:1 SizeLat:1 for: %r246 = uitofp <16 x i32> poison to <16 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:16 CodeSize:1 Lat:1 SizeLat:1 for: %r247 = sitofp <16 x i32> poison to <16 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r248 = uitofp <16 x i64> poison to <16 x double> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r249 = sitofp <16 x i64> poison to <16 x double> ; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef ; ; SVE128-NO-NEON-LABEL: 'casts_no_users' -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui float undef to i1 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi float undef to i1 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui float undef to i8 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi float undef to i8 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui float undef to i16 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi float undef to i16 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui float undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi float undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui float undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi float undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r40 = fptoui double undef to i1 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r41 = fptosi double undef to i1 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r42 = fptoui double undef to i8 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r43 = fptosi double undef to i8 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r44 = fptoui double undef to i16 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r45 = fptosi double undef to i16 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r46 = fptoui double undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r47 = fptosi double undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r48 = fptoui double undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r49 = fptosi double undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r50 = sitofp i1 undef to float -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r51 = uitofp i1 undef to float -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r52 = sitofp i1 undef to double -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r53 = uitofp i1 undef to double -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r54 = sitofp i8 undef to float -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r55 = uitofp i8 undef to float -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r56 = sitofp i8 undef to double -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r57 = uitofp i8 undef to double -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r58 = sitofp i16 undef to float -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r59 = uitofp i16 undef to float -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r60 = sitofp i16 undef to double -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r61 = uitofp i16 undef to double -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r62 = sitofp i32 undef to float -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r63 = uitofp i32 undef to float -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r64 = sitofp i32 undef to double -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r65 = uitofp i32 undef to double -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r66 = sitofp i64 undef to float -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r67 = uitofp i64 undef to float -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r68 = sitofp i64 undef to double -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r69 = uitofp i64 undef to double -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r80 = fptrunc double undef to float -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r81 = fptrunc <2 x double> undef to <2 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r82 = fptrunc <4 x double> undef to <4 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r83 = fptrunc <8 x double> undef to <8 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r84 = fptrunc <16 x double> undef to <16 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r85 = fpext float undef to double -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r86 = fpext <2 x float> undef to <2 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r87 = fpext <4 x float> undef to <4 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r88 = fpext <8 x float> undef to <8 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r89 = fpext <16 x float> undef to <16 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x float> undef to <2 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x float> undef to <2 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x float> undef to <2 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x float> undef to <2 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x float> undef to <2 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x float> undef to <2 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x float> undef to <2 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x float> undef to <2 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r98 = fptoui <2 x float> undef to <2 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r99 = fptosi <2 x float> undef to <2 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r100 = fptoui <2 x double> undef to <2 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r101 = fptosi <2 x double> undef to <2 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r102 = fptoui <2 x double> undef to <2 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r103 = fptosi <2 x double> undef to <2 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r104 = fptoui <2 x double> undef to <2 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r105 = fptosi <2 x double> undef to <2 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r106 = fptoui <2 x double> undef to <2 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r107 = fptosi <2 x double> undef to <2 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r108 = fptoui <2 x double> undef to <2 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r109 = fptosi <2 x double> undef to <2 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x float> undef to <4 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x float> undef to <4 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x float> undef to <4 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x float> undef to <4 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x float> undef to <4 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x float> undef to <4 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r116 = fptoui <4 x float> undef to <4 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r117 = fptosi <4 x float> undef to <4 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x float> undef to <4 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x float> undef to <4 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r120 = fptoui <4 x double> undef to <4 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r121 = fptosi <4 x double> undef to <4 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r122 = fptoui <4 x double> undef to <4 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r123 = fptosi <4 x double> undef to <4 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r124 = fptoui <4 x double> undef to <4 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r125 = fptosi <4 x double> undef to <4 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r126 = fptoui <4 x double> undef to <4 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r127 = fptosi <4 x double> undef to <4 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r128 = fptoui <4 x double> undef to <4 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r129 = fptosi <4 x double> undef to <4 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x float> undef to <8 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x float> undef to <8 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x float> undef to <8 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x float> undef to <8 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r134 = fptoui <8 x float> undef to <8 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r135 = fptosi <8 x float> undef to <8 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x float> undef to <8 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x float> undef to <8 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x float> undef to <8 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x float> undef to <8 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r140 = fptoui <8 x double> undef to <8 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r141 = fptosi <8 x double> undef to <8 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r142 = fptoui <8 x double> undef to <8 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r143 = fptosi <8 x double> undef to <8 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r144 = fptoui <8 x double> undef to <8 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r145 = fptosi <8 x double> undef to <8 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r146 = fptoui <8 x double> undef to <8 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r147 = fptosi <8 x double> undef to <8 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r148 = fptoui <8 x double> undef to <8 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r149 = fptosi <8 x double> undef to <8 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x float> undef to <16 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x float> undef to <16 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x float> undef to <16 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x float> undef to <16 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x float> undef to <16 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x float> undef to <16 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x float> undef to <16 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x float> undef to <16 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x float> undef to <16 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x float> undef to <16 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r160 = fptoui <16 x double> undef to <16 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r161 = fptosi <16 x double> undef to <16 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r162 = fptoui <16 x double> undef to <16 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r163 = fptosi <16 x double> undef to <16 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r164 = fptoui <16 x double> undef to <16 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r165 = fptosi <16 x double> undef to <16 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r166 = fptoui <16 x double> undef to <16 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r167 = fptosi <16 x double> undef to <16 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r168 = fptoui <16 x double> undef to <16 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r169 = fptosi <16 x double> undef to <16 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r170 = uitofp <2 x i1> undef to <2 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r171 = sitofp <2 x i1> undef to <2 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r172 = uitofp <2 x i8> undef to <2 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r173 = sitofp <2 x i8> undef to <2 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r174 = uitofp <2 x i16> undef to <2 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r175 = sitofp <2 x i16> undef to <2 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r176 = uitofp <2 x i32> undef to <2 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r177 = sitofp <2 x i32> undef to <2 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r178 = uitofp <2 x i64> undef to <2 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r179 = sitofp <2 x i64> undef to <2 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r180 = uitofp <2 x i1> undef to <2 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r181 = sitofp <2 x i1> undef to <2 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r182 = uitofp <2 x i8> undef to <2 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r183 = sitofp <2 x i8> undef to <2 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r184 = uitofp <2 x i16> undef to <2 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r185 = sitofp <2 x i16> undef to <2 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r186 = uitofp <2 x i32> undef to <2 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r187 = sitofp <2 x i32> undef to <2 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r188 = uitofp <2 x i64> undef to <2 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r189 = sitofp <2 x i64> undef to <2 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r190 = uitofp <4 x i1> undef to <4 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r191 = sitofp <4 x i1> undef to <4 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r192 = uitofp <4 x i8> undef to <4 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r193 = sitofp <4 x i8> undef to <4 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r194 = uitofp <4 x i16> undef to <4 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r195 = sitofp <4 x i16> undef to <4 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r196 = uitofp <4 x i32> undef to <4 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r197 = sitofp <4 x i32> undef to <4 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r198 = uitofp <4 x i64> undef to <4 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r199 = sitofp <4 x i64> undef to <4 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r200 = uitofp <4 x i1> undef to <4 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r201 = sitofp <4 x i1> undef to <4 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r202 = uitofp <4 x i8> undef to <4 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r203 = sitofp <4 x i8> undef to <4 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r204 = uitofp <4 x i16> undef to <4 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r205 = sitofp <4 x i16> undef to <4 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r206 = uitofp <4 x i32> undef to <4 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r207 = sitofp <4 x i32> undef to <4 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r208 = uitofp <4 x i64> undef to <4 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r209 = sitofp <4 x i64> undef to <4 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r210 = uitofp <8 x i1> undef to <8 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r211 = sitofp <8 x i1> undef to <8 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r212 = uitofp <8 x i8> undef to <8 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r213 = sitofp <8 x i8> undef to <8 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r214 = uitofp <8 x i16> undef to <8 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r215 = sitofp <8 x i16> undef to <8 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r216 = uitofp <8 x i32> undef to <8 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r217 = sitofp <8 x i32> undef to <8 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r218 = uitofp <8 x i64> undef to <8 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r219 = sitofp <8 x i64> undef to <8 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r220 = uitofp <8 x i1> undef to <8 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r221 = sitofp <8 x i1> undef to <8 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r222 = uitofp <8 x i8> undef to <8 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r223 = sitofp <8 x i8> undef to <8 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r224 = uitofp <8 x i16> undef to <8 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r225 = sitofp <8 x i16> undef to <8 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r226 = uitofp <8 x i32> undef to <8 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r227 = sitofp <8 x i32> undef to <8 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r228 = uitofp <8 x i64> undef to <8 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r229 = sitofp <8 x i64> undef to <8 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r230 = uitofp <16 x i1> undef to <16 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r231 = sitofp <16 x i1> undef to <16 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r232 = uitofp <16 x i8> undef to <16 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r233 = sitofp <16 x i8> undef to <16 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r234 = uitofp <16 x i16> undef to <16 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r235 = sitofp <16 x i16> undef to <16 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r236 = uitofp <16 x i32> undef to <16 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r237 = sitofp <16 x i32> undef to <16 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r238 = uitofp <16 x i64> undef to <16 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r239 = sitofp <16 x i64> undef to <16 x float> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r240 = uitofp <16 x i1> undef to <16 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r241 = sitofp <16 x i1> undef to <16 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:16 CodeSize:1 Lat:1 SizeLat:1 for: %r242 = uitofp <16 x i8> undef to <16 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:16 CodeSize:1 Lat:1 SizeLat:1 for: %r243 = sitofp <16 x i8> undef to <16 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r244 = uitofp <16 x i16> undef to <16 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r245 = sitofp <16 x i16> undef to <16 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r246 = uitofp <16 x i32> undef to <16 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r247 = sitofp <16 x i32> undef to <16 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r248 = uitofp <16 x i64> undef to <16 x double> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r249 = sitofp <16 x i64> undef to <16 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui float poison to i1 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi float poison to i1 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui float poison to i8 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi float poison to i8 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui float poison to i16 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi float poison to i16 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui float poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi float poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui float poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi float poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r40 = fptoui double poison to i1 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r41 = fptosi double poison to i1 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r42 = fptoui double poison to i8 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r43 = fptosi double poison to i8 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r44 = fptoui double poison to i16 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r45 = fptosi double poison to i16 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r46 = fptoui double poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r47 = fptosi double poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r48 = fptoui double poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r49 = fptosi double poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r50 = sitofp i1 poison to float +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r51 = uitofp i1 poison to float +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r52 = sitofp i1 poison to double +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r53 = uitofp i1 poison to double +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r54 = sitofp i8 poison to float +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r55 = uitofp i8 poison to float +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r56 = sitofp i8 poison to double +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r57 = uitofp i8 poison to double +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r58 = sitofp i16 poison to float +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r59 = uitofp i16 poison to float +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r60 = sitofp i16 poison to double +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r61 = uitofp i16 poison to double +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r62 = sitofp i32 poison to float +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r63 = uitofp i32 poison to float +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r64 = sitofp i32 poison to double +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r65 = uitofp i32 poison to double +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r66 = sitofp i64 poison to float +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r67 = uitofp i64 poison to float +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r68 = sitofp i64 poison to double +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r69 = uitofp i64 poison to double +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r80 = fptrunc double poison to float +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r81 = fptrunc <2 x double> poison to <2 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r82 = fptrunc <4 x double> poison to <4 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r83 = fptrunc <8 x double> poison to <8 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r84 = fptrunc <16 x double> poison to <16 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r85 = fpext float poison to double +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r86 = fpext <2 x float> poison to <2 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r87 = fpext <4 x float> poison to <4 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r88 = fpext <8 x float> poison to <8 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r89 = fpext <16 x float> poison to <16 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x float> poison to <2 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x float> poison to <2 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x float> poison to <2 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x float> poison to <2 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x float> poison to <2 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x float> poison to <2 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x float> poison to <2 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x float> poison to <2 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r98 = fptoui <2 x float> poison to <2 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r99 = fptosi <2 x float> poison to <2 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r100 = fptoui <2 x double> poison to <2 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r101 = fptosi <2 x double> poison to <2 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r102 = fptoui <2 x double> poison to <2 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r103 = fptosi <2 x double> poison to <2 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r104 = fptoui <2 x double> poison to <2 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r105 = fptosi <2 x double> poison to <2 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r106 = fptoui <2 x double> poison to <2 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r107 = fptosi <2 x double> poison to <2 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r108 = fptoui <2 x double> poison to <2 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r109 = fptosi <2 x double> poison to <2 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x float> poison to <4 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x float> poison to <4 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x float> poison to <4 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x float> poison to <4 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x float> poison to <4 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x float> poison to <4 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r116 = fptoui <4 x float> poison to <4 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r117 = fptosi <4 x float> poison to <4 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x float> poison to <4 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x float> poison to <4 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r120 = fptoui <4 x double> poison to <4 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r121 = fptosi <4 x double> poison to <4 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r122 = fptoui <4 x double> poison to <4 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r123 = fptosi <4 x double> poison to <4 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r124 = fptoui <4 x double> poison to <4 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r125 = fptosi <4 x double> poison to <4 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r126 = fptoui <4 x double> poison to <4 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r127 = fptosi <4 x double> poison to <4 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r128 = fptoui <4 x double> poison to <4 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r129 = fptosi <4 x double> poison to <4 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x float> poison to <8 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x float> poison to <8 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x float> poison to <8 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x float> poison to <8 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r134 = fptoui <8 x float> poison to <8 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r135 = fptosi <8 x float> poison to <8 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x float> poison to <8 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x float> poison to <8 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x float> poison to <8 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x float> poison to <8 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r140 = fptoui <8 x double> poison to <8 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r141 = fptosi <8 x double> poison to <8 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r142 = fptoui <8 x double> poison to <8 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r143 = fptosi <8 x double> poison to <8 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r144 = fptoui <8 x double> poison to <8 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r145 = fptosi <8 x double> poison to <8 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r146 = fptoui <8 x double> poison to <8 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r147 = fptosi <8 x double> poison to <8 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r148 = fptoui <8 x double> poison to <8 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r149 = fptosi <8 x double> poison to <8 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x float> poison to <16 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x float> poison to <16 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x float> poison to <16 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x float> poison to <16 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x float> poison to <16 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x float> poison to <16 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x float> poison to <16 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x float> poison to <16 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x float> poison to <16 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x float> poison to <16 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r160 = fptoui <16 x double> poison to <16 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r161 = fptosi <16 x double> poison to <16 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r162 = fptoui <16 x double> poison to <16 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r163 = fptosi <16 x double> poison to <16 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r164 = fptoui <16 x double> poison to <16 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r165 = fptosi <16 x double> poison to <16 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r166 = fptoui <16 x double> poison to <16 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r167 = fptosi <16 x double> poison to <16 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r168 = fptoui <16 x double> poison to <16 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r169 = fptosi <16 x double> poison to <16 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r170 = uitofp <2 x i1> poison to <2 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r171 = sitofp <2 x i1> poison to <2 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r172 = uitofp <2 x i8> poison to <2 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r173 = sitofp <2 x i8> poison to <2 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r174 = uitofp <2 x i16> poison to <2 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r175 = sitofp <2 x i16> poison to <2 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r176 = uitofp <2 x i32> poison to <2 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r177 = sitofp <2 x i32> poison to <2 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r178 = uitofp <2 x i64> poison to <2 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r179 = sitofp <2 x i64> poison to <2 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r180 = uitofp <2 x i1> poison to <2 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r181 = sitofp <2 x i1> poison to <2 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r182 = uitofp <2 x i8> poison to <2 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r183 = sitofp <2 x i8> poison to <2 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r184 = uitofp <2 x i16> poison to <2 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r185 = sitofp <2 x i16> poison to <2 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r186 = uitofp <2 x i32> poison to <2 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r187 = sitofp <2 x i32> poison to <2 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r188 = uitofp <2 x i64> poison to <2 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r189 = sitofp <2 x i64> poison to <2 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r190 = uitofp <4 x i1> poison to <4 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r191 = sitofp <4 x i1> poison to <4 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r192 = uitofp <4 x i8> poison to <4 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r193 = sitofp <4 x i8> poison to <4 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r194 = uitofp <4 x i16> poison to <4 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r195 = sitofp <4 x i16> poison to <4 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r196 = uitofp <4 x i32> poison to <4 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r197 = sitofp <4 x i32> poison to <4 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r198 = uitofp <4 x i64> poison to <4 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r199 = sitofp <4 x i64> poison to <4 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r200 = uitofp <4 x i1> poison to <4 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r201 = sitofp <4 x i1> poison to <4 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r202 = uitofp <4 x i8> poison to <4 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r203 = sitofp <4 x i8> poison to <4 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r204 = uitofp <4 x i16> poison to <4 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r205 = sitofp <4 x i16> poison to <4 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r206 = uitofp <4 x i32> poison to <4 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r207 = sitofp <4 x i32> poison to <4 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r208 = uitofp <4 x i64> poison to <4 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r209 = sitofp <4 x i64> poison to <4 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r210 = uitofp <8 x i1> poison to <8 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r211 = sitofp <8 x i1> poison to <8 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r212 = uitofp <8 x i8> poison to <8 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r213 = sitofp <8 x i8> poison to <8 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r214 = uitofp <8 x i16> poison to <8 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r215 = sitofp <8 x i16> poison to <8 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r216 = uitofp <8 x i32> poison to <8 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r217 = sitofp <8 x i32> poison to <8 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r218 = uitofp <8 x i64> poison to <8 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r219 = sitofp <8 x i64> poison to <8 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r220 = uitofp <8 x i1> poison to <8 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r221 = sitofp <8 x i1> poison to <8 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r222 = uitofp <8 x i8> poison to <8 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r223 = sitofp <8 x i8> poison to <8 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r224 = uitofp <8 x i16> poison to <8 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r225 = sitofp <8 x i16> poison to <8 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r226 = uitofp <8 x i32> poison to <8 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r227 = sitofp <8 x i32> poison to <8 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r228 = uitofp <8 x i64> poison to <8 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r229 = sitofp <8 x i64> poison to <8 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r230 = uitofp <16 x i1> poison to <16 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r231 = sitofp <16 x i1> poison to <16 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r232 = uitofp <16 x i8> poison to <16 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r233 = sitofp <16 x i8> poison to <16 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r234 = uitofp <16 x i16> poison to <16 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r235 = sitofp <16 x i16> poison to <16 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r236 = uitofp <16 x i32> poison to <16 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r237 = sitofp <16 x i32> poison to <16 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r238 = uitofp <16 x i64> poison to <16 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r239 = sitofp <16 x i64> poison to <16 x float> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r240 = uitofp <16 x i1> poison to <16 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r241 = sitofp <16 x i1> poison to <16 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:16 CodeSize:1 Lat:1 SizeLat:1 for: %r242 = uitofp <16 x i8> poison to <16 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:16 CodeSize:1 Lat:1 SizeLat:1 for: %r243 = sitofp <16 x i8> poison to <16 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r244 = uitofp <16 x i16> poison to <16 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r245 = sitofp <16 x i16> poison to <16 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r246 = uitofp <16 x i32> poison to <16 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r247 = sitofp <16 x i32> poison to <16 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r248 = uitofp <16 x i64> poison to <16 x double> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r249 = sitofp <16 x i64> poison to <16 x double> ; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef ; ; FIXED-MIN-256-LABEL: 'casts_no_users' -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui float undef to i1 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi float undef to i1 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui float undef to i8 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi float undef to i8 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui float undef to i16 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi float undef to i16 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui float undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi float undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui float undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi float undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r40 = fptoui double undef to i1 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r41 = fptosi double undef to i1 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r42 = fptoui double undef to i8 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r43 = fptosi double undef to i8 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r44 = fptoui double undef to i16 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r45 = fptosi double undef to i16 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r46 = fptoui double undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r47 = fptosi double undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r48 = fptoui double undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r49 = fptosi double undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r50 = sitofp i1 undef to float -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r51 = uitofp i1 undef to float -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r52 = sitofp i1 undef to double -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r53 = uitofp i1 undef to double -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r54 = sitofp i8 undef to float -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r55 = uitofp i8 undef to float -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r56 = sitofp i8 undef to double -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r57 = uitofp i8 undef to double -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r58 = sitofp i16 undef to float -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r59 = uitofp i16 undef to float -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r60 = sitofp i16 undef to double -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r61 = uitofp i16 undef to double -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r62 = sitofp i32 undef to float -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r63 = uitofp i32 undef to float -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r64 = sitofp i32 undef to double -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r65 = uitofp i32 undef to double -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r66 = sitofp i64 undef to float -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r67 = uitofp i64 undef to float -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r68 = sitofp i64 undef to double -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r69 = uitofp i64 undef to double -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r80 = fptrunc double undef to float -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r81 = fptrunc <2 x double> undef to <2 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r82 = fptrunc <4 x double> undef to <4 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r83 = fptrunc <8 x double> undef to <8 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r84 = fptrunc <16 x double> undef to <16 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r85 = fpext float undef to double -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r86 = fpext <2 x float> undef to <2 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r87 = fpext <4 x float> undef to <4 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r88 = fpext <8 x float> undef to <8 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r89 = fpext <16 x float> undef to <16 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x float> undef to <2 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x float> undef to <2 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x float> undef to <2 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x float> undef to <2 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x float> undef to <2 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x float> undef to <2 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x float> undef to <2 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x float> undef to <2 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x float> undef to <2 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x float> undef to <2 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r100 = fptoui <2 x double> undef to <2 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r101 = fptosi <2 x double> undef to <2 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r102 = fptoui <2 x double> undef to <2 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r103 = fptosi <2 x double> undef to <2 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r104 = fptoui <2 x double> undef to <2 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r105 = fptosi <2 x double> undef to <2 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r106 = fptoui <2 x double> undef to <2 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r107 = fptosi <2 x double> undef to <2 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r108 = fptoui <2 x double> undef to <2 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r109 = fptosi <2 x double> undef to <2 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r110 = fptoui <4 x float> undef to <4 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r111 = fptosi <4 x float> undef to <4 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r112 = fptoui <4 x float> undef to <4 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r113 = fptosi <4 x float> undef to <4 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r114 = fptoui <4 x float> undef to <4 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r115 = fptosi <4 x float> undef to <4 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r116 = fptoui <4 x float> undef to <4 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r117 = fptosi <4 x float> undef to <4 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r118 = fptoui <4 x float> undef to <4 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r119 = fptosi <4 x float> undef to <4 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r120 = fptoui <4 x double> undef to <4 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r121 = fptosi <4 x double> undef to <4 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r122 = fptoui <4 x double> undef to <4 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r123 = fptosi <4 x double> undef to <4 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r124 = fptoui <4 x double> undef to <4 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r125 = fptosi <4 x double> undef to <4 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r126 = fptoui <4 x double> undef to <4 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r127 = fptosi <4 x double> undef to <4 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r128 = fptoui <4 x double> undef to <4 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r129 = fptosi <4 x double> undef to <4 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r130 = fptoui <8 x float> undef to <8 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r131 = fptosi <8 x float> undef to <8 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r132 = fptoui <8 x float> undef to <8 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r133 = fptosi <8 x float> undef to <8 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x float> undef to <8 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x float> undef to <8 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r136 = fptoui <8 x float> undef to <8 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r137 = fptosi <8 x float> undef to <8 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x float> undef to <8 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x float> undef to <8 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r140 = fptoui <8 x double> undef to <8 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r141 = fptosi <8 x double> undef to <8 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r142 = fptoui <8 x double> undef to <8 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r143 = fptosi <8 x double> undef to <8 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r144 = fptoui <8 x double> undef to <8 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r145 = fptosi <8 x double> undef to <8 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r146 = fptoui <8 x double> undef to <8 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r147 = fptosi <8 x double> undef to <8 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r148 = fptoui <8 x double> undef to <8 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r149 = fptosi <8 x double> undef to <8 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x float> undef to <16 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x float> undef to <16 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x float> undef to <16 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x float> undef to <16 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x float> undef to <16 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x float> undef to <16 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x float> undef to <16 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x float> undef to <16 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x float> undef to <16 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x float> undef to <16 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r160 = fptoui <16 x double> undef to <16 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r161 = fptosi <16 x double> undef to <16 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r162 = fptoui <16 x double> undef to <16 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r163 = fptosi <16 x double> undef to <16 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r164 = fptoui <16 x double> undef to <16 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r165 = fptosi <16 x double> undef to <16 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r166 = fptoui <16 x double> undef to <16 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r167 = fptosi <16 x double> undef to <16 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r168 = fptoui <16 x double> undef to <16 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r169 = fptosi <16 x double> undef to <16 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r170 = uitofp <2 x i1> undef to <2 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r171 = sitofp <2 x i1> undef to <2 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r172 = uitofp <2 x i8> undef to <2 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r173 = sitofp <2 x i8> undef to <2 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r174 = uitofp <2 x i16> undef to <2 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r175 = sitofp <2 x i16> undef to <2 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r176 = uitofp <2 x i32> undef to <2 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r177 = sitofp <2 x i32> undef to <2 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r178 = uitofp <2 x i64> undef to <2 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r179 = sitofp <2 x i64> undef to <2 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r180 = uitofp <2 x i1> undef to <2 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r181 = sitofp <2 x i1> undef to <2 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r182 = uitofp <2 x i8> undef to <2 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r183 = sitofp <2 x i8> undef to <2 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r184 = uitofp <2 x i16> undef to <2 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r185 = sitofp <2 x i16> undef to <2 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r186 = uitofp <2 x i32> undef to <2 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r187 = sitofp <2 x i32> undef to <2 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r188 = uitofp <2 x i64> undef to <2 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r189 = sitofp <2 x i64> undef to <2 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r190 = uitofp <4 x i1> undef to <4 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r191 = sitofp <4 x i1> undef to <4 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r192 = uitofp <4 x i8> undef to <4 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r193 = sitofp <4 x i8> undef to <4 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r194 = uitofp <4 x i16> undef to <4 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r195 = sitofp <4 x i16> undef to <4 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r196 = uitofp <4 x i32> undef to <4 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r197 = sitofp <4 x i32> undef to <4 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r198 = uitofp <4 x i64> undef to <4 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r199 = sitofp <4 x i64> undef to <4 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r200 = uitofp <4 x i1> undef to <4 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r201 = sitofp <4 x i1> undef to <4 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r202 = uitofp <4 x i8> undef to <4 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r203 = sitofp <4 x i8> undef to <4 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r204 = uitofp <4 x i16> undef to <4 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r205 = sitofp <4 x i16> undef to <4 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r206 = uitofp <4 x i32> undef to <4 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r207 = sitofp <4 x i32> undef to <4 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r208 = uitofp <4 x i64> undef to <4 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r209 = sitofp <4 x i64> undef to <4 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r210 = uitofp <8 x i1> undef to <8 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r211 = sitofp <8 x i1> undef to <8 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r212 = uitofp <8 x i8> undef to <8 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r213 = sitofp <8 x i8> undef to <8 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r214 = uitofp <8 x i16> undef to <8 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r215 = sitofp <8 x i16> undef to <8 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r216 = uitofp <8 x i32> undef to <8 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r217 = sitofp <8 x i32> undef to <8 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r218 = uitofp <8 x i64> undef to <8 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r219 = sitofp <8 x i64> undef to <8 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r220 = uitofp <8 x i1> undef to <8 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r221 = sitofp <8 x i1> undef to <8 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r222 = uitofp <8 x i8> undef to <8 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r223 = sitofp <8 x i8> undef to <8 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r224 = uitofp <8 x i16> undef to <8 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r225 = sitofp <8 x i16> undef to <8 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r226 = uitofp <8 x i32> undef to <8 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r227 = sitofp <8 x i32> undef to <8 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r228 = uitofp <8 x i64> undef to <8 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r229 = sitofp <8 x i64> undef to <8 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r230 = uitofp <16 x i1> undef to <16 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r231 = sitofp <16 x i1> undef to <16 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r232 = uitofp <16 x i8> undef to <16 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r233 = sitofp <16 x i8> undef to <16 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r234 = uitofp <16 x i16> undef to <16 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r235 = sitofp <16 x i16> undef to <16 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r236 = uitofp <16 x i32> undef to <16 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r237 = sitofp <16 x i32> undef to <16 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r238 = uitofp <16 x i64> undef to <16 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r239 = sitofp <16 x i64> undef to <16 x float> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r240 = uitofp <16 x i1> undef to <16 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r241 = sitofp <16 x i1> undef to <16 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r242 = uitofp <16 x i8> undef to <16 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r243 = sitofp <16 x i8> undef to <16 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r244 = uitofp <16 x i16> undef to <16 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r245 = sitofp <16 x i16> undef to <16 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r246 = uitofp <16 x i32> undef to <16 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r247 = sitofp <16 x i32> undef to <16 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r248 = uitofp <16 x i64> undef to <16 x double> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r249 = sitofp <16 x i64> undef to <16 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui float poison to i1 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi float poison to i1 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui float poison to i8 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi float poison to i8 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui float poison to i16 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi float poison to i16 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui float poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi float poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui float poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi float poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r40 = fptoui double poison to i1 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r41 = fptosi double poison to i1 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r42 = fptoui double poison to i8 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r43 = fptosi double poison to i8 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r44 = fptoui double poison to i16 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r45 = fptosi double poison to i16 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r46 = fptoui double poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r47 = fptosi double poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r48 = fptoui double poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r49 = fptosi double poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r50 = sitofp i1 poison to float +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r51 = uitofp i1 poison to float +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r52 = sitofp i1 poison to double +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r53 = uitofp i1 poison to double +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r54 = sitofp i8 poison to float +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r55 = uitofp i8 poison to float +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r56 = sitofp i8 poison to double +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r57 = uitofp i8 poison to double +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r58 = sitofp i16 poison to float +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r59 = uitofp i16 poison to float +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r60 = sitofp i16 poison to double +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r61 = uitofp i16 poison to double +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r62 = sitofp i32 poison to float +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r63 = uitofp i32 poison to float +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r64 = sitofp i32 poison to double +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r65 = uitofp i32 poison to double +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r66 = sitofp i64 poison to float +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r67 = uitofp i64 poison to float +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r68 = sitofp i64 poison to double +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r69 = uitofp i64 poison to double +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r80 = fptrunc double poison to float +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r81 = fptrunc <2 x double> poison to <2 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r82 = fptrunc <4 x double> poison to <4 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r83 = fptrunc <8 x double> poison to <8 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r84 = fptrunc <16 x double> poison to <16 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r85 = fpext float poison to double +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r86 = fpext <2 x float> poison to <2 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r87 = fpext <4 x float> poison to <4 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r88 = fpext <8 x float> poison to <8 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r89 = fpext <16 x float> poison to <16 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x float> poison to <2 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x float> poison to <2 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x float> poison to <2 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x float> poison to <2 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x float> poison to <2 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x float> poison to <2 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x float> poison to <2 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x float> poison to <2 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x float> poison to <2 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x float> poison to <2 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r100 = fptoui <2 x double> poison to <2 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r101 = fptosi <2 x double> poison to <2 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r102 = fptoui <2 x double> poison to <2 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r103 = fptosi <2 x double> poison to <2 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r104 = fptoui <2 x double> poison to <2 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r105 = fptosi <2 x double> poison to <2 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r106 = fptoui <2 x double> poison to <2 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r107 = fptosi <2 x double> poison to <2 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r108 = fptoui <2 x double> poison to <2 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r109 = fptosi <2 x double> poison to <2 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r110 = fptoui <4 x float> poison to <4 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r111 = fptosi <4 x float> poison to <4 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r112 = fptoui <4 x float> poison to <4 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r113 = fptosi <4 x float> poison to <4 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r114 = fptoui <4 x float> poison to <4 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r115 = fptosi <4 x float> poison to <4 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r116 = fptoui <4 x float> poison to <4 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r117 = fptosi <4 x float> poison to <4 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r118 = fptoui <4 x float> poison to <4 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r119 = fptosi <4 x float> poison to <4 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r120 = fptoui <4 x double> poison to <4 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r121 = fptosi <4 x double> poison to <4 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r122 = fptoui <4 x double> poison to <4 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r123 = fptosi <4 x double> poison to <4 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r124 = fptoui <4 x double> poison to <4 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r125 = fptosi <4 x double> poison to <4 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r126 = fptoui <4 x double> poison to <4 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r127 = fptosi <4 x double> poison to <4 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r128 = fptoui <4 x double> poison to <4 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r129 = fptosi <4 x double> poison to <4 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r130 = fptoui <8 x float> poison to <8 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r131 = fptosi <8 x float> poison to <8 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r132 = fptoui <8 x float> poison to <8 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r133 = fptosi <8 x float> poison to <8 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x float> poison to <8 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x float> poison to <8 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r136 = fptoui <8 x float> poison to <8 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r137 = fptosi <8 x float> poison to <8 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x float> poison to <8 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x float> poison to <8 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r140 = fptoui <8 x double> poison to <8 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r141 = fptosi <8 x double> poison to <8 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r142 = fptoui <8 x double> poison to <8 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r143 = fptosi <8 x double> poison to <8 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r144 = fptoui <8 x double> poison to <8 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r145 = fptosi <8 x double> poison to <8 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r146 = fptoui <8 x double> poison to <8 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r147 = fptosi <8 x double> poison to <8 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r148 = fptoui <8 x double> poison to <8 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r149 = fptosi <8 x double> poison to <8 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x float> poison to <16 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x float> poison to <16 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x float> poison to <16 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x float> poison to <16 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x float> poison to <16 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x float> poison to <16 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x float> poison to <16 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x float> poison to <16 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x float> poison to <16 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x float> poison to <16 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r160 = fptoui <16 x double> poison to <16 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r161 = fptosi <16 x double> poison to <16 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r162 = fptoui <16 x double> poison to <16 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r163 = fptosi <16 x double> poison to <16 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r164 = fptoui <16 x double> poison to <16 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r165 = fptosi <16 x double> poison to <16 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r166 = fptoui <16 x double> poison to <16 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r167 = fptosi <16 x double> poison to <16 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r168 = fptoui <16 x double> poison to <16 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r169 = fptosi <16 x double> poison to <16 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r170 = uitofp <2 x i1> poison to <2 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r171 = sitofp <2 x i1> poison to <2 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r172 = uitofp <2 x i8> poison to <2 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r173 = sitofp <2 x i8> poison to <2 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r174 = uitofp <2 x i16> poison to <2 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r175 = sitofp <2 x i16> poison to <2 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r176 = uitofp <2 x i32> poison to <2 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r177 = sitofp <2 x i32> poison to <2 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r178 = uitofp <2 x i64> poison to <2 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r179 = sitofp <2 x i64> poison to <2 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r180 = uitofp <2 x i1> poison to <2 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r181 = sitofp <2 x i1> poison to <2 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r182 = uitofp <2 x i8> poison to <2 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r183 = sitofp <2 x i8> poison to <2 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r184 = uitofp <2 x i16> poison to <2 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r185 = sitofp <2 x i16> poison to <2 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r186 = uitofp <2 x i32> poison to <2 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r187 = sitofp <2 x i32> poison to <2 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r188 = uitofp <2 x i64> poison to <2 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r189 = sitofp <2 x i64> poison to <2 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r190 = uitofp <4 x i1> poison to <4 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r191 = sitofp <4 x i1> poison to <4 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r192 = uitofp <4 x i8> poison to <4 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r193 = sitofp <4 x i8> poison to <4 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r194 = uitofp <4 x i16> poison to <4 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r195 = sitofp <4 x i16> poison to <4 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r196 = uitofp <4 x i32> poison to <4 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r197 = sitofp <4 x i32> poison to <4 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r198 = uitofp <4 x i64> poison to <4 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r199 = sitofp <4 x i64> poison to <4 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r200 = uitofp <4 x i1> poison to <4 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r201 = sitofp <4 x i1> poison to <4 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r202 = uitofp <4 x i8> poison to <4 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r203 = sitofp <4 x i8> poison to <4 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r204 = uitofp <4 x i16> poison to <4 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r205 = sitofp <4 x i16> poison to <4 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r206 = uitofp <4 x i32> poison to <4 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r207 = sitofp <4 x i32> poison to <4 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r208 = uitofp <4 x i64> poison to <4 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r209 = sitofp <4 x i64> poison to <4 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r210 = uitofp <8 x i1> poison to <8 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r211 = sitofp <8 x i1> poison to <8 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r212 = uitofp <8 x i8> poison to <8 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r213 = sitofp <8 x i8> poison to <8 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r214 = uitofp <8 x i16> poison to <8 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r215 = sitofp <8 x i16> poison to <8 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r216 = uitofp <8 x i32> poison to <8 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r217 = sitofp <8 x i32> poison to <8 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r218 = uitofp <8 x i64> poison to <8 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r219 = sitofp <8 x i64> poison to <8 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r220 = uitofp <8 x i1> poison to <8 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r221 = sitofp <8 x i1> poison to <8 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r222 = uitofp <8 x i8> poison to <8 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r223 = sitofp <8 x i8> poison to <8 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r224 = uitofp <8 x i16> poison to <8 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r225 = sitofp <8 x i16> poison to <8 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r226 = uitofp <8 x i32> poison to <8 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r227 = sitofp <8 x i32> poison to <8 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r228 = uitofp <8 x i64> poison to <8 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r229 = sitofp <8 x i64> poison to <8 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r230 = uitofp <16 x i1> poison to <16 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r231 = sitofp <16 x i1> poison to <16 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r232 = uitofp <16 x i8> poison to <16 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r233 = sitofp <16 x i8> poison to <16 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r234 = uitofp <16 x i16> poison to <16 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r235 = sitofp <16 x i16> poison to <16 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r236 = uitofp <16 x i32> poison to <16 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r237 = sitofp <16 x i32> poison to <16 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r238 = uitofp <16 x i64> poison to <16 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r239 = sitofp <16 x i64> poison to <16 x float> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r240 = uitofp <16 x i1> poison to <16 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r241 = sitofp <16 x i1> poison to <16 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r242 = uitofp <16 x i8> poison to <16 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r243 = sitofp <16 x i8> poison to <16 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r244 = uitofp <16 x i16> poison to <16 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r245 = sitofp <16 x i16> poison to <16 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r246 = uitofp <16 x i32> poison to <16 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r247 = sitofp <16 x i32> poison to <16 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r248 = uitofp <16 x i64> poison to <16 x double> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r249 = sitofp <16 x i64> poison to <16 x double> ; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef ; ; FIXED-MIN-2048-LABEL: 'casts_no_users' -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui float undef to i1 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi float undef to i1 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui float undef to i8 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi float undef to i8 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui float undef to i16 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi float undef to i16 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui float undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi float undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui float undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi float undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r40 = fptoui double undef to i1 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r41 = fptosi double undef to i1 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r42 = fptoui double undef to i8 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r43 = fptosi double undef to i8 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r44 = fptoui double undef to i16 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r45 = fptosi double undef to i16 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r46 = fptoui double undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r47 = fptosi double undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r48 = fptoui double undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r49 = fptosi double undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r50 = sitofp i1 undef to float -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r51 = uitofp i1 undef to float -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r52 = sitofp i1 undef to double -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r53 = uitofp i1 undef to double -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r54 = sitofp i8 undef to float -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r55 = uitofp i8 undef to float -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r56 = sitofp i8 undef to double -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r57 = uitofp i8 undef to double -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r58 = sitofp i16 undef to float -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r59 = uitofp i16 undef to float -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r60 = sitofp i16 undef to double -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r61 = uitofp i16 undef to double -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r62 = sitofp i32 undef to float -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r63 = uitofp i32 undef to float -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r64 = sitofp i32 undef to double -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r65 = uitofp i32 undef to double -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r66 = sitofp i64 undef to float -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r67 = uitofp i64 undef to float -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r68 = sitofp i64 undef to double -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r69 = uitofp i64 undef to double -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r80 = fptrunc double undef to float -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r81 = fptrunc <2 x double> undef to <2 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r82 = fptrunc <4 x double> undef to <4 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r83 = fptrunc <8 x double> undef to <8 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r84 = fptrunc <16 x double> undef to <16 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r85 = fpext float undef to double -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r86 = fpext <2 x float> undef to <2 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r87 = fpext <4 x float> undef to <4 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r88 = fpext <8 x float> undef to <8 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r89 = fpext <16 x float> undef to <16 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x float> undef to <2 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x float> undef to <2 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x float> undef to <2 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x float> undef to <2 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x float> undef to <2 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x float> undef to <2 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x float> undef to <2 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x float> undef to <2 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x float> undef to <2 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x float> undef to <2 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r100 = fptoui <2 x double> undef to <2 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r101 = fptosi <2 x double> undef to <2 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r102 = fptoui <2 x double> undef to <2 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r103 = fptosi <2 x double> undef to <2 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r104 = fptoui <2 x double> undef to <2 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r105 = fptosi <2 x double> undef to <2 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r106 = fptoui <2 x double> undef to <2 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r107 = fptosi <2 x double> undef to <2 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r108 = fptoui <2 x double> undef to <2 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r109 = fptosi <2 x double> undef to <2 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r110 = fptoui <4 x float> undef to <4 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r111 = fptosi <4 x float> undef to <4 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r112 = fptoui <4 x float> undef to <4 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r113 = fptosi <4 x float> undef to <4 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r114 = fptoui <4 x float> undef to <4 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r115 = fptosi <4 x float> undef to <4 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r116 = fptoui <4 x float> undef to <4 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r117 = fptosi <4 x float> undef to <4 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r118 = fptoui <4 x float> undef to <4 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r119 = fptosi <4 x float> undef to <4 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r120 = fptoui <4 x double> undef to <4 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r121 = fptosi <4 x double> undef to <4 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r122 = fptoui <4 x double> undef to <4 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r123 = fptosi <4 x double> undef to <4 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r124 = fptoui <4 x double> undef to <4 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r125 = fptosi <4 x double> undef to <4 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r126 = fptoui <4 x double> undef to <4 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r127 = fptosi <4 x double> undef to <4 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r128 = fptoui <4 x double> undef to <4 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r129 = fptosi <4 x double> undef to <4 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r130 = fptoui <8 x float> undef to <8 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r131 = fptosi <8 x float> undef to <8 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r132 = fptoui <8 x float> undef to <8 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r133 = fptosi <8 x float> undef to <8 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x float> undef to <8 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x float> undef to <8 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r136 = fptoui <8 x float> undef to <8 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r137 = fptosi <8 x float> undef to <8 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r138 = fptoui <8 x float> undef to <8 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r139 = fptosi <8 x float> undef to <8 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r140 = fptoui <8 x double> undef to <8 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r141 = fptosi <8 x double> undef to <8 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r142 = fptoui <8 x double> undef to <8 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r143 = fptosi <8 x double> undef to <8 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r144 = fptoui <8 x double> undef to <8 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r145 = fptosi <8 x double> undef to <8 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r146 = fptoui <8 x double> undef to <8 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r147 = fptosi <8 x double> undef to <8 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r148 = fptoui <8 x double> undef to <8 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r149 = fptosi <8 x double> undef to <8 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r150 = fptoui <16 x float> undef to <16 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r151 = fptosi <16 x float> undef to <16 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r152 = fptoui <16 x float> undef to <16 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r153 = fptosi <16 x float> undef to <16 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r154 = fptoui <16 x float> undef to <16 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r155 = fptosi <16 x float> undef to <16 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r156 = fptoui <16 x float> undef to <16 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r157 = fptosi <16 x float> undef to <16 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r158 = fptoui <16 x float> undef to <16 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r159 = fptosi <16 x float> undef to <16 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r160 = fptoui <16 x double> undef to <16 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r161 = fptosi <16 x double> undef to <16 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r162 = fptoui <16 x double> undef to <16 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r163 = fptosi <16 x double> undef to <16 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r164 = fptoui <16 x double> undef to <16 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r165 = fptosi <16 x double> undef to <16 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r166 = fptoui <16 x double> undef to <16 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r167 = fptosi <16 x double> undef to <16 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r168 = fptoui <16 x double> undef to <16 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r169 = fptosi <16 x double> undef to <16 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r170 = uitofp <2 x i1> undef to <2 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r171 = sitofp <2 x i1> undef to <2 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r172 = uitofp <2 x i8> undef to <2 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r173 = sitofp <2 x i8> undef to <2 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r174 = uitofp <2 x i16> undef to <2 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r175 = sitofp <2 x i16> undef to <2 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r176 = uitofp <2 x i32> undef to <2 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r177 = sitofp <2 x i32> undef to <2 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r178 = uitofp <2 x i64> undef to <2 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r179 = sitofp <2 x i64> undef to <2 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r180 = uitofp <2 x i1> undef to <2 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r181 = sitofp <2 x i1> undef to <2 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r182 = uitofp <2 x i8> undef to <2 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r183 = sitofp <2 x i8> undef to <2 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r184 = uitofp <2 x i16> undef to <2 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r185 = sitofp <2 x i16> undef to <2 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r186 = uitofp <2 x i32> undef to <2 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r187 = sitofp <2 x i32> undef to <2 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r188 = uitofp <2 x i64> undef to <2 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r189 = sitofp <2 x i64> undef to <2 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r190 = uitofp <4 x i1> undef to <4 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r191 = sitofp <4 x i1> undef to <4 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r192 = uitofp <4 x i8> undef to <4 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r193 = sitofp <4 x i8> undef to <4 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r194 = uitofp <4 x i16> undef to <4 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r195 = sitofp <4 x i16> undef to <4 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r196 = uitofp <4 x i32> undef to <4 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r197 = sitofp <4 x i32> undef to <4 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r198 = uitofp <4 x i64> undef to <4 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r199 = sitofp <4 x i64> undef to <4 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r200 = uitofp <4 x i1> undef to <4 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r201 = sitofp <4 x i1> undef to <4 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r202 = uitofp <4 x i8> undef to <4 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r203 = sitofp <4 x i8> undef to <4 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r204 = uitofp <4 x i16> undef to <4 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r205 = sitofp <4 x i16> undef to <4 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r206 = uitofp <4 x i32> undef to <4 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r207 = sitofp <4 x i32> undef to <4 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r208 = uitofp <4 x i64> undef to <4 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r209 = sitofp <4 x i64> undef to <4 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r210 = uitofp <8 x i1> undef to <8 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r211 = sitofp <8 x i1> undef to <8 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r212 = uitofp <8 x i8> undef to <8 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r213 = sitofp <8 x i8> undef to <8 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r214 = uitofp <8 x i16> undef to <8 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r215 = sitofp <8 x i16> undef to <8 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r216 = uitofp <8 x i32> undef to <8 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r217 = sitofp <8 x i32> undef to <8 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r218 = uitofp <8 x i64> undef to <8 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r219 = sitofp <8 x i64> undef to <8 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r220 = uitofp <8 x i1> undef to <8 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r221 = sitofp <8 x i1> undef to <8 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r222 = uitofp <8 x i8> undef to <8 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r223 = sitofp <8 x i8> undef to <8 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r224 = uitofp <8 x i16> undef to <8 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r225 = sitofp <8 x i16> undef to <8 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r226 = uitofp <8 x i32> undef to <8 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r227 = sitofp <8 x i32> undef to <8 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r228 = uitofp <8 x i64> undef to <8 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r229 = sitofp <8 x i64> undef to <8 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r230 = uitofp <16 x i1> undef to <16 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r231 = sitofp <16 x i1> undef to <16 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r232 = uitofp <16 x i8> undef to <16 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r233 = sitofp <16 x i8> undef to <16 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r234 = uitofp <16 x i16> undef to <16 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r235 = sitofp <16 x i16> undef to <16 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r236 = uitofp <16 x i32> undef to <16 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r237 = sitofp <16 x i32> undef to <16 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r238 = uitofp <16 x i64> undef to <16 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r239 = sitofp <16 x i64> undef to <16 x float> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r240 = uitofp <16 x i1> undef to <16 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r241 = sitofp <16 x i1> undef to <16 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r242 = uitofp <16 x i8> undef to <16 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r243 = sitofp <16 x i8> undef to <16 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r244 = uitofp <16 x i16> undef to <16 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r245 = sitofp <16 x i16> undef to <16 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r246 = uitofp <16 x i32> undef to <16 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r247 = sitofp <16 x i32> undef to <16 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r248 = uitofp <16 x i64> undef to <16 x double> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r249 = sitofp <16 x i64> undef to <16 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui float poison to i1 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi float poison to i1 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui float poison to i8 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi float poison to i8 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui float poison to i16 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi float poison to i16 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui float poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi float poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui float poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi float poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r40 = fptoui double poison to i1 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r41 = fptosi double poison to i1 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r42 = fptoui double poison to i8 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r43 = fptosi double poison to i8 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r44 = fptoui double poison to i16 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r45 = fptosi double poison to i16 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r46 = fptoui double poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r47 = fptosi double poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r48 = fptoui double poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r49 = fptosi double poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r50 = sitofp i1 poison to float +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r51 = uitofp i1 poison to float +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r52 = sitofp i1 poison to double +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r53 = uitofp i1 poison to double +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r54 = sitofp i8 poison to float +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r55 = uitofp i8 poison to float +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r56 = sitofp i8 poison to double +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r57 = uitofp i8 poison to double +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r58 = sitofp i16 poison to float +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r59 = uitofp i16 poison to float +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r60 = sitofp i16 poison to double +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r61 = uitofp i16 poison to double +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r62 = sitofp i32 poison to float +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r63 = uitofp i32 poison to float +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r64 = sitofp i32 poison to double +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r65 = uitofp i32 poison to double +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r66 = sitofp i64 poison to float +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r67 = uitofp i64 poison to float +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r68 = sitofp i64 poison to double +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r69 = uitofp i64 poison to double +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r80 = fptrunc double poison to float +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r81 = fptrunc <2 x double> poison to <2 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r82 = fptrunc <4 x double> poison to <4 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r83 = fptrunc <8 x double> poison to <8 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r84 = fptrunc <16 x double> poison to <16 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r85 = fpext float poison to double +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r86 = fpext <2 x float> poison to <2 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r87 = fpext <4 x float> poison to <4 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r88 = fpext <8 x float> poison to <8 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r89 = fpext <16 x float> poison to <16 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x float> poison to <2 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x float> poison to <2 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x float> poison to <2 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x float> poison to <2 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x float> poison to <2 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x float> poison to <2 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x float> poison to <2 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x float> poison to <2 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x float> poison to <2 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x float> poison to <2 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r100 = fptoui <2 x double> poison to <2 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r101 = fptosi <2 x double> poison to <2 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r102 = fptoui <2 x double> poison to <2 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r103 = fptosi <2 x double> poison to <2 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r104 = fptoui <2 x double> poison to <2 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r105 = fptosi <2 x double> poison to <2 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r106 = fptoui <2 x double> poison to <2 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r107 = fptosi <2 x double> poison to <2 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r108 = fptoui <2 x double> poison to <2 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r109 = fptosi <2 x double> poison to <2 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r110 = fptoui <4 x float> poison to <4 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:20 CodeSize:1 Lat:1 SizeLat:1 for: %r111 = fptosi <4 x float> poison to <4 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r112 = fptoui <4 x float> poison to <4 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r113 = fptosi <4 x float> poison to <4 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r114 = fptoui <4 x float> poison to <4 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r115 = fptosi <4 x float> poison to <4 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r116 = fptoui <4 x float> poison to <4 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r117 = fptosi <4 x float> poison to <4 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r118 = fptoui <4 x float> poison to <4 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r119 = fptosi <4 x float> poison to <4 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r120 = fptoui <4 x double> poison to <4 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r121 = fptosi <4 x double> poison to <4 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r122 = fptoui <4 x double> poison to <4 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r123 = fptosi <4 x double> poison to <4 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r124 = fptoui <4 x double> poison to <4 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r125 = fptosi <4 x double> poison to <4 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r126 = fptoui <4 x double> poison to <4 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r127 = fptosi <4 x double> poison to <4 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r128 = fptoui <4 x double> poison to <4 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r129 = fptosi <4 x double> poison to <4 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r130 = fptoui <8 x float> poison to <8 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r131 = fptosi <8 x float> poison to <8 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r132 = fptoui <8 x float> poison to <8 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r133 = fptosi <8 x float> poison to <8 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x float> poison to <8 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x float> poison to <8 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r136 = fptoui <8 x float> poison to <8 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r137 = fptosi <8 x float> poison to <8 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r138 = fptoui <8 x float> poison to <8 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r139 = fptosi <8 x float> poison to <8 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r140 = fptoui <8 x double> poison to <8 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r141 = fptosi <8 x double> poison to <8 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r142 = fptoui <8 x double> poison to <8 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r143 = fptosi <8 x double> poison to <8 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r144 = fptoui <8 x double> poison to <8 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r145 = fptosi <8 x double> poison to <8 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r146 = fptoui <8 x double> poison to <8 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r147 = fptosi <8 x double> poison to <8 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r148 = fptoui <8 x double> poison to <8 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r149 = fptosi <8 x double> poison to <8 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r150 = fptoui <16 x float> poison to <16 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r151 = fptosi <16 x float> poison to <16 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r152 = fptoui <16 x float> poison to <16 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r153 = fptosi <16 x float> poison to <16 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r154 = fptoui <16 x float> poison to <16 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r155 = fptosi <16 x float> poison to <16 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r156 = fptoui <16 x float> poison to <16 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r157 = fptosi <16 x float> poison to <16 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r158 = fptoui <16 x float> poison to <16 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r159 = fptosi <16 x float> poison to <16 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r160 = fptoui <16 x double> poison to <16 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r161 = fptosi <16 x double> poison to <16 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r162 = fptoui <16 x double> poison to <16 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r163 = fptosi <16 x double> poison to <16 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r164 = fptoui <16 x double> poison to <16 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r165 = fptosi <16 x double> poison to <16 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r166 = fptoui <16 x double> poison to <16 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r167 = fptosi <16 x double> poison to <16 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r168 = fptoui <16 x double> poison to <16 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r169 = fptosi <16 x double> poison to <16 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r170 = uitofp <2 x i1> poison to <2 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r171 = sitofp <2 x i1> poison to <2 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r172 = uitofp <2 x i8> poison to <2 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r173 = sitofp <2 x i8> poison to <2 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r174 = uitofp <2 x i16> poison to <2 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r175 = sitofp <2 x i16> poison to <2 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r176 = uitofp <2 x i32> poison to <2 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r177 = sitofp <2 x i32> poison to <2 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r178 = uitofp <2 x i64> poison to <2 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r179 = sitofp <2 x i64> poison to <2 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r180 = uitofp <2 x i1> poison to <2 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r181 = sitofp <2 x i1> poison to <2 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r182 = uitofp <2 x i8> poison to <2 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r183 = sitofp <2 x i8> poison to <2 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r184 = uitofp <2 x i16> poison to <2 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r185 = sitofp <2 x i16> poison to <2 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r186 = uitofp <2 x i32> poison to <2 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r187 = sitofp <2 x i32> poison to <2 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r188 = uitofp <2 x i64> poison to <2 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r189 = sitofp <2 x i64> poison to <2 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r190 = uitofp <4 x i1> poison to <4 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r191 = sitofp <4 x i1> poison to <4 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r192 = uitofp <4 x i8> poison to <4 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r193 = sitofp <4 x i8> poison to <4 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r194 = uitofp <4 x i16> poison to <4 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r195 = sitofp <4 x i16> poison to <4 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r196 = uitofp <4 x i32> poison to <4 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r197 = sitofp <4 x i32> poison to <4 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r198 = uitofp <4 x i64> poison to <4 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r199 = sitofp <4 x i64> poison to <4 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r200 = uitofp <4 x i1> poison to <4 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r201 = sitofp <4 x i1> poison to <4 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r202 = uitofp <4 x i8> poison to <4 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r203 = sitofp <4 x i8> poison to <4 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r204 = uitofp <4 x i16> poison to <4 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r205 = sitofp <4 x i16> poison to <4 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r206 = uitofp <4 x i32> poison to <4 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r207 = sitofp <4 x i32> poison to <4 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r208 = uitofp <4 x i64> poison to <4 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r209 = sitofp <4 x i64> poison to <4 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r210 = uitofp <8 x i1> poison to <8 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r211 = sitofp <8 x i1> poison to <8 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r212 = uitofp <8 x i8> poison to <8 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r213 = sitofp <8 x i8> poison to <8 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r214 = uitofp <8 x i16> poison to <8 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r215 = sitofp <8 x i16> poison to <8 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r216 = uitofp <8 x i32> poison to <8 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r217 = sitofp <8 x i32> poison to <8 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r218 = uitofp <8 x i64> poison to <8 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r219 = sitofp <8 x i64> poison to <8 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r220 = uitofp <8 x i1> poison to <8 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r221 = sitofp <8 x i1> poison to <8 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r222 = uitofp <8 x i8> poison to <8 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r223 = sitofp <8 x i8> poison to <8 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r224 = uitofp <8 x i16> poison to <8 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r225 = sitofp <8 x i16> poison to <8 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r226 = uitofp <8 x i32> poison to <8 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r227 = sitofp <8 x i32> poison to <8 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r228 = uitofp <8 x i64> poison to <8 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r229 = sitofp <8 x i64> poison to <8 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r230 = uitofp <16 x i1> poison to <16 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r231 = sitofp <16 x i1> poison to <16 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r232 = uitofp <16 x i8> poison to <16 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r233 = sitofp <16 x i8> poison to <16 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r234 = uitofp <16 x i16> poison to <16 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r235 = sitofp <16 x i16> poison to <16 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r236 = uitofp <16 x i32> poison to <16 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r237 = sitofp <16 x i32> poison to <16 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r238 = uitofp <16 x i64> poison to <16 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r239 = sitofp <16 x i64> poison to <16 x float> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r240 = uitofp <16 x i1> poison to <16 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r241 = sitofp <16 x i1> poison to <16 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r242 = uitofp <16 x i8> poison to <16 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r243 = sitofp <16 x i8> poison to <16 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r244 = uitofp <16 x i16> poison to <16 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r245 = sitofp <16 x i16> poison to <16 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r246 = uitofp <16 x i32> poison to <16 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r247 = sitofp <16 x i32> poison to <16 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r248 = uitofp <16 x i64> poison to <16 x double> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r249 = sitofp <16 x i64> poison to <16 x double> ; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef ; - %r30 = fptoui float undef to i1 - %r31 = fptosi float undef to i1 - %r32 = fptoui float undef to i8 - %r33 = fptosi float undef to i8 - %r34 = fptoui float undef to i16 - %r35 = fptosi float undef to i16 - %r36 = fptoui float undef to i32 - %r37 = fptosi float undef to i32 - %r38 = fptoui float undef to i64 - %r39 = fptosi float undef to i64 - %r40 = fptoui double undef to i1 - %r41 = fptosi double undef to i1 - %r42 = fptoui double undef to i8 - %r43 = fptosi double undef to i8 - %r44 = fptoui double undef to i16 - %r45 = fptosi double undef to i16 - %r46 = fptoui double undef to i32 - %r47 = fptosi double undef to i32 - %r48 = fptoui double undef to i64 - %r49 = fptosi double undef to i64 - %r50 = sitofp i1 undef to float - %r51 = uitofp i1 undef to float - %r52 = sitofp i1 undef to double - %r53 = uitofp i1 undef to double - %r54 = sitofp i8 undef to float - %r55 = uitofp i8 undef to float - %r56 = sitofp i8 undef to double - %r57 = uitofp i8 undef to double - %r58 = sitofp i16 undef to float - %r59 = uitofp i16 undef to float - %r60 = sitofp i16 undef to double - %r61 = uitofp i16 undef to double - %r62 = sitofp i32 undef to float - %r63 = uitofp i32 undef to float - %r64 = sitofp i32 undef to double - %r65 = uitofp i32 undef to double - %r66 = sitofp i64 undef to float - %r67 = uitofp i64 undef to float - %r68 = sitofp i64 undef to double - %r69 = uitofp i64 undef to double - %r80 = fptrunc double undef to float - %r81 = fptrunc <2 x double> undef to <2 x float> - %r82 = fptrunc <4 x double> undef to <4 x float> - %r83 = fptrunc <8 x double> undef to <8 x float> - %r84 = fptrunc <16 x double> undef to <16 x float> - %r85 = fpext float undef to double - %r86 = fpext <2 x float> undef to <2 x double> - %r87 = fpext <4 x float> undef to <4 x double> - %r88 = fpext <8 x float> undef to <8 x double> - %r89 = fpext <16 x float> undef to <16 x double> - %r90 = fptoui <2 x float> undef to <2 x i1> - %r91 = fptosi <2 x float> undef to <2 x i1> - %r92 = fptoui <2 x float> undef to <2 x i8> - %r93 = fptosi <2 x float> undef to <2 x i8> - %r94 = fptoui <2 x float> undef to <2 x i16> - %r95 = fptosi <2 x float> undef to <2 x i16> - %r96 = fptoui <2 x float> undef to <2 x i32> - %r97 = fptosi <2 x float> undef to <2 x i32> - %r98 = fptoui <2 x float> undef to <2 x i64> - %r99 = fptosi <2 x float> undef to <2 x i64> - %r100 = fptoui <2 x double> undef to <2 x i1> - %r101 = fptosi <2 x double> undef to <2 x i1> - %r102 = fptoui <2 x double> undef to <2 x i8> - %r103 = fptosi <2 x double> undef to <2 x i8> - %r104 = fptoui <2 x double> undef to <2 x i16> - %r105 = fptosi <2 x double> undef to <2 x i16> - %r106 = fptoui <2 x double> undef to <2 x i32> - %r107 = fptosi <2 x double> undef to <2 x i32> - %r108 = fptoui <2 x double> undef to <2 x i64> - %r109 = fptosi <2 x double> undef to <2 x i64> + %r30 = fptoui float poison to i1 + %r31 = fptosi float poison to i1 + %r32 = fptoui float poison to i8 + %r33 = fptosi float poison to i8 + %r34 = fptoui float poison to i16 + %r35 = fptosi float poison to i16 + %r36 = fptoui float poison to i32 + %r37 = fptosi float poison to i32 + %r38 = fptoui float poison to i64 + %r39 = fptosi float poison to i64 + %r40 = fptoui double poison to i1 + %r41 = fptosi double poison to i1 + %r42 = fptoui double poison to i8 + %r43 = fptosi double poison to i8 + %r44 = fptoui double poison to i16 + %r45 = fptosi double poison to i16 + %r46 = fptoui double poison to i32 + %r47 = fptosi double poison to i32 + %r48 = fptoui double poison to i64 + %r49 = fptosi double poison to i64 + %r50 = sitofp i1 poison to float + %r51 = uitofp i1 poison to float + %r52 = sitofp i1 poison to double + %r53 = uitofp i1 poison to double + %r54 = sitofp i8 poison to float + %r55 = uitofp i8 poison to float + %r56 = sitofp i8 poison to double + %r57 = uitofp i8 poison to double + %r58 = sitofp i16 poison to float + %r59 = uitofp i16 poison to float + %r60 = sitofp i16 poison to double + %r61 = uitofp i16 poison to double + %r62 = sitofp i32 poison to float + %r63 = uitofp i32 poison to float + %r64 = sitofp i32 poison to double + %r65 = uitofp i32 poison to double + %r66 = sitofp i64 poison to float + %r67 = uitofp i64 poison to float + %r68 = sitofp i64 poison to double + %r69 = uitofp i64 poison to double + %r80 = fptrunc double poison to float + %r81 = fptrunc <2 x double> poison to <2 x float> + %r82 = fptrunc <4 x double> poison to <4 x float> + %r83 = fptrunc <8 x double> poison to <8 x float> + %r84 = fptrunc <16 x double> poison to <16 x float> + %r85 = fpext float poison to double + %r86 = fpext <2 x float> poison to <2 x double> + %r87 = fpext <4 x float> poison to <4 x double> + %r88 = fpext <8 x float> poison to <8 x double> + %r89 = fpext <16 x float> poison to <16 x double> + %r90 = fptoui <2 x float> poison to <2 x i1> + %r91 = fptosi <2 x float> poison to <2 x i1> + %r92 = fptoui <2 x float> poison to <2 x i8> + %r93 = fptosi <2 x float> poison to <2 x i8> + %r94 = fptoui <2 x float> poison to <2 x i16> + %r95 = fptosi <2 x float> poison to <2 x i16> + %r96 = fptoui <2 x float> poison to <2 x i32> + %r97 = fptosi <2 x float> poison to <2 x i32> + %r98 = fptoui <2 x float> poison to <2 x i64> + %r99 = fptosi <2 x float> poison to <2 x i64> + %r100 = fptoui <2 x double> poison to <2 x i1> + %r101 = fptosi <2 x double> poison to <2 x i1> + %r102 = fptoui <2 x double> poison to <2 x i8> + %r103 = fptosi <2 x double> poison to <2 x i8> + %r104 = fptoui <2 x double> poison to <2 x i16> + %r105 = fptosi <2 x double> poison to <2 x i16> + %r106 = fptoui <2 x double> poison to <2 x i32> + %r107 = fptosi <2 x double> poison to <2 x i32> + %r108 = fptoui <2 x double> poison to <2 x i64> + %r109 = fptosi <2 x double> poison to <2 x i64> - %r110 = fptoui <4 x float> undef to <4 x i1> - %r111 = fptosi <4 x float> undef to <4 x i1> - %r112 = fptoui <4 x float> undef to <4 x i8> - %r113 = fptosi <4 x float> undef to <4 x i8> - %r114 = fptoui <4 x float> undef to <4 x i16> - %r115 = fptosi <4 x float> undef to <4 x i16> - %r116 = fptoui <4 x float> undef to <4 x i32> - %r117 = fptosi <4 x float> undef to <4 x i32> - %r118 = fptoui <4 x float> undef to <4 x i64> - %r119 = fptosi <4 x float> undef to <4 x i64> + %r110 = fptoui <4 x float> poison to <4 x i1> + %r111 = fptosi <4 x float> poison to <4 x i1> + %r112 = fptoui <4 x float> poison to <4 x i8> + %r113 = fptosi <4 x float> poison to <4 x i8> + %r114 = fptoui <4 x float> poison to <4 x i16> + %r115 = fptosi <4 x float> poison to <4 x i16> + %r116 = fptoui <4 x float> poison to <4 x i32> + %r117 = fptosi <4 x float> poison to <4 x i32> + %r118 = fptoui <4 x float> poison to <4 x i64> + %r119 = fptosi <4 x float> poison to <4 x i64> - %r120 = fptoui <4 x double> undef to <4 x i1> - %r121 = fptosi <4 x double> undef to <4 x i1> - %r122 = fptoui <4 x double> undef to <4 x i8> - %r123 = fptosi <4 x double> undef to <4 x i8> - %r124 = fptoui <4 x double> undef to <4 x i16> - %r125 = fptosi <4 x double> undef to <4 x i16> - %r126 = fptoui <4 x double> undef to <4 x i32> - %r127 = fptosi <4 x double> undef to <4 x i32> - %r128 = fptoui <4 x double> undef to <4 x i64> - %r129 = fptosi <4 x double> undef to <4 x i64> + %r120 = fptoui <4 x double> poison to <4 x i1> + %r121 = fptosi <4 x double> poison to <4 x i1> + %r122 = fptoui <4 x double> poison to <4 x i8> + %r123 = fptosi <4 x double> poison to <4 x i8> + %r124 = fptoui <4 x double> poison to <4 x i16> + %r125 = fptosi <4 x double> poison to <4 x i16> + %r126 = fptoui <4 x double> poison to <4 x i32> + %r127 = fptosi <4 x double> poison to <4 x i32> + %r128 = fptoui <4 x double> poison to <4 x i64> + %r129 = fptosi <4 x double> poison to <4 x i64> - %r130 = fptoui <8 x float> undef to <8 x i1> - %r131 = fptosi <8 x float> undef to <8 x i1> - %r132 = fptoui <8 x float> undef to <8 x i8> - %r133 = fptosi <8 x float> undef to <8 x i8> - %r134 = fptoui <8 x float> undef to <8 x i16> - %r135 = fptosi <8 x float> undef to <8 x i16> - %r136 = fptoui <8 x float> undef to <8 x i32> - %r137 = fptosi <8 x float> undef to <8 x i32> - %r138 = fptoui <8 x float> undef to <8 x i64> - %r139 = fptosi <8 x float> undef to <8 x i64> + %r130 = fptoui <8 x float> poison to <8 x i1> + %r131 = fptosi <8 x float> poison to <8 x i1> + %r132 = fptoui <8 x float> poison to <8 x i8> + %r133 = fptosi <8 x float> poison to <8 x i8> + %r134 = fptoui <8 x float> poison to <8 x i16> + %r135 = fptosi <8 x float> poison to <8 x i16> + %r136 = fptoui <8 x float> poison to <8 x i32> + %r137 = fptosi <8 x float> poison to <8 x i32> + %r138 = fptoui <8 x float> poison to <8 x i64> + %r139 = fptosi <8 x float> poison to <8 x i64> - %r140 = fptoui <8 x double> undef to <8 x i1> - %r141 = fptosi <8 x double> undef to <8 x i1> - %r142 = fptoui <8 x double> undef to <8 x i8> - %r143 = fptosi <8 x double> undef to <8 x i8> - %r144 = fptoui <8 x double> undef to <8 x i16> - %r145 = fptosi <8 x double> undef to <8 x i16> - %r146 = fptoui <8 x double> undef to <8 x i32> - %r147 = fptosi <8 x double> undef to <8 x i32> - %r148 = fptoui <8 x double> undef to <8 x i64> - %r149 = fptosi <8 x double> undef to <8 x i64> + %r140 = fptoui <8 x double> poison to <8 x i1> + %r141 = fptosi <8 x double> poison to <8 x i1> + %r142 = fptoui <8 x double> poison to <8 x i8> + %r143 = fptosi <8 x double> poison to <8 x i8> + %r144 = fptoui <8 x double> poison to <8 x i16> + %r145 = fptosi <8 x double> poison to <8 x i16> + %r146 = fptoui <8 x double> poison to <8 x i32> + %r147 = fptosi <8 x double> poison to <8 x i32> + %r148 = fptoui <8 x double> poison to <8 x i64> + %r149 = fptosi <8 x double> poison to <8 x i64> - %r150 = fptoui <16 x float> undef to <16 x i1> - %r151 = fptosi <16 x float> undef to <16 x i1> - %r152 = fptoui <16 x float> undef to <16 x i8> - %r153 = fptosi <16 x float> undef to <16 x i8> - %r154 = fptoui <16 x float> undef to <16 x i16> - %r155 = fptosi <16 x float> undef to <16 x i16> - %r156 = fptoui <16 x float> undef to <16 x i32> - %r157 = fptosi <16 x float> undef to <16 x i32> - %r158 = fptoui <16 x float> undef to <16 x i64> - %r159 = fptosi <16 x float> undef to <16 x i64> + %r150 = fptoui <16 x float> poison to <16 x i1> + %r151 = fptosi <16 x float> poison to <16 x i1> + %r152 = fptoui <16 x float> poison to <16 x i8> + %r153 = fptosi <16 x float> poison to <16 x i8> + %r154 = fptoui <16 x float> poison to <16 x i16> + %r155 = fptosi <16 x float> poison to <16 x i16> + %r156 = fptoui <16 x float> poison to <16 x i32> + %r157 = fptosi <16 x float> poison to <16 x i32> + %r158 = fptoui <16 x float> poison to <16 x i64> + %r159 = fptosi <16 x float> poison to <16 x i64> - %r160 = fptoui <16 x double> undef to <16 x i1> - %r161 = fptosi <16 x double> undef to <16 x i1> - %r162 = fptoui <16 x double> undef to <16 x i8> - %r163 = fptosi <16 x double> undef to <16 x i8> - %r164 = fptoui <16 x double> undef to <16 x i16> - %r165 = fptosi <16 x double> undef to <16 x i16> - %r166 = fptoui <16 x double> undef to <16 x i32> - %r167 = fptosi <16 x double> undef to <16 x i32> - %r168 = fptoui <16 x double> undef to <16 x i64> - %r169 = fptosi <16 x double> undef to <16 x i64> + %r160 = fptoui <16 x double> poison to <16 x i1> + %r161 = fptosi <16 x double> poison to <16 x i1> + %r162 = fptoui <16 x double> poison to <16 x i8> + %r163 = fptosi <16 x double> poison to <16 x i8> + %r164 = fptoui <16 x double> poison to <16 x i16> + %r165 = fptosi <16 x double> poison to <16 x i16> + %r166 = fptoui <16 x double> poison to <16 x i32> + %r167 = fptosi <16 x double> poison to <16 x i32> + %r168 = fptoui <16 x double> poison to <16 x i64> + %r169 = fptosi <16 x double> poison to <16 x i64> - %r170 = uitofp <2 x i1> undef to <2 x float> - %r171 = sitofp <2 x i1> undef to <2 x float> - %r172 = uitofp <2 x i8> undef to <2 x float> - %r173 = sitofp <2 x i8> undef to <2 x float> - %r174 = uitofp <2 x i16> undef to <2 x float> - %r175 = sitofp <2 x i16> undef to <2 x float> - %r176 = uitofp <2 x i32> undef to <2 x float> - %r177 = sitofp <2 x i32> undef to <2 x float> - %r178 = uitofp <2 x i64> undef to <2 x float> - %r179 = sitofp <2 x i64> undef to <2 x float> + %r170 = uitofp <2 x i1> poison to <2 x float> + %r171 = sitofp <2 x i1> poison to <2 x float> + %r172 = uitofp <2 x i8> poison to <2 x float> + %r173 = sitofp <2 x i8> poison to <2 x float> + %r174 = uitofp <2 x i16> poison to <2 x float> + %r175 = sitofp <2 x i16> poison to <2 x float> + %r176 = uitofp <2 x i32> poison to <2 x float> + %r177 = sitofp <2 x i32> poison to <2 x float> + %r178 = uitofp <2 x i64> poison to <2 x float> + %r179 = sitofp <2 x i64> poison to <2 x float> - %r180 = uitofp <2 x i1> undef to <2 x double> - %r181 = sitofp <2 x i1> undef to <2 x double> - %r182 = uitofp <2 x i8> undef to <2 x double> - %r183 = sitofp <2 x i8> undef to <2 x double> - %r184 = uitofp <2 x i16> undef to <2 x double> - %r185 = sitofp <2 x i16> undef to <2 x double> - %r186 = uitofp <2 x i32> undef to <2 x double> - %r187 = sitofp <2 x i32> undef to <2 x double> - %r188 = uitofp <2 x i64> undef to <2 x double> - %r189 = sitofp <2 x i64> undef to <2 x double> + %r180 = uitofp <2 x i1> poison to <2 x double> + %r181 = sitofp <2 x i1> poison to <2 x double> + %r182 = uitofp <2 x i8> poison to <2 x double> + %r183 = sitofp <2 x i8> poison to <2 x double> + %r184 = uitofp <2 x i16> poison to <2 x double> + %r185 = sitofp <2 x i16> poison to <2 x double> + %r186 = uitofp <2 x i32> poison to <2 x double> + %r187 = sitofp <2 x i32> poison to <2 x double> + %r188 = uitofp <2 x i64> poison to <2 x double> + %r189 = sitofp <2 x i64> poison to <2 x double> - %r190 = uitofp <4 x i1> undef to <4 x float> - %r191 = sitofp <4 x i1> undef to <4 x float> - %r192 = uitofp <4 x i8> undef to <4 x float> - %r193 = sitofp <4 x i8> undef to <4 x float> - %r194 = uitofp <4 x i16> undef to <4 x float> - %r195 = sitofp <4 x i16> undef to <4 x float> - %r196 = uitofp <4 x i32> undef to <4 x float> - %r197 = sitofp <4 x i32> undef to <4 x float> - %r198 = uitofp <4 x i64> undef to <4 x float> - %r199 = sitofp <4 x i64> undef to <4 x float> + %r190 = uitofp <4 x i1> poison to <4 x float> + %r191 = sitofp <4 x i1> poison to <4 x float> + %r192 = uitofp <4 x i8> poison to <4 x float> + %r193 = sitofp <4 x i8> poison to <4 x float> + %r194 = uitofp <4 x i16> poison to <4 x float> + %r195 = sitofp <4 x i16> poison to <4 x float> + %r196 = uitofp <4 x i32> poison to <4 x float> + %r197 = sitofp <4 x i32> poison to <4 x float> + %r198 = uitofp <4 x i64> poison to <4 x float> + %r199 = sitofp <4 x i64> poison to <4 x float> - %r200 = uitofp <4 x i1> undef to <4 x double> - %r201 = sitofp <4 x i1> undef to <4 x double> - %r202 = uitofp <4 x i8> undef to <4 x double> - %r203 = sitofp <4 x i8> undef to <4 x double> - %r204 = uitofp <4 x i16> undef to <4 x double> - %r205 = sitofp <4 x i16> undef to <4 x double> - %r206 = uitofp <4 x i32> undef to <4 x double> - %r207 = sitofp <4 x i32> undef to <4 x double> - %r208 = uitofp <4 x i64> undef to <4 x double> - %r209 = sitofp <4 x i64> undef to <4 x double> + %r200 = uitofp <4 x i1> poison to <4 x double> + %r201 = sitofp <4 x i1> poison to <4 x double> + %r202 = uitofp <4 x i8> poison to <4 x double> + %r203 = sitofp <4 x i8> poison to <4 x double> + %r204 = uitofp <4 x i16> poison to <4 x double> + %r205 = sitofp <4 x i16> poison to <4 x double> + %r206 = uitofp <4 x i32> poison to <4 x double> + %r207 = sitofp <4 x i32> poison to <4 x double> + %r208 = uitofp <4 x i64> poison to <4 x double> + %r209 = sitofp <4 x i64> poison to <4 x double> - %r210 = uitofp <8 x i1> undef to <8 x float> - %r211 = sitofp <8 x i1> undef to <8 x float> - %r212 = uitofp <8 x i8> undef to <8 x float> - %r213 = sitofp <8 x i8> undef to <8 x float> - %r214 = uitofp <8 x i16> undef to <8 x float> - %r215 = sitofp <8 x i16> undef to <8 x float> - %r216 = uitofp <8 x i32> undef to <8 x float> - %r217 = sitofp <8 x i32> undef to <8 x float> - %r218 = uitofp <8 x i64> undef to <8 x float> - %r219 = sitofp <8 x i64> undef to <8 x float> + %r210 = uitofp <8 x i1> poison to <8 x float> + %r211 = sitofp <8 x i1> poison to <8 x float> + %r212 = uitofp <8 x i8> poison to <8 x float> + %r213 = sitofp <8 x i8> poison to <8 x float> + %r214 = uitofp <8 x i16> poison to <8 x float> + %r215 = sitofp <8 x i16> poison to <8 x float> + %r216 = uitofp <8 x i32> poison to <8 x float> + %r217 = sitofp <8 x i32> poison to <8 x float> + %r218 = uitofp <8 x i64> poison to <8 x float> + %r219 = sitofp <8 x i64> poison to <8 x float> - %r220 = uitofp <8 x i1> undef to <8 x double> - %r221 = sitofp <8 x i1> undef to <8 x double> - %r222 = uitofp <8 x i8> undef to <8 x double> - %r223 = sitofp <8 x i8> undef to <8 x double> - %r224 = uitofp <8 x i16> undef to <8 x double> - %r225 = sitofp <8 x i16> undef to <8 x double> - %r226 = uitofp <8 x i32> undef to <8 x double> - %r227 = sitofp <8 x i32> undef to <8 x double> - %r228 = uitofp <8 x i64> undef to <8 x double> - %r229 = sitofp <8 x i64> undef to <8 x double> + %r220 = uitofp <8 x i1> poison to <8 x double> + %r221 = sitofp <8 x i1> poison to <8 x double> + %r222 = uitofp <8 x i8> poison to <8 x double> + %r223 = sitofp <8 x i8> poison to <8 x double> + %r224 = uitofp <8 x i16> poison to <8 x double> + %r225 = sitofp <8 x i16> poison to <8 x double> + %r226 = uitofp <8 x i32> poison to <8 x double> + %r227 = sitofp <8 x i32> poison to <8 x double> + %r228 = uitofp <8 x i64> poison to <8 x double> + %r229 = sitofp <8 x i64> poison to <8 x double> - %r230 = uitofp <16 x i1> undef to <16 x float> - %r231 = sitofp <16 x i1> undef to <16 x float> - %r232 = uitofp <16 x i8> undef to <16 x float> - %r233 = sitofp <16 x i8> undef to <16 x float> - %r234 = uitofp <16 x i16> undef to <16 x float> - %r235 = sitofp <16 x i16> undef to <16 x float> - %r236 = uitofp <16 x i32> undef to <16 x float> - %r237 = sitofp <16 x i32> undef to <16 x float> - %r238 = uitofp <16 x i64> undef to <16 x float> - %r239 = sitofp <16 x i64> undef to <16 x float> + %r230 = uitofp <16 x i1> poison to <16 x float> + %r231 = sitofp <16 x i1> poison to <16 x float> + %r232 = uitofp <16 x i8> poison to <16 x float> + %r233 = sitofp <16 x i8> poison to <16 x float> + %r234 = uitofp <16 x i16> poison to <16 x float> + %r235 = sitofp <16 x i16> poison to <16 x float> + %r236 = uitofp <16 x i32> poison to <16 x float> + %r237 = sitofp <16 x i32> poison to <16 x float> + %r238 = uitofp <16 x i64> poison to <16 x float> + %r239 = sitofp <16 x i64> poison to <16 x float> - %r240 = uitofp <16 x i1> undef to <16 x double> - %r241 = sitofp <16 x i1> undef to <16 x double> - %r242 = uitofp <16 x i8> undef to <16 x double> - %r243 = sitofp <16 x i8> undef to <16 x double> - %r244 = uitofp <16 x i16> undef to <16 x double> - %r245 = sitofp <16 x i16> undef to <16 x double> - %r246 = uitofp <16 x i32> undef to <16 x double> - %r247 = sitofp <16 x i32> undef to <16 x double> - %r248 = uitofp <16 x i64> undef to <16 x double> - %r249 = sitofp <16 x i64> undef to <16 x double> + %r240 = uitofp <16 x i1> poison to <16 x double> + %r241 = sitofp <16 x i1> poison to <16 x double> + %r242 = uitofp <16 x i8> poison to <16 x double> + %r243 = sitofp <16 x i8> poison to <16 x double> + %r244 = uitofp <16 x i16> poison to <16 x double> + %r245 = sitofp <16 x i16> poison to <16 x double> + %r246 = uitofp <16 x i32> poison to <16 x double> + %r247 = sitofp <16 x i32> poison to <16 x double> + %r248 = uitofp <16 x i64> poison to <16 x double> + %r249 = sitofp <16 x i64> poison to <16 x double> ret i32 undef } @@ -1760,24 +1760,24 @@ define i32 @casts_with_users(i8 %a, i16 %b, i32 %c, i64 %d, i1 %e) { define i32 @bitcasts() { ; CHECK-LABEL: 'bitcasts' -; CHECK-NEXT: Cost Model: Found costs of 0 for: %a = bitcast i32 undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %b = bitcast float undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %c = bitcast i32 undef to float -; CHECK-NEXT: Cost Model: Found costs of 1 for: %d = bitcast float undef to i32 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %e = bitcast i64 undef to double -; CHECK-NEXT: Cost Model: Found costs of 1 for: %f = bitcast double undef to i64 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %g = bitcast half undef to i16 -; CHECK-NEXT: Cost Model: Found costs of 1 for: %h = bitcast i16 undef to half +; CHECK-NEXT: Cost Model: Found costs of 0 for: %a = bitcast i32 poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %b = bitcast float poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %c = bitcast i32 poison to float +; CHECK-NEXT: Cost Model: Found costs of 1 for: %d = bitcast float poison to i32 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %e = bitcast i64 poison to double +; CHECK-NEXT: Cost Model: Found costs of 1 for: %f = bitcast double poison to i64 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %g = bitcast half poison to i16 +; CHECK-NEXT: Cost Model: Found costs of 1 for: %h = bitcast i16 poison to half ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef ; - %a = bitcast i32 undef to i32 - %b = bitcast float undef to float - %c = bitcast i32 undef to float - %d = bitcast float undef to i32 - %e = bitcast i64 undef to double - %f = bitcast double undef to i64 - %g = bitcast half undef to i16 - %h = bitcast i16 undef to half + %a = bitcast i32 poison to i32 + %b = bitcast float poison to float + %c = bitcast i32 poison to float + %d = bitcast float poison to i32 + %e = bitcast i64 poison to double + %f = bitcast double poison to i64 + %g = bitcast half poison to i16 + %h = bitcast i16 poison to half ret i32 undef } @@ -2012,31 +2012,31 @@ define i32 @load_extends() #0 { define i32 @store_truncs() { ; CHECK-LABEL: 'store_truncs' -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r0 = trunc i64 undef to i8 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r0 = trunc i64 poison to i8 ; CHECK-NEXT: Cost Model: Found costs of 1 for: store i8 %r0, ptr undef, align 1 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r1 = trunc i64 undef to i16 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r1 = trunc i64 poison to i16 ; CHECK-NEXT: Cost Model: Found costs of 1 for: store i16 %r1, ptr undef, align 2 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r2 = trunc i64 undef to i32 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r2 = trunc i64 poison to i32 ; CHECK-NEXT: Cost Model: Found costs of 1 for: store i32 %r2, ptr undef, align 4 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r3 = trunc i32 undef to i8 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r3 = trunc i32 poison to i8 ; CHECK-NEXT: Cost Model: Found costs of 1 for: store i8 %r3, ptr undef, align 1 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r4 = trunc i32 undef to i16 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r4 = trunc i32 poison to i16 ; CHECK-NEXT: Cost Model: Found costs of 1 for: store i16 %r4, ptr undef, align 2 -; CHECK-NEXT: Cost Model: Found costs of 0 for: %r5 = trunc i16 undef to i8 +; CHECK-NEXT: Cost Model: Found costs of 0 for: %r5 = trunc i16 poison to i8 ; CHECK-NEXT: Cost Model: Found costs of 1 for: store i8 %r5, ptr undef, align 1 ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret i32 undef ; - %r0 = trunc i64 undef to i8 + %r0 = trunc i64 poison to i8 store i8 %r0, ptr undef - %r1 = trunc i64 undef to i16 + %r1 = trunc i64 poison to i16 store i16 %r1, ptr undef - %r2 = trunc i64 undef to i32 + %r2 = trunc i64 poison to i32 store i32 %r2, ptr undef - %r3 = trunc i32 undef to i8 + %r3 = trunc i32 poison to i8 store i8 %r3, ptr undef - %r4 = trunc i32 undef to i16 + %r4 = trunc i32 poison to i16 store i16 %r4, ptr undef - %r5 = trunc i16 undef to i8 + %r5 = trunc i16 poison to i8 store i8 %r5, ptr undef ret i32 undef } @@ -2084,372 +2084,372 @@ declare void @use(i16, i16, i32, i32, i64, i64, i32, i32, i64, i64, i64, i64) define void @fp16cast() { ; CHECK-SVE-LABEL: 'fp16cast' -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui half undef to i1 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi half undef to i1 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui half undef to i8 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi half undef to i8 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui half undef to i16 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi half undef to i16 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui half undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi half undef to i32 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui half undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi half undef to i64 -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x half> undef to <2 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x half> undef to <2 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x half> undef to <2 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x half> undef to <2 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x half> undef to <2 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x half> undef to <2 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x half> undef to <2 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x half> undef to <2 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x half> undef to <2 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x half> undef to <2 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x half> undef to <4 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x half> undef to <4 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x half> undef to <4 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x half> undef to <4 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x half> undef to <4 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x half> undef to <4 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r116 = fptoui <4 x half> undef to <4 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r117 = fptosi <4 x half> undef to <4 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x half> undef to <4 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x half> undef to <4 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x half> undef to <8 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x half> undef to <8 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x half> undef to <8 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x half> undef to <8 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x half> undef to <8 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x half> undef to <8 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x half> undef to <8 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x half> undef to <8 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x half> undef to <8 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x half> undef to <8 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x half> undef to <16 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x half> undef to <16 x i1> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x half> undef to <16 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x half> undef to <16 x i8> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x half> undef to <16 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x half> undef to <16 x i16> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x half> undef to <16 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x half> undef to <16 x i32> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:86 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x half> undef to <16 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:86 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x half> undef to <16 x i64> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r250 = uitofp <8 x i1> undef to <8 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r251 = sitofp <8 x i1> undef to <8 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r252 = uitofp <8 x i8> undef to <8 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r253 = sitofp <8 x i8> undef to <8 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r254 = uitofp <8 x i16> undef to <8 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r255 = sitofp <8 x i16> undef to <8 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r256 = uitofp <8 x i32> undef to <8 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r257 = sitofp <8 x i32> undef to <8 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r258 = uitofp <8 x i64> undef to <8 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r259 = sitofp <8 x i64> undef to <8 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r260 = uitofp <16 x i1> undef to <16 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r261 = sitofp <16 x i1> undef to <16 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r262 = uitofp <16 x i8> undef to <16 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r263 = sitofp <16 x i8> undef to <16 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r264 = uitofp <16 x i16> undef to <16 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r265 = sitofp <16 x i16> undef to <16 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %r266 = uitofp <16 x i32> undef to <16 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %r267 = sitofp <16 x i32> undef to <16 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %r268 = uitofp <16 x i64> undef to <16 x half> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %r269 = sitofp <16 x i64> undef to <16 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui half poison to i1 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi half poison to i1 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui half poison to i8 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi half poison to i8 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui half poison to i16 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi half poison to i16 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui half poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi half poison to i32 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui half poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi half poison to i64 +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x half> poison to <2 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x half> poison to <2 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x half> poison to <2 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x half> poison to <2 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x half> poison to <2 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x half> poison to <2 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x half> poison to <2 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x half> poison to <2 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x half> poison to <2 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x half> poison to <2 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x half> poison to <4 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x half> poison to <4 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x half> poison to <4 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x half> poison to <4 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x half> poison to <4 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x half> poison to <4 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r116 = fptoui <4 x half> poison to <4 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r117 = fptosi <4 x half> poison to <4 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x half> poison to <4 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:21 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x half> poison to <4 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x half> poison to <8 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x half> poison to <8 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x half> poison to <8 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x half> poison to <8 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x half> poison to <8 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x half> poison to <8 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x half> poison to <8 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x half> poison to <8 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x half> poison to <8 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:43 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x half> poison to <8 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x half> poison to <16 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:81 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x half> poison to <16 x i1> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x half> poison to <16 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x half> poison to <16 x i8> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x half> poison to <16 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x half> poison to <16 x i16> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x half> poison to <16 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x half> poison to <16 x i32> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:86 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x half> poison to <16 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:86 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x half> poison to <16 x i64> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r250 = uitofp <8 x i1> poison to <8 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r251 = sitofp <8 x i1> poison to <8 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r252 = uitofp <8 x i8> poison to <8 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r253 = sitofp <8 x i8> poison to <8 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r254 = uitofp <8 x i16> poison to <8 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of 1 for: %r255 = sitofp <8 x i16> poison to <8 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r256 = uitofp <8 x i32> poison to <8 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r257 = sitofp <8 x i32> poison to <8 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r258 = uitofp <8 x i64> poison to <8 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %r259 = sitofp <8 x i64> poison to <8 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r260 = uitofp <16 x i1> poison to <16 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %r261 = sitofp <16 x i1> poison to <16 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r262 = uitofp <16 x i8> poison to <16 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r263 = sitofp <16 x i8> poison to <16 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r264 = uitofp <16 x i16> poison to <16 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r265 = sitofp <16 x i16> poison to <16 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %r266 = uitofp <16 x i32> poison to <16 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %r267 = sitofp <16 x i32> poison to <16 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %r268 = uitofp <16 x i64> poison to <16 x half> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %r269 = sitofp <16 x i64> poison to <16 x half> ; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; SVE128-NO-NEON-LABEL: 'fp16cast' -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui half undef to i1 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi half undef to i1 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui half undef to i8 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi half undef to i8 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui half undef to i16 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi half undef to i16 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui half undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi half undef to i32 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui half undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi half undef to i64 -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x half> undef to <2 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x half> undef to <2 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x half> undef to <2 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x half> undef to <2 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x half> undef to <2 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x half> undef to <2 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x half> undef to <2 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x half> undef to <2 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r98 = fptoui <2 x half> undef to <2 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r99 = fptosi <2 x half> undef to <2 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x half> undef to <4 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x half> undef to <4 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x half> undef to <4 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x half> undef to <4 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x half> undef to <4 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x half> undef to <4 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r116 = fptoui <4 x half> undef to <4 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r117 = fptosi <4 x half> undef to <4 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x half> undef to <4 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x half> undef to <4 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r130 = fptoui <8 x half> undef to <8 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r131 = fptosi <8 x half> undef to <8 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r132 = fptoui <8 x half> undef to <8 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r133 = fptosi <8 x half> undef to <8 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x half> undef to <8 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x half> undef to <8 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x half> undef to <8 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x half> undef to <8 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x half> undef to <8 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x half> undef to <8 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x half> undef to <16 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x half> undef to <16 x i1> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x half> undef to <16 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x half> undef to <16 x i8> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x half> undef to <16 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x half> undef to <16 x i16> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x half> undef to <16 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x half> undef to <16 x i32> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x half> undef to <16 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x half> undef to <16 x i64> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r250 = uitofp <8 x i1> undef to <8 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r251 = sitofp <8 x i1> undef to <8 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r252 = uitofp <8 x i8> undef to <8 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r253 = sitofp <8 x i8> undef to <8 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r254 = uitofp <8 x i16> undef to <8 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r255 = sitofp <8 x i16> undef to <8 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r256 = uitofp <8 x i32> undef to <8 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r257 = sitofp <8 x i32> undef to <8 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r258 = uitofp <8 x i64> undef to <8 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r259 = sitofp <8 x i64> undef to <8 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r260 = uitofp <16 x i1> undef to <16 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r261 = sitofp <16 x i1> undef to <16 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r262 = uitofp <16 x i8> undef to <16 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r263 = sitofp <16 x i8> undef to <16 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r264 = uitofp <16 x i16> undef to <16 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r265 = sitofp <16 x i16> undef to <16 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r266 = uitofp <16 x i32> undef to <16 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r267 = sitofp <16 x i32> undef to <16 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r268 = uitofp <16 x i64> undef to <16 x half> -; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r269 = sitofp <16 x i64> undef to <16 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui half poison to i1 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi half poison to i1 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui half poison to i8 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi half poison to i8 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui half poison to i16 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi half poison to i16 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui half poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi half poison to i32 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui half poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi half poison to i64 +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x half> poison to <2 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x half> poison to <2 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x half> poison to <2 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x half> poison to <2 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x half> poison to <2 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x half> poison to <2 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x half> poison to <2 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x half> poison to <2 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r98 = fptoui <2 x half> poison to <2 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r99 = fptosi <2 x half> poison to <2 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x half> poison to <4 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x half> poison to <4 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x half> poison to <4 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x half> poison to <4 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x half> poison to <4 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x half> poison to <4 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r116 = fptoui <4 x half> poison to <4 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r117 = fptosi <4 x half> poison to <4 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r118 = fptoui <4 x half> poison to <4 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r119 = fptosi <4 x half> poison to <4 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r130 = fptoui <8 x half> poison to <8 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r131 = fptosi <8 x half> poison to <8 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r132 = fptoui <8 x half> poison to <8 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r133 = fptosi <8 x half> poison to <8 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x half> poison to <8 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x half> poison to <8 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r136 = fptoui <8 x half> poison to <8 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r137 = fptosi <8 x half> poison to <8 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x half> poison to <8 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x half> poison to <8 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r150 = fptoui <16 x half> poison to <16 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r151 = fptosi <16 x half> poison to <16 x i1> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r152 = fptoui <16 x half> poison to <16 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r153 = fptosi <16 x half> poison to <16 x i8> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r154 = fptoui <16 x half> poison to <16 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r155 = fptosi <16 x half> poison to <16 x i16> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x half> poison to <16 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x half> poison to <16 x i32> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x half> poison to <16 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x half> poison to <16 x i64> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r250 = uitofp <8 x i1> poison to <8 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r251 = sitofp <8 x i1> poison to <8 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r252 = uitofp <8 x i8> poison to <8 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r253 = sitofp <8 x i8> poison to <8 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r254 = uitofp <8 x i16> poison to <8 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of 1 for: %r255 = sitofp <8 x i16> poison to <8 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r256 = uitofp <8 x i32> poison to <8 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r257 = sitofp <8 x i32> poison to <8 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r258 = uitofp <8 x i64> poison to <8 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r259 = sitofp <8 x i64> poison to <8 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r260 = uitofp <16 x i1> poison to <16 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r261 = sitofp <16 x i1> poison to <16 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r262 = uitofp <16 x i8> poison to <16 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r263 = sitofp <16 x i8> poison to <16 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r264 = uitofp <16 x i16> poison to <16 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r265 = sitofp <16 x i16> poison to <16 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r266 = uitofp <16 x i32> poison to <16 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r267 = sitofp <16 x i32> poison to <16 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r268 = uitofp <16 x i64> poison to <16 x half> +; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %r269 = sitofp <16 x i64> poison to <16 x half> ; SVE128-NO-NEON-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; FIXED-MIN-256-LABEL: 'fp16cast' -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui half undef to i1 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi half undef to i1 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui half undef to i8 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi half undef to i8 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui half undef to i16 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi half undef to i16 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui half undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi half undef to i32 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui half undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi half undef to i64 -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x half> undef to <2 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x half> undef to <2 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x half> undef to <2 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x half> undef to <2 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x half> undef to <2 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x half> undef to <2 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x half> undef to <2 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x half> undef to <2 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x half> undef to <2 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x half> undef to <2 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x half> undef to <4 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x half> undef to <4 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x half> undef to <4 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x half> undef to <4 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x half> undef to <4 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x half> undef to <4 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r116 = fptoui <4 x half> undef to <4 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r117 = fptosi <4 x half> undef to <4 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r118 = fptoui <4 x half> undef to <4 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r119 = fptosi <4 x half> undef to <4 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x half> undef to <8 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x half> undef to <8 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x half> undef to <8 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x half> undef to <8 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x half> undef to <8 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x half> undef to <8 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r136 = fptoui <8 x half> undef to <8 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r137 = fptosi <8 x half> undef to <8 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x half> undef to <8 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x half> undef to <8 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r150 = fptoui <16 x half> undef to <16 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r151 = fptosi <16 x half> undef to <16 x i1> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r152 = fptoui <16 x half> undef to <16 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r153 = fptosi <16 x half> undef to <16 x i8> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r154 = fptoui <16 x half> undef to <16 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r155 = fptosi <16 x half> undef to <16 x i16> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x half> undef to <16 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x half> undef to <16 x i32> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x half> undef to <16 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x half> undef to <16 x i64> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r250 = uitofp <8 x i1> undef to <8 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r251 = sitofp <8 x i1> undef to <8 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r252 = uitofp <8 x i8> undef to <8 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r253 = sitofp <8 x i8> undef to <8 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r254 = uitofp <8 x i16> undef to <8 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r255 = sitofp <8 x i16> undef to <8 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r256 = uitofp <8 x i32> undef to <8 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r257 = sitofp <8 x i32> undef to <8 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r258 = uitofp <8 x i64> undef to <8 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r259 = sitofp <8 x i64> undef to <8 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r260 = uitofp <16 x i1> undef to <16 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r261 = sitofp <16 x i1> undef to <16 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r262 = uitofp <16 x i8> undef to <16 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r263 = sitofp <16 x i8> undef to <16 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r264 = uitofp <16 x i16> undef to <16 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r265 = sitofp <16 x i16> undef to <16 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r266 = uitofp <16 x i32> undef to <16 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r267 = sitofp <16 x i32> undef to <16 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r268 = uitofp <16 x i64> undef to <16 x half> -; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r269 = sitofp <16 x i64> undef to <16 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui half poison to i1 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi half poison to i1 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui half poison to i8 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi half poison to i8 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui half poison to i16 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi half poison to i16 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui half poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi half poison to i32 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui half poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi half poison to i64 +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x half> poison to <2 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x half> poison to <2 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x half> poison to <2 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x half> poison to <2 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x half> poison to <2 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x half> poison to <2 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x half> poison to <2 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x half> poison to <2 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x half> poison to <2 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x half> poison to <2 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x half> poison to <4 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x half> poison to <4 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x half> poison to <4 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x half> poison to <4 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x half> poison to <4 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x half> poison to <4 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r116 = fptoui <4 x half> poison to <4 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r117 = fptosi <4 x half> poison to <4 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r118 = fptoui <4 x half> poison to <4 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r119 = fptosi <4 x half> poison to <4 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x half> poison to <8 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x half> poison to <8 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x half> poison to <8 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x half> poison to <8 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x half> poison to <8 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x half> poison to <8 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r136 = fptoui <8 x half> poison to <8 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r137 = fptosi <8 x half> poison to <8 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r138 = fptoui <8 x half> poison to <8 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r139 = fptosi <8 x half> poison to <8 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r150 = fptoui <16 x half> poison to <16 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r151 = fptosi <16 x half> poison to <16 x i1> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r152 = fptoui <16 x half> poison to <16 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r153 = fptosi <16 x half> poison to <16 x i8> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r154 = fptoui <16 x half> poison to <16 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r155 = fptosi <16 x half> poison to <16 x i16> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r156 = fptoui <16 x half> poison to <16 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r157 = fptosi <16 x half> poison to <16 x i32> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r158 = fptoui <16 x half> poison to <16 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r159 = fptosi <16 x half> poison to <16 x i64> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r250 = uitofp <8 x i1> poison to <8 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r251 = sitofp <8 x i1> poison to <8 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r252 = uitofp <8 x i8> poison to <8 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r253 = sitofp <8 x i8> poison to <8 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r254 = uitofp <8 x i16> poison to <8 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r255 = sitofp <8 x i16> poison to <8 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r256 = uitofp <8 x i32> poison to <8 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r257 = sitofp <8 x i32> poison to <8 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r258 = uitofp <8 x i64> poison to <8 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r259 = sitofp <8 x i64> poison to <8 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r260 = uitofp <16 x i1> poison to <16 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r261 = sitofp <16 x i1> poison to <16 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r262 = uitofp <16 x i8> poison to <16 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r263 = sitofp <16 x i8> poison to <16 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r264 = uitofp <16 x i16> poison to <16 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of 1 for: %r265 = sitofp <16 x i16> poison to <16 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r266 = uitofp <16 x i32> poison to <16 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r267 = sitofp <16 x i32> poison to <16 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r268 = uitofp <16 x i64> poison to <16 x half> +; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %r269 = sitofp <16 x i64> poison to <16 x half> ; FIXED-MIN-256-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; FIXED-MIN-2048-LABEL: 'fp16cast' -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui half undef to i1 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi half undef to i1 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui half undef to i8 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi half undef to i8 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui half undef to i16 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi half undef to i16 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui half undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi half undef to i32 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui half undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi half undef to i64 -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x half> undef to <2 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x half> undef to <2 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x half> undef to <2 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x half> undef to <2 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x half> undef to <2 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x half> undef to <2 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x half> undef to <2 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x half> undef to <2 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x half> undef to <2 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x half> undef to <2 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x half> undef to <4 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x half> undef to <4 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x half> undef to <4 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x half> undef to <4 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x half> undef to <4 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x half> undef to <4 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r116 = fptoui <4 x half> undef to <4 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r117 = fptosi <4 x half> undef to <4 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r118 = fptoui <4 x half> undef to <4 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r119 = fptosi <4 x half> undef to <4 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x half> undef to <8 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x half> undef to <8 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x half> undef to <8 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x half> undef to <8 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x half> undef to <8 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x half> undef to <8 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r136 = fptoui <8 x half> undef to <8 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r137 = fptosi <8 x half> undef to <8 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r138 = fptoui <8 x half> undef to <8 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r139 = fptosi <8 x half> undef to <8 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r150 = fptoui <16 x half> undef to <16 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r151 = fptosi <16 x half> undef to <16 x i1> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r152 = fptoui <16 x half> undef to <16 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r153 = fptosi <16 x half> undef to <16 x i8> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r154 = fptoui <16 x half> undef to <16 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r155 = fptosi <16 x half> undef to <16 x i16> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r156 = fptoui <16 x half> undef to <16 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r157 = fptosi <16 x half> undef to <16 x i32> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r158 = fptoui <16 x half> undef to <16 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r159 = fptosi <16 x half> undef to <16 x i64> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r250 = uitofp <8 x i1> undef to <8 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r251 = sitofp <8 x i1> undef to <8 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r252 = uitofp <8 x i8> undef to <8 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r253 = sitofp <8 x i8> undef to <8 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r254 = uitofp <8 x i16> undef to <8 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r255 = sitofp <8 x i16> undef to <8 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r256 = uitofp <8 x i32> undef to <8 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r257 = sitofp <8 x i32> undef to <8 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r258 = uitofp <8 x i64> undef to <8 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r259 = sitofp <8 x i64> undef to <8 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r260 = uitofp <16 x i1> undef to <16 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r261 = sitofp <16 x i1> undef to <16 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r262 = uitofp <16 x i8> undef to <16 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r263 = sitofp <16 x i8> undef to <16 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r264 = uitofp <16 x i16> undef to <16 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r265 = sitofp <16 x i16> undef to <16 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r266 = uitofp <16 x i32> undef to <16 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r267 = sitofp <16 x i32> undef to <16 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r268 = uitofp <16 x i64> undef to <16 x half> -; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r269 = sitofp <16 x i64> undef to <16 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r30 = fptoui half poison to i1 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r31 = fptosi half poison to i1 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r32 = fptoui half poison to i8 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r33 = fptosi half poison to i8 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r34 = fptoui half poison to i16 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r35 = fptosi half poison to i16 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r36 = fptoui half poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r37 = fptosi half poison to i32 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r38 = fptoui half poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r39 = fptosi half poison to i64 +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r90 = fptoui <2 x half> poison to <2 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r91 = fptosi <2 x half> poison to <2 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r92 = fptoui <2 x half> poison to <2 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r93 = fptosi <2 x half> poison to <2 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r94 = fptoui <2 x half> poison to <2 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r95 = fptosi <2 x half> poison to <2 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r96 = fptoui <2 x half> poison to <2 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r97 = fptosi <2 x half> poison to <2 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r98 = fptoui <2 x half> poison to <2 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %r99 = fptosi <2 x half> poison to <2 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r110 = fptoui <4 x half> poison to <4 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r111 = fptosi <4 x half> poison to <4 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r112 = fptoui <4 x half> poison to <4 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r113 = fptosi <4 x half> poison to <4 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r114 = fptoui <4 x half> poison to <4 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r115 = fptosi <4 x half> poison to <4 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r116 = fptoui <4 x half> poison to <4 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r117 = fptosi <4 x half> poison to <4 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r118 = fptoui <4 x half> poison to <4 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r119 = fptosi <4 x half> poison to <4 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r130 = fptoui <8 x half> poison to <8 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:40 CodeSize:1 Lat:1 SizeLat:1 for: %r131 = fptosi <8 x half> poison to <8 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r132 = fptoui <8 x half> poison to <8 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r133 = fptosi <8 x half> poison to <8 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r134 = fptoui <8 x half> poison to <8 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r135 = fptosi <8 x half> poison to <8 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r136 = fptoui <8 x half> poison to <8 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r137 = fptosi <8 x half> poison to <8 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r138 = fptoui <8 x half> poison to <8 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r139 = fptosi <8 x half> poison to <8 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r150 = fptoui <16 x half> poison to <16 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r151 = fptosi <16 x half> poison to <16 x i1> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r152 = fptoui <16 x half> poison to <16 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r153 = fptosi <16 x half> poison to <16 x i8> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r154 = fptoui <16 x half> poison to <16 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r155 = fptosi <16 x half> poison to <16 x i16> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r156 = fptoui <16 x half> poison to <16 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r157 = fptosi <16 x half> poison to <16 x i32> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r158 = fptoui <16 x half> poison to <16 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r159 = fptosi <16 x half> poison to <16 x i64> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r250 = uitofp <8 x i1> poison to <8 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r251 = sitofp <8 x i1> poison to <8 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r252 = uitofp <8 x i8> poison to <8 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r253 = sitofp <8 x i8> poison to <8 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r254 = uitofp <8 x i16> poison to <8 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r255 = sitofp <8 x i16> poison to <8 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r256 = uitofp <8 x i32> poison to <8 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r257 = sitofp <8 x i32> poison to <8 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r258 = uitofp <8 x i64> poison to <8 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r259 = sitofp <8 x i64> poison to <8 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r260 = uitofp <16 x i1> poison to <16 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r261 = sitofp <16 x i1> poison to <16 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r262 = uitofp <16 x i8> poison to <16 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %r263 = sitofp <16 x i8> poison to <16 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r264 = uitofp <16 x i16> poison to <16 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r265 = sitofp <16 x i16> poison to <16 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r266 = uitofp <16 x i32> poison to <16 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r267 = sitofp <16 x i32> poison to <16 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r268 = uitofp <16 x i64> poison to <16 x half> +; FIXED-MIN-2048-NEXT: Cost Model: Found costs of 1 for: %r269 = sitofp <16 x i64> poison to <16 x half> ; FIXED-MIN-2048-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %r30 = fptoui half undef to i1 - %r31 = fptosi half undef to i1 - %r32 = fptoui half undef to i8 - %r33 = fptosi half undef to i8 - %r34 = fptoui half undef to i16 - %r35 = fptosi half undef to i16 - %r36 = fptoui half undef to i32 - %r37 = fptosi half undef to i32 - %r38 = fptoui half undef to i64 - %r39 = fptosi half undef to i64 + %r30 = fptoui half poison to i1 + %r31 = fptosi half poison to i1 + %r32 = fptoui half poison to i8 + %r33 = fptosi half poison to i8 + %r34 = fptoui half poison to i16 + %r35 = fptosi half poison to i16 + %r36 = fptoui half poison to i32 + %r37 = fptosi half poison to i32 + %r38 = fptoui half poison to i64 + %r39 = fptosi half poison to i64 - %r90 = fptoui <2 x half> undef to <2 x i1> - %r91 = fptosi <2 x half> undef to <2 x i1> - %r92 = fptoui <2 x half> undef to <2 x i8> - %r93 = fptosi <2 x half> undef to <2 x i8> - %r94 = fptoui <2 x half> undef to <2 x i16> - %r95 = fptosi <2 x half> undef to <2 x i16> - %r96 = fptoui <2 x half> undef to <2 x i32> - %r97 = fptosi <2 x half> undef to <2 x i32> - %r98 = fptoui <2 x half> undef to <2 x i64> - %r99 = fptosi <2 x half> undef to <2 x i64> + %r90 = fptoui <2 x half> poison to <2 x i1> + %r91 = fptosi <2 x half> poison to <2 x i1> + %r92 = fptoui <2 x half> poison to <2 x i8> + %r93 = fptosi <2 x half> poison to <2 x i8> + %r94 = fptoui <2 x half> poison to <2 x i16> + %r95 = fptosi <2 x half> poison to <2 x i16> + %r96 = fptoui <2 x half> poison to <2 x i32> + %r97 = fptosi <2 x half> poison to <2 x i32> + %r98 = fptoui <2 x half> poison to <2 x i64> + %r99 = fptosi <2 x half> poison to <2 x i64> - %r110 = fptoui <4 x half> undef to <4 x i1> - %r111 = fptosi <4 x half> undef to <4 x i1> - %r112 = fptoui <4 x half> undef to <4 x i8> - %r113 = fptosi <4 x half> undef to <4 x i8> - %r114 = fptoui <4 x half> undef to <4 x i16> - %r115 = fptosi <4 x half> undef to <4 x i16> - %r116 = fptoui <4 x half> undef to <4 x i32> - %r117 = fptosi <4 x half> undef to <4 x i32> - %r118 = fptoui <4 x half> undef to <4 x i64> - %r119 = fptosi <4 x half> undef to <4 x i64> + %r110 = fptoui <4 x half> poison to <4 x i1> + %r111 = fptosi <4 x half> poison to <4 x i1> + %r112 = fptoui <4 x half> poison to <4 x i8> + %r113 = fptosi <4 x half> poison to <4 x i8> + %r114 = fptoui <4 x half> poison to <4 x i16> + %r115 = fptosi <4 x half> poison to <4 x i16> + %r116 = fptoui <4 x half> poison to <4 x i32> + %r117 = fptosi <4 x half> poison to <4 x i32> + %r118 = fptoui <4 x half> poison to <4 x i64> + %r119 = fptosi <4 x half> poison to <4 x i64> - %r130 = fptoui <8 x half> undef to <8 x i1> - %r131 = fptosi <8 x half> undef to <8 x i1> - %r132 = fptoui <8 x half> undef to <8 x i8> - %r133 = fptosi <8 x half> undef to <8 x i8> - %r134 = fptoui <8 x half> undef to <8 x i16> - %r135 = fptosi <8 x half> undef to <8 x i16> - %r136 = fptoui <8 x half> undef to <8 x i32> - %r137 = fptosi <8 x half> undef to <8 x i32> - %r138 = fptoui <8 x half> undef to <8 x i64> - %r139 = fptosi <8 x half> undef to <8 x i64> + %r130 = fptoui <8 x half> poison to <8 x i1> + %r131 = fptosi <8 x half> poison to <8 x i1> + %r132 = fptoui <8 x half> poison to <8 x i8> + %r133 = fptosi <8 x half> poison to <8 x i8> + %r134 = fptoui <8 x half> poison to <8 x i16> + %r135 = fptosi <8 x half> poison to <8 x i16> + %r136 = fptoui <8 x half> poison to <8 x i32> + %r137 = fptosi <8 x half> poison to <8 x i32> + %r138 = fptoui <8 x half> poison to <8 x i64> + %r139 = fptosi <8 x half> poison to <8 x i64> - %r150 = fptoui <16 x half> undef to <16 x i1> - %r151 = fptosi <16 x half> undef to <16 x i1> - %r152 = fptoui <16 x half> undef to <16 x i8> - %r153 = fptosi <16 x half> undef to <16 x i8> - %r154 = fptoui <16 x half> undef to <16 x i16> - %r155 = fptosi <16 x half> undef to <16 x i16> - %r156 = fptoui <16 x half> undef to <16 x i32> - %r157 = fptosi <16 x half> undef to <16 x i32> - %r158 = fptoui <16 x half> undef to <16 x i64> - %r159 = fptosi <16 x half> undef to <16 x i64> + %r150 = fptoui <16 x half> poison to <16 x i1> + %r151 = fptosi <16 x half> poison to <16 x i1> + %r152 = fptoui <16 x half> poison to <16 x i8> + %r153 = fptosi <16 x half> poison to <16 x i8> + %r154 = fptoui <16 x half> poison to <16 x i16> + %r155 = fptosi <16 x half> poison to <16 x i16> + %r156 = fptoui <16 x half> poison to <16 x i32> + %r157 = fptosi <16 x half> poison to <16 x i32> + %r158 = fptoui <16 x half> poison to <16 x i64> + %r159 = fptosi <16 x half> poison to <16 x i64> - %r250 = uitofp <8 x i1> undef to <8 x half> - %r251 = sitofp <8 x i1> undef to <8 x half> - %r252 = uitofp <8 x i8> undef to <8 x half> - %r253 = sitofp <8 x i8> undef to <8 x half> - %r254 = uitofp <8 x i16> undef to <8 x half> - %r255 = sitofp <8 x i16> undef to <8 x half> - %r256 = uitofp <8 x i32> undef to <8 x half> - %r257 = sitofp <8 x i32> undef to <8 x half> - %r258 = uitofp <8 x i64> undef to <8 x half> - %r259 = sitofp <8 x i64> undef to <8 x half> + %r250 = uitofp <8 x i1> poison to <8 x half> + %r251 = sitofp <8 x i1> poison to <8 x half> + %r252 = uitofp <8 x i8> poison to <8 x half> + %r253 = sitofp <8 x i8> poison to <8 x half> + %r254 = uitofp <8 x i16> poison to <8 x half> + %r255 = sitofp <8 x i16> poison to <8 x half> + %r256 = uitofp <8 x i32> poison to <8 x half> + %r257 = sitofp <8 x i32> poison to <8 x half> + %r258 = uitofp <8 x i64> poison to <8 x half> + %r259 = sitofp <8 x i64> poison to <8 x half> - %r260 = uitofp <16 x i1> undef to <16 x half> - %r261 = sitofp <16 x i1> undef to <16 x half> - %r262 = uitofp <16 x i8> undef to <16 x half> - %r263 = sitofp <16 x i8> undef to <16 x half> - %r264 = uitofp <16 x i16> undef to <16 x half> - %r265 = sitofp <16 x i16> undef to <16 x half> - %r266 = uitofp <16 x i32> undef to <16 x half> - %r267 = sitofp <16 x i32> undef to <16 x half> - %r268 = uitofp <16 x i64> undef to <16 x half> - %r269 = sitofp <16 x i64> undef to <16 x half> + %r260 = uitofp <16 x i1> poison to <16 x half> + %r261 = sitofp <16 x i1> poison to <16 x half> + %r262 = uitofp <16 x i8> poison to <16 x half> + %r263 = sitofp <16 x i8> poison to <16 x half> + %r264 = uitofp <16 x i16> poison to <16 x half> + %r265 = sitofp <16 x i16> poison to <16 x half> + %r266 = uitofp <16 x i32> poison to <16 x half> + %r267 = sitofp <16 x i32> poison to <16 x half> + %r268 = uitofp <16 x i64> poison to <16 x half> + %r269 = sitofp <16 x i64> poison to <16 x half> ret void } diff --git a/llvm/test/Analysis/CostModel/AArch64/sve-ext.ll b/llvm/test/Analysis/CostModel/AArch64/sve-ext.ll index b887654..91aaea2 100644 --- a/llvm/test/Analysis/CostModel/AArch64/sve-ext.ll +++ b/llvm/test/Analysis/CostModel/AArch64/sve-ext.ll @@ -5,49 +5,49 @@ target triple = "aarch64-unknown-linux-gnu" define void @sve_ext() { ; CHECK-LABEL: 'sve_ext' -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv16_i8_to_i16 = zext <vscale x 16 x i8> undef to <vscale x 16 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv16_i8_to_i32 = zext <vscale x 16 x i8> undef to <vscale x 16 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv16_i8_to_i64 = zext <vscale x 16 x i8> undef to <vscale x 16 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv8_i16_to_i32 = zext <vscale x 8 x i16> undef to <vscale x 8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv8_i16_to_i64 = zext <vscale x 8 x i16> undef to <vscale x 8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv4_i32_to_i64 = zext <vscale x 4 x i32> undef to <vscale x 4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv4_i8_to_i64 = zext <vscale x 4 x i8> undef to <vscale x 4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv8_i8_to_i32 = zext <vscale x 8 x i8> undef to <vscale x 8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv4_i16_to_i64 = zext <vscale x 4 x i16> undef to <vscale x 4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv8_i8_to_i64 = zext <vscale x 8 x i8> undef to <vscale x 8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv16_i8_to_i16 = sext <vscale x 16 x i8> undef to <vscale x 16 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv16_i8_to_i32 = sext <vscale x 16 x i8> undef to <vscale x 16 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv16_i8_to_i64 = sext <vscale x 16 x i8> undef to <vscale x 16 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv8_i16_to_i32 = sext <vscale x 8 x i16> undef to <vscale x 8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv8_i16_to_i64 = sext <vscale x 8 x i16> undef to <vscale x 8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv4_i32_to_i64 = sext <vscale x 4 x i32> undef to <vscale x 4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv4_i8_to_i64 = sext <vscale x 4 x i8> undef to <vscale x 4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv8_i8_to_i32 = sext <vscale x 8 x i8> undef to <vscale x 8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv4_i16_to_i64 = sext <vscale x 4 x i16> undef to <vscale x 4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv8_i8_to_i64 = sext <vscale x 8 x i8> undef to <vscale x 8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv16_i8_to_i16 = zext <vscale x 16 x i8> poison to <vscale x 16 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv16_i8_to_i32 = zext <vscale x 16 x i8> poison to <vscale x 16 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv16_i8_to_i64 = zext <vscale x 16 x i8> poison to <vscale x 16 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv8_i16_to_i32 = zext <vscale x 8 x i16> poison to <vscale x 8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv8_i16_to_i64 = zext <vscale x 8 x i16> poison to <vscale x 8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv4_i32_to_i64 = zext <vscale x 4 x i32> poison to <vscale x 4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv4_i8_to_i64 = zext <vscale x 4 x i8> poison to <vscale x 4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv8_i8_to_i32 = zext <vscale x 8 x i8> poison to <vscale x 8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv4_i16_to_i64 = zext <vscale x 4 x i16> poison to <vscale x 4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %zext_nxv8_i8_to_i64 = zext <vscale x 8 x i8> poison to <vscale x 8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv16_i8_to_i16 = sext <vscale x 16 x i8> poison to <vscale x 16 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv16_i8_to_i32 = sext <vscale x 16 x i8> poison to <vscale x 16 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv16_i8_to_i64 = sext <vscale x 16 x i8> poison to <vscale x 16 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv8_i16_to_i32 = sext <vscale x 8 x i16> poison to <vscale x 8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv8_i16_to_i64 = sext <vscale x 8 x i16> poison to <vscale x 8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv4_i32_to_i64 = sext <vscale x 4 x i32> poison to <vscale x 4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv4_i8_to_i64 = sext <vscale x 4 x i8> poison to <vscale x 4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv8_i8_to_i32 = sext <vscale x 8 x i8> poison to <vscale x 8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv4_i16_to_i64 = sext <vscale x 4 x i16> poison to <vscale x 4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %sext_nxv8_i8_to_i64 = sext <vscale x 8 x i8> poison to <vscale x 8 x i64> ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %zext_nxv16_i8_to_i16 = zext <vscale x 16 x i8> undef to <vscale x 16 x i16> - %zext_nxv16_i8_to_i32 = zext <vscale x 16 x i8> undef to <vscale x 16 x i32> - %zext_nxv16_i8_to_i64 = zext <vscale x 16 x i8> undef to <vscale x 16 x i64> - %zext_nxv8_i16_to_i32 = zext <vscale x 8 x i16> undef to <vscale x 8 x i32> - %zext_nxv8_i16_to_i64 = zext <vscale x 8 x i16> undef to <vscale x 8 x i64> - %zext_nxv4_i32_to_i64 = zext <vscale x 4 x i32> undef to <vscale x 4 x i64> - %zext_nxv4_i8_to_i64 = zext <vscale x 4 x i8> undef to <vscale x 4 x i64> - %zext_nxv8_i8_to_i32 = zext <vscale x 8 x i8> undef to <vscale x 8 x i32> - %zext_nxv4_i16_to_i64 = zext <vscale x 4 x i16> undef to <vscale x 4 x i64> - %zext_nxv8_i8_to_i64 = zext <vscale x 8 x i8> undef to <vscale x 8 x i64> + %zext_nxv16_i8_to_i16 = zext <vscale x 16 x i8> poison to <vscale x 16 x i16> + %zext_nxv16_i8_to_i32 = zext <vscale x 16 x i8> poison to <vscale x 16 x i32> + %zext_nxv16_i8_to_i64 = zext <vscale x 16 x i8> poison to <vscale x 16 x i64> + %zext_nxv8_i16_to_i32 = zext <vscale x 8 x i16> poison to <vscale x 8 x i32> + %zext_nxv8_i16_to_i64 = zext <vscale x 8 x i16> poison to <vscale x 8 x i64> + %zext_nxv4_i32_to_i64 = zext <vscale x 4 x i32> poison to <vscale x 4 x i64> + %zext_nxv4_i8_to_i64 = zext <vscale x 4 x i8> poison to <vscale x 4 x i64> + %zext_nxv8_i8_to_i32 = zext <vscale x 8 x i8> poison to <vscale x 8 x i32> + %zext_nxv4_i16_to_i64 = zext <vscale x 4 x i16> poison to <vscale x 4 x i64> + %zext_nxv8_i8_to_i64 = zext <vscale x 8 x i8> poison to <vscale x 8 x i64> - %sext_nxv16_i8_to_i16 = sext <vscale x 16 x i8> undef to <vscale x 16 x i16> - %sext_nxv16_i8_to_i32 = sext <vscale x 16 x i8> undef to <vscale x 16 x i32> - %sext_nxv16_i8_to_i64 = sext <vscale x 16 x i8> undef to <vscale x 16 x i64> - %sext_nxv8_i16_to_i32 = sext <vscale x 8 x i16> undef to <vscale x 8 x i32> - %sext_nxv8_i16_to_i64 = sext <vscale x 8 x i16> undef to <vscale x 8 x i64> - %sext_nxv4_i32_to_i64 = sext <vscale x 4 x i32> undef to <vscale x 4 x i64> - %sext_nxv4_i8_to_i64 = sext <vscale x 4 x i8> undef to <vscale x 4 x i64> - %sext_nxv8_i8_to_i32 = sext <vscale x 8 x i8> undef to <vscale x 8 x i32> - %sext_nxv4_i16_to_i64 = sext <vscale x 4 x i16> undef to <vscale x 4 x i64> - %sext_nxv8_i8_to_i64 = sext <vscale x 8 x i8> undef to <vscale x 8 x i64> + %sext_nxv16_i8_to_i16 = sext <vscale x 16 x i8> poison to <vscale x 16 x i16> + %sext_nxv16_i8_to_i32 = sext <vscale x 16 x i8> poison to <vscale x 16 x i32> + %sext_nxv16_i8_to_i64 = sext <vscale x 16 x i8> poison to <vscale x 16 x i64> + %sext_nxv8_i16_to_i32 = sext <vscale x 8 x i16> poison to <vscale x 8 x i32> + %sext_nxv8_i16_to_i64 = sext <vscale x 8 x i16> poison to <vscale x 8 x i64> + %sext_nxv4_i32_to_i64 = sext <vscale x 4 x i32> poison to <vscale x 4 x i64> + %sext_nxv4_i8_to_i64 = sext <vscale x 4 x i8> poison to <vscale x 4 x i64> + %sext_nxv8_i8_to_i32 = sext <vscale x 8 x i8> poison to <vscale x 8 x i32> + %sext_nxv4_i16_to_i64 = sext <vscale x 4 x i16> poison to <vscale x 4 x i64> + %sext_nxv8_i8_to_i64 = sext <vscale x 8 x i8> poison to <vscale x 8 x i64> ret void } diff --git a/llvm/test/Analysis/CostModel/AArch64/sve-fpext.ll b/llvm/test/Analysis/CostModel/AArch64/sve-fpext.ll index 4ad0e3f..1e698b1 100644 --- a/llvm/test/Analysis/CostModel/AArch64/sve-fpext.ll +++ b/llvm/test/Analysis/CostModel/AArch64/sve-fpext.ll @@ -6,49 +6,49 @@ target triple = "aarch64-unknown-linux-gnu" define void @sve_fpext() { ; CHECK-LABEL: 'sve_fpext' -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f16_to_f32 = fpext <vscale x 2 x half> undef to <vscale x 2 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv4_f16_to_f32 = fpext <vscale x 4 x half> undef to <vscale x 4 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_to_f32 = fpext <vscale x 8 x half> undef to <vscale x 8 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f16_to_f64 = fpext <vscale x 2 x half> undef to <vscale x 2 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_to_f64 = fpext <vscale x 4 x half> undef to <vscale x 4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_to_f64 = fpext <vscale x 8 x half> undef to <vscale x 8 x double> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f32_to_f64 = fpext <vscale x 2 x float> undef to <vscale x 2 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f32_to_f64 = fpext <vscale x 4 x float> undef to <vscale x 4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f32_to_f64 = fpext <vscale x 8 x float> undef to <vscale x 8 x double> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f16_to_f32 = fpext <vscale x 2 x half> poison to <vscale x 2 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv4_f16_to_f32 = fpext <vscale x 4 x half> poison to <vscale x 4 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_to_f32 = fpext <vscale x 8 x half> poison to <vscale x 8 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f16_to_f64 = fpext <vscale x 2 x half> poison to <vscale x 2 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_to_f64 = fpext <vscale x 4 x half> poison to <vscale x 4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_to_f64 = fpext <vscale x 8 x half> poison to <vscale x 8 x double> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f32_to_f64 = fpext <vscale x 2 x float> poison to <vscale x 2 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f32_to_f64 = fpext <vscale x 4 x float> poison to <vscale x 4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f32_to_f64 = fpext <vscale x 8 x float> poison to <vscale x 8 x double> ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %nxv2_f16_to_f32 = fpext <vscale x 2 x half> undef to <vscale x 2 x float> - %nxv4_f16_to_f32 = fpext <vscale x 4 x half> undef to <vscale x 4 x float> - %nxv8_f16_to_f32 = fpext <vscale x 8 x half> undef to <vscale x 8 x float> + %nxv2_f16_to_f32 = fpext <vscale x 2 x half> poison to <vscale x 2 x float> + %nxv4_f16_to_f32 = fpext <vscale x 4 x half> poison to <vscale x 4 x float> + %nxv8_f16_to_f32 = fpext <vscale x 8 x half> poison to <vscale x 8 x float> - %nxv2_f16_to_f64 = fpext <vscale x 2 x half> undef to <vscale x 2 x double> - %nxv4_f16_to_f64 = fpext <vscale x 4 x half> undef to <vscale x 4 x double> - %nxv8_f16_to_f64 = fpext <vscale x 8 x half> undef to <vscale x 8 x double> + %nxv2_f16_to_f64 = fpext <vscale x 2 x half> poison to <vscale x 2 x double> + %nxv4_f16_to_f64 = fpext <vscale x 4 x half> poison to <vscale x 4 x double> + %nxv8_f16_to_f64 = fpext <vscale x 8 x half> poison to <vscale x 8 x double> - %nxv2_f32_to_f64 = fpext <vscale x 2 x float> undef to <vscale x 2 x double> - %nxv4_f32_to_f64 = fpext <vscale x 4 x float> undef to <vscale x 4 x double> - %nxv8_f32_to_f64 = fpext <vscale x 8 x float> undef to <vscale x 8 x double> + %nxv2_f32_to_f64 = fpext <vscale x 2 x float> poison to <vscale x 2 x double> + %nxv4_f32_to_f64 = fpext <vscale x 4 x float> poison to <vscale x 4 x double> + %nxv8_f32_to_f64 = fpext <vscale x 8 x float> poison to <vscale x 8 x double> ret void } define void @sve_fpext_bf16() { ; CHECK-LABEL: 'sve_fpext_bf16' -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f16_to_f32 = fpext <vscale x 2 x bfloat> undef to <vscale x 2 x float> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv4_f16_to_f32 = fpext <vscale x 4 x bfloat> undef to <vscale x 4 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_to_f32 = fpext <vscale x 8 x bfloat> undef to <vscale x 8 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %nxv2_f16_to_f64 = fpext <vscale x 2 x bfloat> undef to <vscale x 2 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_to_f64 = fpext <vscale x 4 x bfloat> undef to <vscale x 4 x double> -; CHECK-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_to_f64 = fpext <vscale x 8 x bfloat> undef to <vscale x 8 x double> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f16_to_f32 = fpext <vscale x 2 x bfloat> poison to <vscale x 2 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv4_f16_to_f32 = fpext <vscale x 4 x bfloat> poison to <vscale x 4 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_to_f32 = fpext <vscale x 8 x bfloat> poison to <vscale x 8 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %nxv2_f16_to_f64 = fpext <vscale x 2 x bfloat> poison to <vscale x 2 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_to_f64 = fpext <vscale x 4 x bfloat> poison to <vscale x 4 x double> +; CHECK-NEXT: Cost Model: Found costs of RThru:14 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_to_f64 = fpext <vscale x 8 x bfloat> poison to <vscale x 8 x double> ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %nxv2_f16_to_f32 = fpext <vscale x 2 x bfloat> undef to <vscale x 2 x float> - %nxv4_f16_to_f32 = fpext <vscale x 4 x bfloat> undef to <vscale x 4 x float> - %nxv8_f16_to_f32 = fpext <vscale x 8 x bfloat> undef to <vscale x 8 x float> + %nxv2_f16_to_f32 = fpext <vscale x 2 x bfloat> poison to <vscale x 2 x float> + %nxv4_f16_to_f32 = fpext <vscale x 4 x bfloat> poison to <vscale x 4 x float> + %nxv8_f16_to_f32 = fpext <vscale x 8 x bfloat> poison to <vscale x 8 x float> - %nxv2_f16_to_f64 = fpext <vscale x 2 x bfloat> undef to <vscale x 2 x double> - %nxv4_f16_to_f64 = fpext <vscale x 4 x bfloat> undef to <vscale x 4 x double> - %nxv8_f16_to_f64 = fpext <vscale x 8 x bfloat> undef to <vscale x 8 x double> + %nxv2_f16_to_f64 = fpext <vscale x 2 x bfloat> poison to <vscale x 2 x double> + %nxv4_f16_to_f64 = fpext <vscale x 4 x bfloat> poison to <vscale x 4 x double> + %nxv8_f16_to_f64 = fpext <vscale x 8 x bfloat> poison to <vscale x 8 x double> ret void } diff --git a/llvm/test/Analysis/CostModel/AArch64/sve-fptoi.ll b/llvm/test/Analysis/CostModel/AArch64/sve-fptoi.ll index 06ed58d..ce624a1 100644 --- a/llvm/test/Analysis/CostModel/AArch64/sve-fptoi.ll +++ b/llvm/test/Analysis/CostModel/AArch64/sve-fptoi.ll @@ -6,163 +6,163 @@ target triple = "aarch64-unknown-linux-gnu" define void @sve-fptoi() { ; CHECK-LABEL: 'sve-fptoi' -; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f16_to_si8 = fptosi <vscale x 1 x half> undef to <vscale x 1 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f16_to_ui8 = fptoui <vscale x 1 x half> undef to <vscale x 1 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f16_to_si32 = fptosi <vscale x 1 x half> undef to <vscale x 1 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f16_to_ui32 = fptoui <vscale x 1 x half> undef to <vscale x 1 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f16_to_si64 = fptosi <vscale x 1 x half> undef to <vscale x 1 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f16_to_ui64 = fptoui <vscale x 1 x half> undef to <vscale x 1 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f32_to_si8 = fptosi <vscale x 1 x float> undef to <vscale x 1 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f32_to_ui8 = fptoui <vscale x 1 x float> undef to <vscale x 1 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f32_to_si16 = fptosi <vscale x 1 x float> undef to <vscale x 1 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f32_to_ui16 = fptoui <vscale x 1 x float> undef to <vscale x 1 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f32_to_si64 = fptosi <vscale x 1 x float> undef to <vscale x 1 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f32_to_ui64 = fptoui <vscale x 1 x float> undef to <vscale x 1 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv1f64_to_si8 = fptosi <vscale x 1 x double> undef to <vscale x 1 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv1f64_to_ui8 = fptoui <vscale x 1 x double> undef to <vscale x 1 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv1f64_to_si16 = fptosi <vscale x 1 x double> undef to <vscale x 1 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv1f64_to_ui16 = fptoui <vscale x 1 x double> undef to <vscale x 1 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv1f64_to_si32 = fptosi <vscale x 1 x double> undef to <vscale x 1 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv1f64_to_ui32 = fptoui <vscale x 1 x double> undef to <vscale x 1 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f16_to_si8 = fptosi <vscale x 2 x half> undef to <vscale x 2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f16_to_ui8 = fptoui <vscale x 2 x half> undef to <vscale x 2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f16_to_si32 = fptosi <vscale x 2 x half> undef to <vscale x 2 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f16_to_ui32 = fptoui <vscale x 2 x half> undef to <vscale x 2 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f16_to_si64 = fptosi <vscale x 2 x half> undef to <vscale x 2 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f16_to_ui64 = fptoui <vscale x 2 x half> undef to <vscale x 2 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f32_to_si8 = fptosi <vscale x 2 x float> undef to <vscale x 2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f32_to_ui8 = fptoui <vscale x 2 x float> undef to <vscale x 2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f32_to_si16 = fptosi <vscale x 2 x float> undef to <vscale x 2 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f32_to_ui16 = fptoui <vscale x 2 x float> undef to <vscale x 2 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f32_to_si64 = fptosi <vscale x 2 x float> undef to <vscale x 2 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f32_to_ui64 = fptoui <vscale x 2 x float> undef to <vscale x 2 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f64_to_si8 = fptosi <vscale x 2 x double> undef to <vscale x 2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f64_to_ui8 = fptoui <vscale x 2 x double> undef to <vscale x 2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f64_to_si16 = fptosi <vscale x 2 x double> undef to <vscale x 2 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f64_to_ui16 = fptoui <vscale x 2 x double> undef to <vscale x 2 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f64_to_si32 = fptosi <vscale x 2 x double> undef to <vscale x 2 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f64_to_ui32 = fptoui <vscale x 2 x double> undef to <vscale x 2 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f16_to_si8 = fptosi <vscale x 4 x half> undef to <vscale x 4 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f16_to_ui8 = fptoui <vscale x 4 x half> undef to <vscale x 4 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f16_to_si32 = fptosi <vscale x 4 x half> undef to <vscale x 4 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f16_to_ui32 = fptoui <vscale x 4 x half> undef to <vscale x 4 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f16_to_si64 = fptosi <vscale x 4 x half> undef to <vscale x 4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f16_to_ui64 = fptoui <vscale x 4 x half> undef to <vscale x 4 x i64> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f32_to_si8 = fptosi <vscale x 4 x float> undef to <vscale x 4 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f32_to_ui8 = fptoui <vscale x 4 x float> undef to <vscale x 4 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f32_to_si16 = fptosi <vscale x 4 x float> undef to <vscale x 4 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f32_to_ui16 = fptoui <vscale x 4 x float> undef to <vscale x 4 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f32_to_si64 = fptosi <vscale x 4 x float> undef to <vscale x 4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f32_to_ui64 = fptoui <vscale x 4 x float> undef to <vscale x 4 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f64_to_si8 = fptosi <vscale x 4 x double> undef to <vscale x 4 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f64_to_ui8 = fptoui <vscale x 4 x double> undef to <vscale x 4 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f64_to_si16 = fptosi <vscale x 4 x double> undef to <vscale x 4 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f64_to_ui16 = fptoui <vscale x 4 x double> undef to <vscale x 4 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f64_to_si32 = fptosi <vscale x 4 x double> undef to <vscale x 4 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f64_to_ui32 = fptoui <vscale x 4 x double> undef to <vscale x 4 x i32> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv8f16_to_si8 = fptosi <vscale x 8 x half> undef to <vscale x 8 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv8f16_to_ui8 = fptoui <vscale x 8 x half> undef to <vscale x 8 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f16_to_si32 = fptosi <vscale x 8 x half> undef to <vscale x 8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f16_to_ui32 = fptoui <vscale x 8 x half> undef to <vscale x 8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f16_to_si64 = fptosi <vscale x 8 x half> undef to <vscale x 8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f16_to_ui64 = fptoui <vscale x 8 x half> undef to <vscale x 8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f32_to_si8 = fptosi <vscale x 8 x float> undef to <vscale x 8 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f32_to_ui8 = fptoui <vscale x 8 x float> undef to <vscale x 8 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f32_to_si16 = fptosi <vscale x 8 x float> undef to <vscale x 8 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f32_to_ui16 = fptoui <vscale x 8 x float> undef to <vscale x 8 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f32_to_si64 = fptosi <vscale x 8 x float> undef to <vscale x 8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f32_to_ui64 = fptoui <vscale x 8 x float> undef to <vscale x 8 x i64> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f64_to_si8 = fptosi <vscale x 8 x double> undef to <vscale x 8 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f64_to_ui8 = fptoui <vscale x 8 x double> undef to <vscale x 8 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f64_to_si16 = fptosi <vscale x 8 x double> undef to <vscale x 8 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f64_to_ui16 = fptoui <vscale x 8 x double> undef to <vscale x 8 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f64_to_si32 = fptosi <vscale x 8 x double> undef to <vscale x 8 x i32> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f64_to_ui32 = fptoui <vscale x 8 x double> undef to <vscale x 8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f16_to_si8 = fptosi <vscale x 1 x half> poison to <vscale x 1 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f16_to_ui8 = fptoui <vscale x 1 x half> poison to <vscale x 1 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f16_to_si32 = fptosi <vscale x 1 x half> poison to <vscale x 1 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f16_to_ui32 = fptoui <vscale x 1 x half> poison to <vscale x 1 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f16_to_si64 = fptosi <vscale x 1 x half> poison to <vscale x 1 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f16_to_ui64 = fptoui <vscale x 1 x half> poison to <vscale x 1 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f32_to_si8 = fptosi <vscale x 1 x float> poison to <vscale x 1 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f32_to_ui8 = fptoui <vscale x 1 x float> poison to <vscale x 1 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f32_to_si16 = fptosi <vscale x 1 x float> poison to <vscale x 1 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f32_to_ui16 = fptoui <vscale x 1 x float> poison to <vscale x 1 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f32_to_si64 = fptosi <vscale x 1 x float> poison to <vscale x 1 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:Invalid CodeSize:1 Lat:1 SizeLat:1 for: %nv1f32_to_ui64 = fptoui <vscale x 1 x float> poison to <vscale x 1 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv1f64_to_si8 = fptosi <vscale x 1 x double> poison to <vscale x 1 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv1f64_to_ui8 = fptoui <vscale x 1 x double> poison to <vscale x 1 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv1f64_to_si16 = fptosi <vscale x 1 x double> poison to <vscale x 1 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv1f64_to_ui16 = fptoui <vscale x 1 x double> poison to <vscale x 1 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv1f64_to_si32 = fptosi <vscale x 1 x double> poison to <vscale x 1 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv1f64_to_ui32 = fptoui <vscale x 1 x double> poison to <vscale x 1 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f16_to_si8 = fptosi <vscale x 2 x half> poison to <vscale x 2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f16_to_ui8 = fptoui <vscale x 2 x half> poison to <vscale x 2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f16_to_si32 = fptosi <vscale x 2 x half> poison to <vscale x 2 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f16_to_ui32 = fptoui <vscale x 2 x half> poison to <vscale x 2 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f16_to_si64 = fptosi <vscale x 2 x half> poison to <vscale x 2 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f16_to_ui64 = fptoui <vscale x 2 x half> poison to <vscale x 2 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f32_to_si8 = fptosi <vscale x 2 x float> poison to <vscale x 2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f32_to_ui8 = fptoui <vscale x 2 x float> poison to <vscale x 2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f32_to_si16 = fptosi <vscale x 2 x float> poison to <vscale x 2 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f32_to_ui16 = fptoui <vscale x 2 x float> poison to <vscale x 2 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f32_to_si64 = fptosi <vscale x 2 x float> poison to <vscale x 2 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f32_to_ui64 = fptoui <vscale x 2 x float> poison to <vscale x 2 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f64_to_si8 = fptosi <vscale x 2 x double> poison to <vscale x 2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f64_to_ui8 = fptoui <vscale x 2 x double> poison to <vscale x 2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f64_to_si16 = fptosi <vscale x 2 x double> poison to <vscale x 2 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f64_to_ui16 = fptoui <vscale x 2 x double> poison to <vscale x 2 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f64_to_si32 = fptosi <vscale x 2 x double> poison to <vscale x 2 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv2f64_to_ui32 = fptoui <vscale x 2 x double> poison to <vscale x 2 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f16_to_si8 = fptosi <vscale x 4 x half> poison to <vscale x 4 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f16_to_ui8 = fptoui <vscale x 4 x half> poison to <vscale x 4 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f16_to_si32 = fptosi <vscale x 4 x half> poison to <vscale x 4 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f16_to_ui32 = fptoui <vscale x 4 x half> poison to <vscale x 4 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f16_to_si64 = fptosi <vscale x 4 x half> poison to <vscale x 4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f16_to_ui64 = fptoui <vscale x 4 x half> poison to <vscale x 4 x i64> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f32_to_si8 = fptosi <vscale x 4 x float> poison to <vscale x 4 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f32_to_ui8 = fptoui <vscale x 4 x float> poison to <vscale x 4 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f32_to_si16 = fptosi <vscale x 4 x float> poison to <vscale x 4 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv4f32_to_ui16 = fptoui <vscale x 4 x float> poison to <vscale x 4 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f32_to_si64 = fptosi <vscale x 4 x float> poison to <vscale x 4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f32_to_ui64 = fptoui <vscale x 4 x float> poison to <vscale x 4 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f64_to_si8 = fptosi <vscale x 4 x double> poison to <vscale x 4 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f64_to_ui8 = fptoui <vscale x 4 x double> poison to <vscale x 4 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f64_to_si16 = fptosi <vscale x 4 x double> poison to <vscale x 4 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f64_to_ui16 = fptoui <vscale x 4 x double> poison to <vscale x 4 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f64_to_si32 = fptosi <vscale x 4 x double> poison to <vscale x 4 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv4f64_to_ui32 = fptoui <vscale x 4 x double> poison to <vscale x 4 x i32> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv8f16_to_si8 = fptosi <vscale x 8 x half> poison to <vscale x 8 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nv8f16_to_ui8 = fptoui <vscale x 8 x half> poison to <vscale x 8 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f16_to_si32 = fptosi <vscale x 8 x half> poison to <vscale x 8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:4 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f16_to_ui32 = fptoui <vscale x 8 x half> poison to <vscale x 8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f16_to_si64 = fptosi <vscale x 8 x half> poison to <vscale x 8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:10 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f16_to_ui64 = fptoui <vscale x 8 x half> poison to <vscale x 8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f32_to_si8 = fptosi <vscale x 8 x float> poison to <vscale x 8 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f32_to_ui8 = fptoui <vscale x 8 x float> poison to <vscale x 8 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f32_to_si16 = fptosi <vscale x 8 x float> poison to <vscale x 8 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f32_to_ui16 = fptoui <vscale x 8 x float> poison to <vscale x 8 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f32_to_si64 = fptosi <vscale x 8 x float> poison to <vscale x 8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f32_to_ui64 = fptoui <vscale x 8 x float> poison to <vscale x 8 x i64> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f64_to_si8 = fptosi <vscale x 8 x double> poison to <vscale x 8 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f64_to_ui8 = fptoui <vscale x 8 x double> poison to <vscale x 8 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f64_to_si16 = fptosi <vscale x 8 x double> poison to <vscale x 8 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f64_to_ui16 = fptoui <vscale x 8 x double> poison to <vscale x 8 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f64_to_si32 = fptosi <vscale x 8 x double> poison to <vscale x 8 x i32> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %nv8f64_to_ui32 = fptoui <vscale x 8 x double> poison to <vscale x 8 x i32> ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %nv1f16_to_si8 = fptosi <vscale x 1 x half> undef to <vscale x 1 x i8> - %nv1f16_to_ui8 = fptoui <vscale x 1 x half> undef to <vscale x 1 x i8> - %nv1f16_to_si32 = fptosi <vscale x 1 x half> undef to <vscale x 1 x i32> - %nv1f16_to_ui32 = fptoui <vscale x 1 x half> undef to <vscale x 1 x i32> - %nv1f16_to_si64 = fptosi <vscale x 1 x half> undef to <vscale x 1 x i64> - %nv1f16_to_ui64 = fptoui <vscale x 1 x half> undef to <vscale x 1 x i64> + %nv1f16_to_si8 = fptosi <vscale x 1 x half> poison to <vscale x 1 x i8> + %nv1f16_to_ui8 = fptoui <vscale x 1 x half> poison to <vscale x 1 x i8> + %nv1f16_to_si32 = fptosi <vscale x 1 x half> poison to <vscale x 1 x i32> + %nv1f16_to_ui32 = fptoui <vscale x 1 x half> poison to <vscale x 1 x i32> + %nv1f16_to_si64 = fptosi <vscale x 1 x half> poison to <vscale x 1 x i64> + %nv1f16_to_ui64 = fptoui <vscale x 1 x half> poison to <vscale x 1 x i64> - %nv1f32_to_si8 = fptosi <vscale x 1 x float> undef to <vscale x 1 x i8> - %nv1f32_to_ui8 = fptoui <vscale x 1 x float> undef to <vscale x 1 x i8> - %nv1f32_to_si16 = fptosi <vscale x 1 x float> undef to <vscale x 1 x i16> - %nv1f32_to_ui16 = fptoui <vscale x 1 x float> undef to <vscale x 1 x i16> - %nv1f32_to_si64 = fptosi <vscale x 1 x float> undef to <vscale x 1 x i64> - %nv1f32_to_ui64 = fptoui <vscale x 1 x float> undef to <vscale x 1 x i64> + %nv1f32_to_si8 = fptosi <vscale x 1 x float> poison to <vscale x 1 x i8> + %nv1f32_to_ui8 = fptoui <vscale x 1 x float> poison to <vscale x 1 x i8> + %nv1f32_to_si16 = fptosi <vscale x 1 x float> poison to <vscale x 1 x i16> + %nv1f32_to_ui16 = fptoui <vscale x 1 x float> poison to <vscale x 1 x i16> + %nv1f32_to_si64 = fptosi <vscale x 1 x float> poison to <vscale x 1 x i64> + %nv1f32_to_ui64 = fptoui <vscale x 1 x float> poison to <vscale x 1 x i64> - %nv1f64_to_si8 = fptosi <vscale x 1 x double> undef to <vscale x 1 x i8> - %nv1f64_to_ui8 = fptoui <vscale x 1 x double> undef to <vscale x 1 x i8> - %nv1f64_to_si16 = fptosi <vscale x 1 x double> undef to <vscale x 1 x i16> - %nv1f64_to_ui16 = fptoui <vscale x 1 x double> undef to <vscale x 1 x i16> - %nv1f64_to_si32 = fptosi <vscale x 1 x double> undef to <vscale x 1 x i32> - %nv1f64_to_ui32 = fptoui <vscale x 1 x double> undef to <vscale x 1 x i32> + %nv1f64_to_si8 = fptosi <vscale x 1 x double> poison to <vscale x 1 x i8> + %nv1f64_to_ui8 = fptoui <vscale x 1 x double> poison to <vscale x 1 x i8> + %nv1f64_to_si16 = fptosi <vscale x 1 x double> poison to <vscale x 1 x i16> + %nv1f64_to_ui16 = fptoui <vscale x 1 x double> poison to <vscale x 1 x i16> + %nv1f64_to_si32 = fptosi <vscale x 1 x double> poison to <vscale x 1 x i32> + %nv1f64_to_ui32 = fptoui <vscale x 1 x double> poison to <vscale x 1 x i32> - %nv2f16_to_si8 = fptosi <vscale x 2 x half> undef to <vscale x 2 x i8> - %nv2f16_to_ui8 = fptoui <vscale x 2 x half> undef to <vscale x 2 x i8> - %nv2f16_to_si32 = fptosi <vscale x 2 x half> undef to <vscale x 2 x i32> - %nv2f16_to_ui32 = fptoui <vscale x 2 x half> undef to <vscale x 2 x i32> - %nv2f16_to_si64 = fptosi <vscale x 2 x half> undef to <vscale x 2 x i64> - %nv2f16_to_ui64 = fptoui <vscale x 2 x half> undef to <vscale x 2 x i64> + %nv2f16_to_si8 = fptosi <vscale x 2 x half> poison to <vscale x 2 x i8> + %nv2f16_to_ui8 = fptoui <vscale x 2 x half> poison to <vscale x 2 x i8> + %nv2f16_to_si32 = fptosi <vscale x 2 x half> poison to <vscale x 2 x i32> + %nv2f16_to_ui32 = fptoui <vscale x 2 x half> poison to <vscale x 2 x i32> + %nv2f16_to_si64 = fptosi <vscale x 2 x half> poison to <vscale x 2 x i64> + %nv2f16_to_ui64 = fptoui <vscale x 2 x half> poison to <vscale x 2 x i64> - %nv2f32_to_si8 = fptosi <vscale x 2 x float> undef to <vscale x 2 x i8> - %nv2f32_to_ui8 = fptoui <vscale x 2 x float> undef to <vscale x 2 x i8> - %nv2f32_to_si16 = fptosi <vscale x 2 x float> undef to <vscale x 2 x i16> - %nv2f32_to_ui16 = fptoui <vscale x 2 x float> undef to <vscale x 2 x i16> - %nv2f32_to_si64 = fptosi <vscale x 2 x float> undef to <vscale x 2 x i64> - %nv2f32_to_ui64 = fptoui <vscale x 2 x float> undef to <vscale x 2 x i64> + %nv2f32_to_si8 = fptosi <vscale x 2 x float> poison to <vscale x 2 x i8> + %nv2f32_to_ui8 = fptoui <vscale x 2 x float> poison to <vscale x 2 x i8> + %nv2f32_to_si16 = fptosi <vscale x 2 x float> poison to <vscale x 2 x i16> + %nv2f32_to_ui16 = fptoui <vscale x 2 x float> poison to <vscale x 2 x i16> + %nv2f32_to_si64 = fptosi <vscale x 2 x float> poison to <vscale x 2 x i64> + %nv2f32_to_ui64 = fptoui <vscale x 2 x float> poison to <vscale x 2 x i64> - %nv2f64_to_si8 = fptosi <vscale x 2 x double> undef to <vscale x 2 x i8> - %nv2f64_to_ui8 = fptoui <vscale x 2 x double> undef to <vscale x 2 x i8> - %nv2f64_to_si16 = fptosi <vscale x 2 x double> undef to <vscale x 2 x i16> - %nv2f64_to_ui16 = fptoui <vscale x 2 x double> undef to <vscale x 2 x i16> - %nv2f64_to_si32 = fptosi <vscale x 2 x double> undef to <vscale x 2 x i32> - %nv2f64_to_ui32 = fptoui <vscale x 2 x double> undef to <vscale x 2 x i32> + %nv2f64_to_si8 = fptosi <vscale x 2 x double> poison to <vscale x 2 x i8> + %nv2f64_to_ui8 = fptoui <vscale x 2 x double> poison to <vscale x 2 x i8> + %nv2f64_to_si16 = fptosi <vscale x 2 x double> poison to <vscale x 2 x i16> + %nv2f64_to_ui16 = fptoui <vscale x 2 x double> poison to <vscale x 2 x i16> + %nv2f64_to_si32 = fptosi <vscale x 2 x double> poison to <vscale x 2 x i32> + %nv2f64_to_ui32 = fptoui <vscale x 2 x double> poison to <vscale x 2 x i32> - %nv4f16_to_si8 = fptosi <vscale x 4 x half> undef to <vscale x 4 x i8> - %nv4f16_to_ui8 = fptoui <vscale x 4 x half> undef to <vscale x 4 x i8> - %nv4f16_to_si32 = fptosi <vscale x 4 x half> undef to <vscale x 4 x i32> - %nv4f16_to_ui32 = fptoui <vscale x 4 x half> undef to <vscale x 4 x i32> - %nv4f16_to_si64 = fptosi <vscale x 4 x half> undef to <vscale x 4 x i64> - %nv4f16_to_ui64 = fptoui <vscale x 4 x half> undef to <vscale x 4 x i64> + %nv4f16_to_si8 = fptosi <vscale x 4 x half> poison to <vscale x 4 x i8> + %nv4f16_to_ui8 = fptoui <vscale x 4 x half> poison to <vscale x 4 x i8> + %nv4f16_to_si32 = fptosi <vscale x 4 x half> poison to <vscale x 4 x i32> + %nv4f16_to_ui32 = fptoui <vscale x 4 x half> poison to <vscale x 4 x i32> + %nv4f16_to_si64 = fptosi <vscale x 4 x half> poison to <vscale x 4 x i64> + %nv4f16_to_ui64 = fptoui <vscale x 4 x half> poison to <vscale x 4 x i64> - %nv4f32_to_si8 = fptosi <vscale x 4 x float> undef to <vscale x 4 x i8> - %nv4f32_to_ui8 = fptoui <vscale x 4 x float> undef to <vscale x 4 x i8> - %nv4f32_to_si16 = fptosi <vscale x 4 x float> undef to <vscale x 4 x i16> - %nv4f32_to_ui16 = fptoui <vscale x 4 x float> undef to <vscale x 4 x i16> - %nv4f32_to_si64 = fptosi <vscale x 4 x float> undef to <vscale x 4 x i64> - %nv4f32_to_ui64 = fptoui <vscale x 4 x float> undef to <vscale x 4 x i64> + %nv4f32_to_si8 = fptosi <vscale x 4 x float> poison to <vscale x 4 x i8> + %nv4f32_to_ui8 = fptoui <vscale x 4 x float> poison to <vscale x 4 x i8> + %nv4f32_to_si16 = fptosi <vscale x 4 x float> poison to <vscale x 4 x i16> + %nv4f32_to_ui16 = fptoui <vscale x 4 x float> poison to <vscale x 4 x i16> + %nv4f32_to_si64 = fptosi <vscale x 4 x float> poison to <vscale x 4 x i64> + %nv4f32_to_ui64 = fptoui <vscale x 4 x float> poison to <vscale x 4 x i64> - %nv4f64_to_si8 = fptosi <vscale x 4 x double> undef to <vscale x 4 x i8> - %nv4f64_to_ui8 = fptoui <vscale x 4 x double> undef to <vscale x 4 x i8> - %nv4f64_to_si16 = fptosi <vscale x 4 x double> undef to <vscale x 4 x i16> - %nv4f64_to_ui16 = fptoui <vscale x 4 x double> undef to <vscale x 4 x i16> - %nv4f64_to_si32 = fptosi <vscale x 4 x double> undef to <vscale x 4 x i32> - %nv4f64_to_ui32 = fptoui <vscale x 4 x double> undef to <vscale x 4 x i32> + %nv4f64_to_si8 = fptosi <vscale x 4 x double> poison to <vscale x 4 x i8> + %nv4f64_to_ui8 = fptoui <vscale x 4 x double> poison to <vscale x 4 x i8> + %nv4f64_to_si16 = fptosi <vscale x 4 x double> poison to <vscale x 4 x i16> + %nv4f64_to_ui16 = fptoui <vscale x 4 x double> poison to <vscale x 4 x i16> + %nv4f64_to_si32 = fptosi <vscale x 4 x double> poison to <vscale x 4 x i32> + %nv4f64_to_ui32 = fptoui <vscale x 4 x double> poison to <vscale x 4 x i32> - %nv8f16_to_si8 = fptosi <vscale x 8 x half> undef to <vscale x 8 x i8> - %nv8f16_to_ui8 = fptoui <vscale x 8 x half> undef to <vscale x 8 x i8> - %nv8f16_to_si32 = fptosi <vscale x 8 x half> undef to <vscale x 8 x i32> - %nv8f16_to_ui32 = fptoui <vscale x 8 x half> undef to <vscale x 8 x i32> - %nv8f16_to_si64 = fptosi <vscale x 8 x half> undef to <vscale x 8 x i64> - %nv8f16_to_ui64 = fptoui <vscale x 8 x half> undef to <vscale x 8 x i64> + %nv8f16_to_si8 = fptosi <vscale x 8 x half> poison to <vscale x 8 x i8> + %nv8f16_to_ui8 = fptoui <vscale x 8 x half> poison to <vscale x 8 x i8> + %nv8f16_to_si32 = fptosi <vscale x 8 x half> poison to <vscale x 8 x i32> + %nv8f16_to_ui32 = fptoui <vscale x 8 x half> poison to <vscale x 8 x i32> + %nv8f16_to_si64 = fptosi <vscale x 8 x half> poison to <vscale x 8 x i64> + %nv8f16_to_ui64 = fptoui <vscale x 8 x half> poison to <vscale x 8 x i64> - %nv8f32_to_si8 = fptosi <vscale x 8 x float> undef to <vscale x 8 x i8> - %nv8f32_to_ui8 = fptoui <vscale x 8 x float> undef to <vscale x 8 x i8> - %nv8f32_to_si16 = fptosi <vscale x 8 x float> undef to <vscale x 8 x i16> - %nv8f32_to_ui16 = fptoui <vscale x 8 x float> undef to <vscale x 8 x i16> - %nv8f32_to_si64 = fptosi <vscale x 8 x float> undef to <vscale x 8 x i64> - %nv8f32_to_ui64 = fptoui <vscale x 8 x float> undef to <vscale x 8 x i64> + %nv8f32_to_si8 = fptosi <vscale x 8 x float> poison to <vscale x 8 x i8> + %nv8f32_to_ui8 = fptoui <vscale x 8 x float> poison to <vscale x 8 x i8> + %nv8f32_to_si16 = fptosi <vscale x 8 x float> poison to <vscale x 8 x i16> + %nv8f32_to_ui16 = fptoui <vscale x 8 x float> poison to <vscale x 8 x i16> + %nv8f32_to_si64 = fptosi <vscale x 8 x float> poison to <vscale x 8 x i64> + %nv8f32_to_ui64 = fptoui <vscale x 8 x float> poison to <vscale x 8 x i64> - %nv8f64_to_si8 = fptosi <vscale x 8 x double> undef to <vscale x 8 x i8> - %nv8f64_to_ui8 = fptoui <vscale x 8 x double> undef to <vscale x 8 x i8> - %nv8f64_to_si16 = fptosi <vscale x 8 x double> undef to <vscale x 8 x i16> - %nv8f64_to_ui16 = fptoui <vscale x 8 x double> undef to <vscale x 8 x i16> - %nv8f64_to_si32 = fptosi <vscale x 8 x double> undef to <vscale x 8 x i32> - %nv8f64_to_ui32 = fptoui <vscale x 8 x double> undef to <vscale x 8 x i32> + %nv8f64_to_si8 = fptosi <vscale x 8 x double> poison to <vscale x 8 x i8> + %nv8f64_to_ui8 = fptoui <vscale x 8 x double> poison to <vscale x 8 x i8> + %nv8f64_to_si16 = fptosi <vscale x 8 x double> poison to <vscale x 8 x i16> + %nv8f64_to_ui16 = fptoui <vscale x 8 x double> poison to <vscale x 8 x i16> + %nv8f64_to_si32 = fptosi <vscale x 8 x double> poison to <vscale x 8 x i32> + %nv8f64_to_ui32 = fptoui <vscale x 8 x double> poison to <vscale x 8 x i32> ret void } diff --git a/llvm/test/Analysis/CostModel/AArch64/sve-fptrunc.ll b/llvm/test/Analysis/CostModel/AArch64/sve-fptrunc.ll index 73556d7e..5b30c33 100644 --- a/llvm/test/Analysis/CostModel/AArch64/sve-fptrunc.ll +++ b/llvm/test/Analysis/CostModel/AArch64/sve-fptrunc.ll @@ -8,67 +8,67 @@ target triple = "aarch64-unknown-linux-gnu" define void @sve_fptruncs() { ; CHECK-LABEL: 'sve_fptruncs' -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f16_from_f32 = fptrunc <vscale x 2 x float> undef to <vscale x 2 x half> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv4_f16_from_f32 = fptrunc <vscale x 4 x float> undef to <vscale x 4 x half> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f32 = fptrunc <vscale x 8 x float> undef to <vscale x 8 x half> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f16_from_f64 = fptrunc <vscale x 2 x double> undef to <vscale x 2 x half> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_from_f64 = fptrunc <vscale x 4 x double> undef to <vscale x 4 x half> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f64 = fptrunc <vscale x 8 x double> undef to <vscale x 8 x half> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f32_from_f64 = fptrunc <vscale x 2 x double> undef to <vscale x 2 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f32_from_f64 = fptrunc <vscale x 4 x double> undef to <vscale x 4 x float> -; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f32_from_f64 = fptrunc <vscale x 8 x double> undef to <vscale x 8 x float> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f16_from_f32 = fptrunc <vscale x 2 x float> poison to <vscale x 2 x half> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv4_f16_from_f32 = fptrunc <vscale x 4 x float> poison to <vscale x 4 x half> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f32 = fptrunc <vscale x 8 x float> poison to <vscale x 8 x half> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f16_from_f64 = fptrunc <vscale x 2 x double> poison to <vscale x 2 x half> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_from_f64 = fptrunc <vscale x 4 x double> poison to <vscale x 4 x half> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f64 = fptrunc <vscale x 8 x double> poison to <vscale x 8 x half> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %nxv2_f32_from_f64 = fptrunc <vscale x 2 x double> poison to <vscale x 2 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f32_from_f64 = fptrunc <vscale x 4 x double> poison to <vscale x 4 x float> +; CHECK-NEXT: Cost Model: Found costs of RThru:6 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f32_from_f64 = fptrunc <vscale x 8 x double> poison to <vscale x 8 x float> ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %nxv2_f16_from_f32 = fptrunc <vscale x 2 x float> undef to <vscale x 2 x half> - %nxv4_f16_from_f32 = fptrunc <vscale x 4 x float> undef to <vscale x 4 x half> - %nxv8_f16_from_f32 = fptrunc <vscale x 8 x float> undef to <vscale x 8 x half> + %nxv2_f16_from_f32 = fptrunc <vscale x 2 x float> poison to <vscale x 2 x half> + %nxv4_f16_from_f32 = fptrunc <vscale x 4 x float> poison to <vscale x 4 x half> + %nxv8_f16_from_f32 = fptrunc <vscale x 8 x float> poison to <vscale x 8 x half> - %nxv2_f16_from_f64 = fptrunc <vscale x 2 x double> undef to <vscale x 2 x half> - %nxv4_f16_from_f64 = fptrunc <vscale x 4 x double> undef to <vscale x 4 x half> - %nxv8_f16_from_f64 = fptrunc <vscale x 8 x double> undef to <vscale x 8 x half> + %nxv2_f16_from_f64 = fptrunc <vscale x 2 x double> poison to <vscale x 2 x half> + %nxv4_f16_from_f64 = fptrunc <vscale x 4 x double> poison to <vscale x 4 x half> + %nxv8_f16_from_f64 = fptrunc <vscale x 8 x double> poison to <vscale x 8 x half> - %nxv2_f32_from_f64 = fptrunc <vscale x 2 x double> undef to <vscale x 2 x float> - %nxv4_f32_from_f64 = fptrunc <vscale x 4 x double> undef to <vscale x 4 x float> - %nxv8_f32_from_f64 = fptrunc <vscale x 8 x double> undef to <vscale x 8 x float> + %nxv2_f32_from_f64 = fptrunc <vscale x 2 x double> poison to <vscale x 2 x float> + %nxv4_f32_from_f64 = fptrunc <vscale x 4 x double> poison to <vscale x 4 x float> + %nxv8_f32_from_f64 = fptrunc <vscale x 8 x double> poison to <vscale x 8 x float> ret void } define void @sve_fptruncs_bf16() { ; CHECK-SVE-LABEL: 'sve_fptruncs_bf16' -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %nxv2_f16_from_f32 = fptrunc <vscale x 2 x float> undef to <vscale x 2 x bfloat> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_from_f32 = fptrunc <vscale x 4 x float> undef to <vscale x 4 x bfloat> -; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:17 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f32 = fptrunc <vscale x 8 x float> undef to <vscale x 8 x bfloat> -; CHECK-SVE-NEXT: Cost Model: Found costs of Invalid for: %nxv2_f16_from_f64 = fptrunc <vscale x 2 x double> undef to <vscale x 2 x bfloat> -; CHECK-SVE-NEXT: Cost Model: Found costs of Invalid for: %nxv4_f16_from_f64 = fptrunc <vscale x 4 x double> undef to <vscale x 4 x bfloat> -; CHECK-SVE-NEXT: Cost Model: Found costs of Invalid for: %nxv8_f16_from_f64 = fptrunc <vscale x 8 x double> undef to <vscale x 8 x bfloat> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %nxv2_f16_from_f32 = fptrunc <vscale x 2 x float> poison to <vscale x 2 x bfloat> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_from_f32 = fptrunc <vscale x 4 x float> poison to <vscale x 4 x bfloat> +; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:17 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f32 = fptrunc <vscale x 8 x float> poison to <vscale x 8 x bfloat> +; CHECK-SVE-NEXT: Cost Model: Found costs of Invalid for: %nxv2_f16_from_f64 = fptrunc <vscale x 2 x double> poison to <vscale x 2 x bfloat> +; CHECK-SVE-NEXT: Cost Model: Found costs of Invalid for: %nxv4_f16_from_f64 = fptrunc <vscale x 4 x double> poison to <vscale x 4 x bfloat> +; CHECK-SVE-NEXT: Cost Model: Found costs of Invalid for: %nxv8_f16_from_f64 = fptrunc <vscale x 8 x double> poison to <vscale x 8 x bfloat> ; CHECK-SVE-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-SVE2-LABEL: 'sve_fptruncs_bf16' -; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %nxv2_f16_from_f32 = fptrunc <vscale x 2 x float> undef to <vscale x 2 x bfloat> -; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_from_f32 = fptrunc <vscale x 4 x float> undef to <vscale x 4 x bfloat> -; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:17 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f32 = fptrunc <vscale x 8 x float> undef to <vscale x 8 x bfloat> -; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %nxv2_f16_from_f64 = fptrunc <vscale x 2 x double> undef to <vscale x 2 x bfloat> -; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_from_f64 = fptrunc <vscale x 4 x double> undef to <vscale x 4 x bfloat> -; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:39 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f64 = fptrunc <vscale x 8 x double> undef to <vscale x 8 x bfloat> +; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %nxv2_f16_from_f32 = fptrunc <vscale x 2 x float> poison to <vscale x 2 x bfloat> +; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:8 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_from_f32 = fptrunc <vscale x 4 x float> poison to <vscale x 4 x bfloat> +; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:17 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f32 = fptrunc <vscale x 8 x float> poison to <vscale x 8 x bfloat> +; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:9 CodeSize:1 Lat:1 SizeLat:1 for: %nxv2_f16_from_f64 = fptrunc <vscale x 2 x double> poison to <vscale x 2 x bfloat> +; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:19 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_from_f64 = fptrunc <vscale x 4 x double> poison to <vscale x 4 x bfloat> +; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:39 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f64 = fptrunc <vscale x 8 x double> poison to <vscale x 8 x bfloat> ; CHECK-SVE2-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-BF16-LABEL: 'sve_fptruncs_bf16' -; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %nxv2_f16_from_f32 = fptrunc <vscale x 2 x float> undef to <vscale x 2 x bfloat> -; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %nxv4_f16_from_f32 = fptrunc <vscale x 4 x float> undef to <vscale x 4 x bfloat> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f32 = fptrunc <vscale x 8 x float> undef to <vscale x 8 x bfloat> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %nxv2_f16_from_f64 = fptrunc <vscale x 2 x double> undef to <vscale x 2 x bfloat> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_from_f64 = fptrunc <vscale x 4 x double> undef to <vscale x 4 x bfloat> -; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f64 = fptrunc <vscale x 8 x double> undef to <vscale x 8 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %nxv2_f16_from_f32 = fptrunc <vscale x 2 x float> poison to <vscale x 2 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of 1 for: %nxv4_f16_from_f32 = fptrunc <vscale x 4 x float> poison to <vscale x 4 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f32 = fptrunc <vscale x 8 x float> poison to <vscale x 8 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %nxv2_f16_from_f64 = fptrunc <vscale x 2 x double> poison to <vscale x 2 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %nxv4_f16_from_f64 = fptrunc <vscale x 4 x double> poison to <vscale x 4 x bfloat> +; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %nxv8_f16_from_f64 = fptrunc <vscale x 8 x double> poison to <vscale x 8 x bfloat> ; CHECK-BF16-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %nxv2_f16_from_f32 = fptrunc <vscale x 2 x float> undef to <vscale x 2 x bfloat> - %nxv4_f16_from_f32 = fptrunc <vscale x 4 x float> undef to <vscale x 4 x bfloat> - %nxv8_f16_from_f32 = fptrunc <vscale x 8 x float> undef to <vscale x 8 x bfloat> + %nxv2_f16_from_f32 = fptrunc <vscale x 2 x float> poison to <vscale x 2 x bfloat> + %nxv4_f16_from_f32 = fptrunc <vscale x 4 x float> poison to <vscale x 4 x bfloat> + %nxv8_f16_from_f32 = fptrunc <vscale x 8 x float> poison to <vscale x 8 x bfloat> - %nxv2_f16_from_f64 = fptrunc <vscale x 2 x double> undef to <vscale x 2 x bfloat> - %nxv4_f16_from_f64 = fptrunc <vscale x 4 x double> undef to <vscale x 4 x bfloat> - %nxv8_f16_from_f64 = fptrunc <vscale x 8 x double> undef to <vscale x 8 x bfloat> + %nxv2_f16_from_f64 = fptrunc <vscale x 2 x double> poison to <vscale x 2 x bfloat> + %nxv4_f16_from_f64 = fptrunc <vscale x 4 x double> poison to <vscale x 4 x bfloat> + %nxv8_f16_from_f64 = fptrunc <vscale x 8 x double> poison to <vscale x 8 x bfloat> ret void } diff --git a/llvm/test/Analysis/CostModel/AArch64/sve-illegal-types.ll b/llvm/test/Analysis/CostModel/AArch64/sve-illegal-types.ll index 3e85760..2c838e2 100644 --- a/llvm/test/Analysis/CostModel/AArch64/sve-illegal-types.ll +++ b/llvm/test/Analysis/CostModel/AArch64/sve-illegal-types.ll @@ -1,14 +1,17 @@ +; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py UTC_ARGS: --version 6 ; RUN: opt -passes="print<cost-model>" 2>&1 -disable-output -mtriple=aarch64--linux-gnu -mattr=+sve < %s | FileCheck %s target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128" define void @load_store(ptr %ptrs) { ; CHECK-LABEL: 'load_store' -; CHECK-NEXT: Invalid cost for instruction: %load1 = load <vscale x 1 x i128>, ptr undef -; CHECK-NEXT: Invalid cost for instruction: %load2 = load <vscale x 2 x i128>, ptr undef -; CHECK-NEXT: Invalid cost for instruction: %load3 = load <vscale x 1 x fp128>, ptr undef -; CHECK-NEXT: Invalid cost for instruction: %load4 = load <vscale x 2 x fp128>, ptr undef -; CHECK-NEXT: Invalid cost for instruction: store <vscale x 1 x i128> %load1, ptr %ptrs +; CHECK-NEXT: Cost Model: Invalid cost for instruction: %load1 = load <vscale x 1 x i128>, ptr undef, align 16 +; CHECK-NEXT: Cost Model: Invalid cost for instruction: %load2 = load <vscale x 2 x i128>, ptr undef, align 32 +; CHECK-NEXT: Cost Model: Invalid cost for instruction: %load3 = load <vscale x 1 x fp128>, ptr undef, align 16 +; CHECK-NEXT: Cost Model: Invalid cost for instruction: %load4 = load <vscale x 2 x fp128>, ptr undef, align 32 +; CHECK-NEXT: Cost Model: Invalid cost for instruction: store <vscale x 1 x i128> %load1, ptr %ptrs, align 16 +; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void +; %load1 = load <vscale x 1 x i128>, ptr undef %load2 = load <vscale x 2 x i128>, ptr undef %load3 = load <vscale x 1 x fp128>, ptr undef @@ -19,8 +22,10 @@ define void @load_store(ptr %ptrs) { define void @masked_load_store(ptr %ptrs, ptr %val, <vscale x 1 x i1> %mask, <vscale x 1 x i128> %passthru) { ; CHECK-LABEL: 'masked_load_store' -; CHECK-NEXT: Invalid cost for instruction: %mload = call <vscale x 1 x i128> @llvm.masked.load.nxv1i128.p0(ptr %val, i32 8, <vscale x 1 x i1> %mask, <vscale x 1 x i128> %passthru) -; CHECK-NEXT: Invalid cost for instruction: call void @llvm.masked.store.nxv1i128.p0(<vscale x 1 x i128> %mload, ptr %ptrs, i32 8, <vscale x 1 x i1> %mask) +; CHECK-NEXT: Cost Model: Invalid cost for instruction: %mload = call <vscale x 1 x i128> @llvm.masked.load.nxv1i128.p0(ptr %val, i32 8, <vscale x 1 x i1> %mask, <vscale x 1 x i128> %passthru) +; CHECK-NEXT: Cost Model: Invalid cost for instruction: call void @llvm.masked.store.nxv1i128.p0(<vscale x 1 x i128> %mload, ptr %ptrs, i32 8, <vscale x 1 x i1> %mask) +; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void +; %mload = call <vscale x 1 x i128> @llvm.masked.load.nxv1i128(ptr %val, i32 8, <vscale x 1 x i1> %mask, <vscale x 1 x i128> %passthru) call void @llvm.masked.store.nxv1i128(<vscale x 1 x i128> %mload, ptr %ptrs, i32 8, <vscale x 1 x i1> %mask) ret void @@ -28,8 +33,10 @@ define void @masked_load_store(ptr %ptrs, ptr %val, <vscale x 1 x i1> %mask, <vs define void @masked_gather_scatter(<vscale x 1 x ptr> %ptrs, <vscale x 1 x ptr> %val, <vscale x 1 x i1> %mask, <vscale x 1 x i128> %passthru) { ; CHECK-LABEL: 'masked_gather_scatter' -; CHECK-NEXT: Invalid cost for instruction: %mgather = call <vscale x 1 x i128> @llvm.masked.gather.nxv1i128.nxv1p0(<vscale x 1 x ptr> %val, i32 0, <vscale x 1 x i1> %mask, <vscale x 1 x i128> %passthru) -; CHECK-NEXT: Invalid cost for instruction: call void @llvm.masked.scatter.nxv1i128.nxv1p0(<vscale x 1 x i128> %mgather, <vscale x 1 x ptr> %ptrs, i32 0, <vscale x 1 x i1> %mask) +; CHECK-NEXT: Cost Model: Invalid cost for instruction: %mgather = call <vscale x 1 x i128> @llvm.masked.gather.nxv1i128.nxv1p0(<vscale x 1 x ptr> %val, i32 0, <vscale x 1 x i1> %mask, <vscale x 1 x i128> %passthru) +; CHECK-NEXT: Cost Model: Invalid cost for instruction: call void @llvm.masked.scatter.nxv1i128.nxv1p0(<vscale x 1 x i128> %mgather, <vscale x 1 x ptr> %ptrs, i32 0, <vscale x 1 x i1> %mask) +; CHECK-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret void +; %mgather = call <vscale x 1 x i128> @llvm.masked.gather.nxv1i128(<vscale x 1 x ptr> %val, i32 0, <vscale x 1 x i1> %mask, <vscale x 1 x i128> %passthru) call void @llvm.masked.scatter.nxv1i128(<vscale x 1 x i128> %mgather, <vscale x 1 x ptr> %ptrs, i32 0, <vscale x 1 x i1> %mask) ret void diff --git a/llvm/test/Analysis/CostModel/AArch64/sve-intrinsics.ll b/llvm/test/Analysis/CostModel/AArch64/sve-intrinsics.ll index 609a23b..0976a10 100644 --- a/llvm/test/Analysis/CostModel/AArch64/sve-intrinsics.ll +++ b/llvm/test/Analysis/CostModel/AArch64/sve-intrinsics.ll @@ -40,50 +40,50 @@ declare <vscale x 16 x i32> @llvm.vector.insert.nxv16i32.nxv4i32(<vscale x 16 x define void @vector_insert_extract_idxzero_128b() #1 { ; CHECK-VSCALE-1-LABEL: 'vector_insert_extract_idxzero_128b' -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 0 for: %insert_legal_fixed_into_scalable = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.v4f32(<vscale x 4 x float> undef, <4 x float> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 0 for: %extract_legal_fixed_from_scalable = call <2 x double> @llvm.vector.extract.v2f64.nxv2f64(<vscale x 2 x double> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.nxv2i1(<vscale x 16 x i1> undef, <vscale x 2 x i1> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:48 CodeSize:32 Lat:48 SizeLat:48 for: %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> undef, <2 x float> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:12 CodeSize:6 Lat:12 SizeLat:12 for: %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> undef, <vscale x 2 x float> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> undef, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 0 for: %insert_legal_fixed_into_scalable = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.v4f32(<vscale x 4 x float> poison, <4 x float> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 0 for: %extract_legal_fixed_from_scalable = call <2 x double> @llvm.vector.extract.v2f64.nxv2f64(<vscale x 2 x double> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.nxv2i1(<vscale x 16 x i1> poison, <vscale x 2 x i1> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:48 CodeSize:32 Lat:48 SizeLat:48 for: %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> poison, <2 x float> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:12 CodeSize:6 Lat:12 SizeLat:12 for: %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> poison, <vscale x 2 x float> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> poison, i64 0) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-VSCALE-2-LABEL: 'vector_insert_extract_idxzero_128b' -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 0 for: %insert_legal_fixed_into_scalable = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.v4f32(<vscale x 4 x float> undef, <4 x float> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 0 for: %extract_legal_fixed_from_scalable = call <2 x double> @llvm.vector.extract.v2f64.nxv2f64(<vscale x 2 x double> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.nxv2i1(<vscale x 16 x i1> undef, <vscale x 2 x i1> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:48 CodeSize:32 Lat:48 SizeLat:48 for: %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> undef, <2 x float> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:12 CodeSize:6 Lat:12 SizeLat:12 for: %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> undef, <vscale x 2 x float> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> undef, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 0 for: %insert_legal_fixed_into_scalable = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.v4f32(<vscale x 4 x float> poison, <4 x float> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 0 for: %extract_legal_fixed_from_scalable = call <2 x double> @llvm.vector.extract.v2f64.nxv2f64(<vscale x 2 x double> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.nxv2i1(<vscale x 16 x i1> poison, <vscale x 2 x i1> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:48 CodeSize:32 Lat:48 SizeLat:48 for: %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> poison, <2 x float> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:12 CodeSize:6 Lat:12 SizeLat:12 for: %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> poison, <vscale x 2 x float> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> poison, i64 0) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; TYPE_BASED_ONLY-LABEL: 'vector_insert_extract_idxzero_128b' -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_legal_fixed_into_scalable = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.v4f32(<vscale x 4 x float> undef, <4 x float> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_legal_fixed_from_scalable = call <2 x double> @llvm.vector.extract.v2f64.nxv2f64(<vscale x 2 x double> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.nxv2i1(<vscale x 16 x i1> undef, <vscale x 2 x i1> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> undef, <2 x float> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> undef, <vscale x 2 x float> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> undef, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_legal_fixed_into_scalable = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.v4f32(<vscale x 4 x float> poison, <4 x float> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_legal_fixed_from_scalable = call <2 x double> @llvm.vector.extract.v2f64.nxv2f64(<vscale x 2 x double> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.nxv2i1(<vscale x 16 x i1> poison, <vscale x 2 x i1> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> poison, <2 x float> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> poison, <vscale x 2 x float> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> poison, i64 0) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %insert_legal_fixed_into_scalable = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.v4f32(<vscale x 4 x float> undef, <4 x float> undef, i64 0) - %extract_legal_fixed_from_scalable = call <2 x double> @llvm.vector.extract.v2f64.nxv2f64(<vscale x 2 x double> undef, i64 0) - %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.v2i1(<vscale x 16 x i1> undef, <vscale x 2 x i1> undef, i64 0) - %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> undef, i64 0) - %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> undef, i64 0) - %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> undef, <2 x float> undef, i64 0) - %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> undef, i64 0) - %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> undef, <vscale x 2 x float> undef, i64 0) - %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> undef, i64 0) + %insert_legal_fixed_into_scalable = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.v4f32(<vscale x 4 x float> poison, <4 x float> poison, i64 0) + %extract_legal_fixed_from_scalable = call <2 x double> @llvm.vector.extract.v2f64.nxv2f64(<vscale x 2 x double> poison, i64 0) + %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.v2i1(<vscale x 16 x i1> poison, <vscale x 2 x i1> poison, i64 0) + %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> poison, i64 0) + %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> poison, i64 0) + %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> poison, <2 x float> poison, i64 0) + %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> poison, i64 0) + %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> poison, <vscale x 2 x float> poison, i64 0) + %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> poison, i64 0) ret void } declare <vscale x 4 x float> @llvm.vector.insert.nxv4f32.v4f32(<vscale x 4 x float>, <4 x float>, i64) @@ -97,50 +97,50 @@ declare <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x define void @vector_insert_extract_idxzero_256b() #2 { ; CHECK-VSCALE-1-LABEL: 'vector_insert_extract_idxzero_256b' -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 0 for: %insert_legal_fixed_into_scalable = call <vscale x 8 x i16> @llvm.vector.insert.nxv8i16.v16i16(<vscale x 8 x i16> undef, <16 x i16> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 0 for: %extract_legal_fixed_from_scalable = call <8 x float> @llvm.vector.extract.v8f32.nxv4f32(<vscale x 4 x float> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.nxv2i1(<vscale x 16 x i1> undef, <vscale x 2 x i1> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:48 CodeSize:32 Lat:48 SizeLat:48 for: %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> undef, <2 x float> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:12 CodeSize:6 Lat:12 SizeLat:12 for: %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> undef, <vscale x 2 x float> undef, i64 0) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> undef, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 0 for: %insert_legal_fixed_into_scalable = call <vscale x 8 x i16> @llvm.vector.insert.nxv8i16.v16i16(<vscale x 8 x i16> poison, <16 x i16> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 0 for: %extract_legal_fixed_from_scalable = call <8 x float> @llvm.vector.extract.v8f32.nxv4f32(<vscale x 4 x float> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.nxv2i1(<vscale x 16 x i1> poison, <vscale x 2 x i1> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:48 CodeSize:32 Lat:48 SizeLat:48 for: %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> poison, <2 x float> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:12 CodeSize:6 Lat:12 SizeLat:12 for: %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> poison, <vscale x 2 x float> poison, i64 0) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> poison, i64 0) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-VSCALE-2-LABEL: 'vector_insert_extract_idxzero_256b' -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 0 for: %insert_legal_fixed_into_scalable = call <vscale x 8 x i16> @llvm.vector.insert.nxv8i16.v16i16(<vscale x 8 x i16> undef, <16 x i16> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 0 for: %extract_legal_fixed_from_scalable = call <8 x float> @llvm.vector.extract.v8f32.nxv4f32(<vscale x 4 x float> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.nxv2i1(<vscale x 16 x i1> undef, <vscale x 2 x i1> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:48 CodeSize:32 Lat:48 SizeLat:48 for: %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> undef, <2 x float> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:12 CodeSize:6 Lat:12 SizeLat:12 for: %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> undef, <vscale x 2 x float> undef, i64 0) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> undef, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 0 for: %insert_legal_fixed_into_scalable = call <vscale x 8 x i16> @llvm.vector.insert.nxv8i16.v16i16(<vscale x 8 x i16> poison, <16 x i16> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 0 for: %extract_legal_fixed_from_scalable = call <8 x float> @llvm.vector.extract.v8f32.nxv4f32(<vscale x 4 x float> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.nxv2i1(<vscale x 16 x i1> poison, <vscale x 2 x i1> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:48 CodeSize:32 Lat:48 SizeLat:48 for: %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:4 CodeSize:2 Lat:4 SizeLat:4 for: %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> poison, <2 x float> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:12 CodeSize:6 Lat:12 SizeLat:12 for: %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> poison, <vscale x 2 x float> poison, i64 0) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> poison, i64 0) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; TYPE_BASED_ONLY-LABEL: 'vector_insert_extract_idxzero_256b' -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_legal_fixed_into_scalable = call <vscale x 8 x i16> @llvm.vector.insert.nxv8i16.v16i16(<vscale x 8 x i16> undef, <16 x i16> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_legal_fixed_from_scalable = call <8 x float> @llvm.vector.extract.v8f32.nxv4f32(<vscale x 4 x float> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.nxv2i1(<vscale x 16 x i1> undef, <vscale x 2 x i1> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> undef, <2 x float> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> undef, <vscale x 2 x float> undef, i64 0) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> undef, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_legal_fixed_into_scalable = call <vscale x 8 x i16> @llvm.vector.insert.nxv8i16.v16i16(<vscale x 8 x i16> poison, <16 x i16> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_legal_fixed_from_scalable = call <8 x float> @llvm.vector.extract.v8f32.nxv4f32(<vscale x 4 x float> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.nxv2i1(<vscale x 16 x i1> poison, <vscale x 2 x i1> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> poison, <2 x float> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> poison, <vscale x 2 x float> poison, i64 0) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> poison, i64 0) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %insert_legal_fixed_into_scalable = call <vscale x 8 x i16> @llvm.vector.insert.nxv8i16.v16i16(<vscale x 8 x i16> undef, <16 x i16> undef, i64 0) - %extract_legal_fixed_from_scalable = call <8 x float> @llvm.vector.extract.v8f32.nx4f32(<vscale x 4 x float> undef, i64 0) - %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.v2i1(<vscale x 16 x i1> undef, <vscale x 2 x i1> undef, i64 0) - %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> undef, i64 0) - %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> undef, i64 0) - %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> undef, <2 x float> undef, i64 0) - %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> undef, i64 0) - %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> undef, <vscale x 2 x float> undef, i64 0) - %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> undef, i64 0) + %insert_legal_fixed_into_scalable = call <vscale x 8 x i16> @llvm.vector.insert.nxv8i16.v16i16(<vscale x 8 x i16> poison, <16 x i16> poison, i64 0) + %extract_legal_fixed_from_scalable = call <8 x float> @llvm.vector.extract.v8f32.nx4f32(<vscale x 4 x float> poison, i64 0) + %insert_nxv16i1_nxv2i1 = call <vscale x 16 x i1> @llvm.vector.insert.nxv16i1.v2i1(<vscale x 16 x i1> poison, <vscale x 2 x i1> poison, i64 0) + %extract_nxv4i1_nxv16i1 = call <vscale x 4 x i1> @llvm.vector.extract.nxv4i1.nxv16i1(<vscale x 16 x i1> poison, i64 0) + %extract_v8i1_nxv8i1 = call <8 x i1> @llvm.vector.extract.v8i1.nxv8i1(<vscale x 8 x i1> poison, i64 0) + %insert_v2f32_nxv2f32 = call <vscale x 2 x float> @llvm.vector.insert.nxv2f32.v2f32(<vscale x 2 x float> poison, <2 x float> poison, i64 0) + %extract_v4f16_nxv4f16 = call <4 x half> @llvm.vector.extract.v4f16.nxv4f16(<vscale x 4 x half> poison, i64 0) + %insert_nxv2f32_nxv4f32 = call <vscale x 4 x float> @llvm.vector.insert.nxv4f32.nxv2f32(<vscale x 4 x float> poison, <vscale x 2 x float> poison, i64 0) + %extract_nxv4f32_nxv8f32 = call <vscale x 4 x float> @llvm.vector.extract.nxv4f32.nxv8f32(<vscale x 8 x float> poison, i64 0) ret void } declare <vscale x 8 x i16> @llvm.vector.insert.nxv8i16.v16i16(<vscale x 8 x i16>, <16 x i16>, i64) @@ -148,157 +148,157 @@ declare <8 x float> @llvm.vector.extract.v8f32.nxv4f32(<vscale x 4 x float>, i64 define void @reductions(<vscale x 4 x i32> %v0, <vscale x 4 x i64> %v1, <vscale x 4 x float> %v2, <vscale x 4 x double> %v3) { ; CHECK-VSCALE-1-LABEL: 'reductions' -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %add_nxv1i32 = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %add_nxv1i32 = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %add_nxv4i32 = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 3 for: %add_nxv4i64 = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %mul_nxv1i32 = call i32 @llvm.vector.reduce.mul.nxv1i32(<vscale x 1 x i32> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %mul_nxv1i32 = call i32 @llvm.vector.reduce.mul.nxv1i32(<vscale x 1 x i32> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %mul_nxv4i32 = call i32 @llvm.vector.reduce.mul.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %mul_nxv4i64 = call i64 @llvm.vector.reduce.mul.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %and_nxv1i32 = call i32 @llvm.vector.reduce.and.nxv1i32(<vscale x 1 x i32> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %and_nxv1i32 = call i32 @llvm.vector.reduce.and.nxv1i32(<vscale x 1 x i32> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %and_nxv4i32 = call i32 @llvm.vector.reduce.and.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 3 for: %and_nxv4i64 = call i64 @llvm.vector.reduce.and.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %or_nxv1i32 = call i32 @llvm.vector.reduce.or.nxv1i32(<vscale x 1 x i32> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %or_nxv1i32 = call i32 @llvm.vector.reduce.or.nxv1i32(<vscale x 1 x i32> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %or_nxv4i32 = call i32 @llvm.vector.reduce.or.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 3 for: %or_nxv4i64 = call i64 @llvm.vector.reduce.or.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %xor_nxv1i32 = call i32 @llvm.vector.reduce.xor.nxv1i32(<vscale x 1 x i32> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %xor_nxv1i32 = call i32 @llvm.vector.reduce.xor.nxv1i32(<vscale x 1 x i32> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %xor_nxv4i32 = call i32 @llvm.vector.reduce.xor.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 3 for: %xor_nxv4i64 = call i64 @llvm.vector.reduce.xor.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %umin_nxv1i64 = call i64 @llvm.vector.reduce.umin.nxv1i64(<vscale x 1 x i64> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %umin_nxv1i64 = call i64 @llvm.vector.reduce.umin.nxv1i64(<vscale x 1 x i64> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %umin_nxv4i32 = call i32 @llvm.vector.reduce.umin.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 3 for: %umin_nxv4i64 = call i64 @llvm.vector.reduce.umin.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %smin_nxv1i64 = call i64 @llvm.vector.reduce.smin.nxv1i64(<vscale x 1 x i64> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %smin_nxv1i64 = call i64 @llvm.vector.reduce.smin.nxv1i64(<vscale x 1 x i64> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %smin_nxv4i32 = call i32 @llvm.vector.reduce.smin.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 3 for: %smin_nxv4i64 = call i64 @llvm.vector.reduce.smin.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %umax_nxv1i64 = call i64 @llvm.vector.reduce.umax.nxv1i64(<vscale x 1 x i64> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %umax_nxv1i64 = call i64 @llvm.vector.reduce.umax.nxv1i64(<vscale x 1 x i64> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %umax_nxv4i32 = call i32 @llvm.vector.reduce.umax.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 3 for: %umax_nxv4i64 = call i64 @llvm.vector.reduce.umax.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %smax_nxv1i64 = call i64 @llvm.vector.reduce.smax.nxv1i64(<vscale x 1 x i64> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %smax_nxv1i64 = call i64 @llvm.vector.reduce.smax.nxv1i64(<vscale x 1 x i64> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %smax_nxv4i32 = call i32 @llvm.vector.reduce.smax.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 3 for: %smax_nxv4i64 = call i64 @llvm.vector.reduce.smax.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %fadd_nxv1f32 = call fast float @llvm.vector.reduce.fadd.nxv1f32(float 0.000000e+00, <vscale x 1 x float> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %fadd_nxv1f32 = call fast float @llvm.vector.reduce.fadd.nxv1f32(float 0.000000e+00, <vscale x 1 x float> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %fadd_nxv4f32 = call fast float @llvm.vector.reduce.fadd.nxv4f32(float 0.000000e+00, <vscale x 4 x float> %v2) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:3 CodeSize:3 Lat:5 SizeLat:3 for: %fadd_nxv4f64 = call fast double @llvm.vector.reduce.fadd.nxv4f64(double 0.000000e+00, <vscale x 4 x double> %v3) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %fmin_nxv1f32 = call fast float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %fmin_nxv1f32 = call fast float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %fmin_nxv4f32 = call fast float @llvm.vector.reduce.fmin.nxv4f32(<vscale x 4 x float> %v2) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 4 for: %fmin_nxv4f64 = call fast double @llvm.vector.reduce.fmin.nxv4f64(<vscale x 4 x double> %v3) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %fmax_nxv1f32 = call fast float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of Invalid for: %fmax_nxv1f32 = call fast float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %fmax_nxv4f32 = call fast float @llvm.vector.reduce.fmax.nxv4f32(<vscale x 4 x float> %v2) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 4 for: %fmax_nxv4f64 = call fast double @llvm.vector.reduce.fmax.nxv4f64(<vscale x 4 x double> %v3) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-VSCALE-2-LABEL: 'reductions' -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %add_nxv1i32 = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %add_nxv1i32 = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %add_nxv4i32 = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 3 for: %add_nxv4i64 = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %mul_nxv1i32 = call i32 @llvm.vector.reduce.mul.nxv1i32(<vscale x 1 x i32> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %mul_nxv1i32 = call i32 @llvm.vector.reduce.mul.nxv1i32(<vscale x 1 x i32> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %mul_nxv4i32 = call i32 @llvm.vector.reduce.mul.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %mul_nxv4i64 = call i64 @llvm.vector.reduce.mul.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %and_nxv1i32 = call i32 @llvm.vector.reduce.and.nxv1i32(<vscale x 1 x i32> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %and_nxv1i32 = call i32 @llvm.vector.reduce.and.nxv1i32(<vscale x 1 x i32> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %and_nxv4i32 = call i32 @llvm.vector.reduce.and.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 3 for: %and_nxv4i64 = call i64 @llvm.vector.reduce.and.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %or_nxv1i32 = call i32 @llvm.vector.reduce.or.nxv1i32(<vscale x 1 x i32> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %or_nxv1i32 = call i32 @llvm.vector.reduce.or.nxv1i32(<vscale x 1 x i32> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %or_nxv4i32 = call i32 @llvm.vector.reduce.or.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 3 for: %or_nxv4i64 = call i64 @llvm.vector.reduce.or.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %xor_nxv1i32 = call i32 @llvm.vector.reduce.xor.nxv1i32(<vscale x 1 x i32> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %xor_nxv1i32 = call i32 @llvm.vector.reduce.xor.nxv1i32(<vscale x 1 x i32> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %xor_nxv4i32 = call i32 @llvm.vector.reduce.xor.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 3 for: %xor_nxv4i64 = call i64 @llvm.vector.reduce.xor.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %umin_nxv1i64 = call i64 @llvm.vector.reduce.umin.nxv1i64(<vscale x 1 x i64> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %umin_nxv1i64 = call i64 @llvm.vector.reduce.umin.nxv1i64(<vscale x 1 x i64> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %umin_nxv4i32 = call i32 @llvm.vector.reduce.umin.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 3 for: %umin_nxv4i64 = call i64 @llvm.vector.reduce.umin.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %smin_nxv1i64 = call i64 @llvm.vector.reduce.smin.nxv1i64(<vscale x 1 x i64> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %smin_nxv1i64 = call i64 @llvm.vector.reduce.smin.nxv1i64(<vscale x 1 x i64> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %smin_nxv4i32 = call i32 @llvm.vector.reduce.smin.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 3 for: %smin_nxv4i64 = call i64 @llvm.vector.reduce.smin.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %umax_nxv1i64 = call i64 @llvm.vector.reduce.umax.nxv1i64(<vscale x 1 x i64> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %umax_nxv1i64 = call i64 @llvm.vector.reduce.umax.nxv1i64(<vscale x 1 x i64> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %umax_nxv4i32 = call i32 @llvm.vector.reduce.umax.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 3 for: %umax_nxv4i64 = call i64 @llvm.vector.reduce.umax.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %smax_nxv1i64 = call i64 @llvm.vector.reduce.smax.nxv1i64(<vscale x 1 x i64> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %smax_nxv1i64 = call i64 @llvm.vector.reduce.smax.nxv1i64(<vscale x 1 x i64> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %smax_nxv4i32 = call i32 @llvm.vector.reduce.smax.nxv4i32(<vscale x 4 x i32> %v0) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 3 for: %smax_nxv4i64 = call i64 @llvm.vector.reduce.smax.nxv4i64(<vscale x 4 x i64> %v1) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %fadd_nxv1f32 = call fast float @llvm.vector.reduce.fadd.nxv1f32(float 0.000000e+00, <vscale x 1 x float> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %fadd_nxv1f32 = call fast float @llvm.vector.reduce.fadd.nxv1f32(float 0.000000e+00, <vscale x 1 x float> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %fadd_nxv4f32 = call fast float @llvm.vector.reduce.fadd.nxv4f32(float 0.000000e+00, <vscale x 4 x float> %v2) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:3 CodeSize:3 Lat:5 SizeLat:3 for: %fadd_nxv4f64 = call fast double @llvm.vector.reduce.fadd.nxv4f64(double 0.000000e+00, <vscale x 4 x double> %v3) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %fmin_nxv1f32 = call fast float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %fmin_nxv1f32 = call fast float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %fmin_nxv4f32 = call fast float @llvm.vector.reduce.fmin.nxv4f32(<vscale x 4 x float> %v2) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 4 for: %fmin_nxv4f64 = call fast double @llvm.vector.reduce.fmin.nxv4f64(<vscale x 4 x double> %v3) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %fmax_nxv1f32 = call fast float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of Invalid for: %fmax_nxv1f32 = call fast float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %fmax_nxv4f32 = call fast float @llvm.vector.reduce.fmax.nxv4f32(<vscale x 4 x float> %v2) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 4 for: %fmax_nxv4f64 = call fast double @llvm.vector.reduce.fmax.nxv4f64(<vscale x 4 x double> %v3) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; TYPE_BASED_ONLY-LABEL: 'reductions' -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %add_nxv1i32 = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %add_nxv1i32 = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %add_nxv4i32 = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %v0) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 3 for: %add_nxv4i64 = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %v1) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %mul_nxv1i32 = call i32 @llvm.vector.reduce.mul.nxv1i32(<vscale x 1 x i32> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %mul_nxv1i32 = call i32 @llvm.vector.reduce.mul.nxv1i32(<vscale x 1 x i32> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %mul_nxv4i32 = call i32 @llvm.vector.reduce.mul.nxv4i32(<vscale x 4 x i32> %v0) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %mul_nxv4i64 = call i64 @llvm.vector.reduce.mul.nxv4i64(<vscale x 4 x i64> %v1) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %and_nxv1i32 = call i32 @llvm.vector.reduce.and.nxv1i32(<vscale x 1 x i32> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %and_nxv1i32 = call i32 @llvm.vector.reduce.and.nxv1i32(<vscale x 1 x i32> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %and_nxv4i32 = call i32 @llvm.vector.reduce.and.nxv4i32(<vscale x 4 x i32> %v0) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 3 for: %and_nxv4i64 = call i64 @llvm.vector.reduce.and.nxv4i64(<vscale x 4 x i64> %v1) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %or_nxv1i32 = call i32 @llvm.vector.reduce.or.nxv1i32(<vscale x 1 x i32> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %or_nxv1i32 = call i32 @llvm.vector.reduce.or.nxv1i32(<vscale x 1 x i32> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %or_nxv4i32 = call i32 @llvm.vector.reduce.or.nxv4i32(<vscale x 4 x i32> %v0) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 3 for: %or_nxv4i64 = call i64 @llvm.vector.reduce.or.nxv4i64(<vscale x 4 x i64> %v1) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %xor_nxv1i32 = call i32 @llvm.vector.reduce.xor.nxv1i32(<vscale x 1 x i32> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %xor_nxv1i32 = call i32 @llvm.vector.reduce.xor.nxv1i32(<vscale x 1 x i32> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %xor_nxv4i32 = call i32 @llvm.vector.reduce.xor.nxv4i32(<vscale x 4 x i32> %v0) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 3 for: %xor_nxv4i64 = call i64 @llvm.vector.reduce.xor.nxv4i64(<vscale x 4 x i64> %v1) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %umin_nxv1i64 = call i64 @llvm.vector.reduce.umin.nxv1i64(<vscale x 1 x i64> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %umin_nxv1i64 = call i64 @llvm.vector.reduce.umin.nxv1i64(<vscale x 1 x i64> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %umin_nxv4i32 = call i32 @llvm.vector.reduce.umin.nxv4i32(<vscale x 4 x i32> %v0) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 3 for: %umin_nxv4i64 = call i64 @llvm.vector.reduce.umin.nxv4i64(<vscale x 4 x i64> %v1) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %smin_nxv1i64 = call i64 @llvm.vector.reduce.smin.nxv1i64(<vscale x 1 x i64> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %smin_nxv1i64 = call i64 @llvm.vector.reduce.smin.nxv1i64(<vscale x 1 x i64> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %smin_nxv4i32 = call i32 @llvm.vector.reduce.smin.nxv4i32(<vscale x 4 x i32> %v0) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 3 for: %smin_nxv4i64 = call i64 @llvm.vector.reduce.smin.nxv4i64(<vscale x 4 x i64> %v1) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %umax_nxv1i64 = call i64 @llvm.vector.reduce.umax.nxv1i64(<vscale x 1 x i64> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %umax_nxv1i64 = call i64 @llvm.vector.reduce.umax.nxv1i64(<vscale x 1 x i64> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %umax_nxv4i32 = call i32 @llvm.vector.reduce.umax.nxv4i32(<vscale x 4 x i32> %v0) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 3 for: %umax_nxv4i64 = call i64 @llvm.vector.reduce.umax.nxv4i64(<vscale x 4 x i64> %v1) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %smax_nxv1i64 = call i64 @llvm.vector.reduce.smax.nxv1i64(<vscale x 1 x i64> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %smax_nxv1i64 = call i64 @llvm.vector.reduce.smax.nxv1i64(<vscale x 1 x i64> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %smax_nxv4i32 = call i32 @llvm.vector.reduce.smax.nxv4i32(<vscale x 4 x i32> %v0) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 3 for: %smax_nxv4i64 = call i64 @llvm.vector.reduce.smax.nxv4i64(<vscale x 4 x i64> %v1) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %fadd_nxv1f32 = call fast float @llvm.vector.reduce.fadd.nxv1f32(float 0.000000e+00, <vscale x 1 x float> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %fadd_nxv1f32 = call fast float @llvm.vector.reduce.fadd.nxv1f32(float 0.000000e+00, <vscale x 1 x float> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %fadd_nxv4f32 = call fast float @llvm.vector.reduce.fadd.nxv4f32(float 0.000000e+00, <vscale x 4 x float> %v2) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:3 CodeSize:3 Lat:5 SizeLat:3 for: %fadd_nxv4f64 = call fast double @llvm.vector.reduce.fadd.nxv4f64(double 0.000000e+00, <vscale x 4 x double> %v3) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %fmin_nxv1f32 = call fast float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %fmin_nxv1f32 = call fast float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %fmin_nxv4f32 = call fast float @llvm.vector.reduce.fmin.nxv4f32(<vscale x 4 x float> %v2) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 4 for: %fmin_nxv4f64 = call fast double @llvm.vector.reduce.fmin.nxv4f64(<vscale x 4 x double> %v3) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %fmax_nxv1f32 = call fast float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of Invalid for: %fmax_nxv1f32 = call fast float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %fmax_nxv4f32 = call fast float @llvm.vector.reduce.fmax.nxv4f32(<vscale x 4 x float> %v2) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 4 for: %fmax_nxv4f64 = call fast double @llvm.vector.reduce.fmax.nxv4f64(<vscale x 4 x double> %v3) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %add_nxv1i32 = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> undef) + %add_nxv1i32 = call i32 @llvm.vector.reduce.add.nxv1i32(<vscale x 1 x i32> poison) %add_nxv4i32 = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> %v0) %add_nxv4i64 = call i64 @llvm.vector.reduce.add.nxv4i64(<vscale x 4 x i64> %v1) - %mul_nxv1i32 = call i32 @llvm.vector.reduce.mul.nxv1i32(<vscale x 1 x i32> undef) + %mul_nxv1i32 = call i32 @llvm.vector.reduce.mul.nxv1i32(<vscale x 1 x i32> poison) %mul_nxv4i32 = call i32 @llvm.vector.reduce.mul.nxv4i32(<vscale x 4 x i32> %v0) %mul_nxv4i64 = call i64 @llvm.vector.reduce.mul.nxv4i64(<vscale x 4 x i64> %v1) - %and_nxv1i32 = call i32 @llvm.vector.reduce.and.nxv1i32(<vscale x 1 x i32> undef) + %and_nxv1i32 = call i32 @llvm.vector.reduce.and.nxv1i32(<vscale x 1 x i32> poison) %and_nxv4i32 = call i32 @llvm.vector.reduce.and.nxv4i32(<vscale x 4 x i32> %v0) %and_nxv4i64 = call i64 @llvm.vector.reduce.and.nxv4i64(<vscale x 4 x i64> %v1) - %or_nxv1i32 = call i32 @llvm.vector.reduce.or.nxv1i32(<vscale x 1 x i32> undef) + %or_nxv1i32 = call i32 @llvm.vector.reduce.or.nxv1i32(<vscale x 1 x i32> poison) %or_nxv4i32 = call i32 @llvm.vector.reduce.or.nxv4i32(<vscale x 4 x i32> %v0) %or_nxv4i64 = call i64 @llvm.vector.reduce.or.nxv4i64(<vscale x 4 x i64> %v1) - %xor_nxv1i32 = call i32 @llvm.vector.reduce.xor.nxv1i32(<vscale x 1 x i32> undef) + %xor_nxv1i32 = call i32 @llvm.vector.reduce.xor.nxv1i32(<vscale x 1 x i32> poison) %xor_nxv4i32 = call i32 @llvm.vector.reduce.xor.nxv4i32(<vscale x 4 x i32> %v0) %xor_nxv4i64 = call i64 @llvm.vector.reduce.xor.nxv4i64(<vscale x 4 x i64> %v1) - %umin_nxv1i64 = call i64 @llvm.vector.reduce.umin.nxv1i64(<vscale x 1 x i64> undef) + %umin_nxv1i64 = call i64 @llvm.vector.reduce.umin.nxv1i64(<vscale x 1 x i64> poison) %umin_nxv4i32 = call i32 @llvm.vector.reduce.umin.nxv4i32(<vscale x 4 x i32> %v0) %umin_nxv4i64 = call i64 @llvm.vector.reduce.umin.nxv4i64(<vscale x 4 x i64> %v1) - %smin_nxv1i64 = call i64 @llvm.vector.reduce.smin.nxv1i64(<vscale x 1 x i64> undef) + %smin_nxv1i64 = call i64 @llvm.vector.reduce.smin.nxv1i64(<vscale x 1 x i64> poison) %smin_nxv4i32 = call i32 @llvm.vector.reduce.smin.nxv4i32(<vscale x 4 x i32> %v0) %smin_nxv4i64 = call i64 @llvm.vector.reduce.smin.nxv4i64(<vscale x 4 x i64> %v1) - %umax_nxv1i64 = call i64 @llvm.vector.reduce.umax.nxv1i64(<vscale x 1 x i64> undef) + %umax_nxv1i64 = call i64 @llvm.vector.reduce.umax.nxv1i64(<vscale x 1 x i64> poison) %umax_nxv4i32 = call i32 @llvm.vector.reduce.umax.nxv4i32(<vscale x 4 x i32> %v0) %umax_nxv4i64 = call i64 @llvm.vector.reduce.umax.nxv4i64(<vscale x 4 x i64> %v1) - %smax_nxv1i64 = call i64 @llvm.vector.reduce.smax.nxv1i64(<vscale x 1 x i64> undef) + %smax_nxv1i64 = call i64 @llvm.vector.reduce.smax.nxv1i64(<vscale x 1 x i64> poison) %smax_nxv4i32 = call i32 @llvm.vector.reduce.smax.nxv4i32(<vscale x 4 x i32> %v0) %smax_nxv4i64 = call i64 @llvm.vector.reduce.smax.nxv4i64(<vscale x 4 x i64> %v1) - %fadd_nxv1f32 = call fast float @llvm.vector.reduce.fadd.nxv1f32(float 0.0, <vscale x 1 x float> undef) + %fadd_nxv1f32 = call fast float @llvm.vector.reduce.fadd.nxv1f32(float 0.0, <vscale x 1 x float> poison) %fadd_nxv4f32 = call fast float @llvm.vector.reduce.fadd.nxv4f32(float 0.0, <vscale x 4 x float> %v2) %fadd_nxv4f64 = call fast double @llvm.vector.reduce.fadd.nxv4f64(double 0.0, <vscale x 4 x double> %v3) - %fmin_nxv1f32 = call fast float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> undef) + %fmin_nxv1f32 = call fast float @llvm.vector.reduce.fmin.nxv1f32(<vscale x 1 x float> poison) %fmin_nxv4f32 = call fast float @llvm.vector.reduce.fmin.nxv4f32(<vscale x 4 x float> %v2) %fmin_nxv4f64 = call fast double @llvm.vector.reduce.fmin.nxv4f64(<vscale x 4 x double> %v3) - %fmax_nxv1f32 = call fast float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> undef) + %fmax_nxv1f32 = call fast float @llvm.vector.reduce.fmax.nxv1f32(<vscale x 1 x float> poison) %fmax_nxv4f32 = call fast float @llvm.vector.reduce.fmax.nxv4f32(<vscale x 4 x float> %v2) %fmax_nxv4f64 = call fast double @llvm.vector.reduce.fmax.nxv4f64(<vscale x 4 x double> %v3) @@ -389,123 +389,123 @@ declare <vscale x 4 x i32> @llvm.cttz.nxv4i32(<vscale x 4 x i32>, i1) define void @vector_reverse() #0 { ; CHECK-VSCALE-1-LABEL: 'vector_reverse' -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv16i8 = call <vscale x 16 x i8> @llvm.vector.reverse.nxv16i8(<vscale x 16 x i8> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv32i8 = call <vscale x 32 x i8> @llvm.vector.reverse.nxv32i8(<vscale x 32 x i8> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i16 = call <vscale x 2 x i16> @llvm.vector.reverse.nxv2i16(<vscale x 2 x i16> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i16 = call <vscale x 4 x i16> @llvm.vector.reverse.nxv4i16(<vscale x 4 x i16> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8i16 = call <vscale x 8 x i16> @llvm.vector.reverse.nxv8i16(<vscale x 8 x i16> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16i16 = call <vscale x 16 x i16> @llvm.vector.reverse.nxv16i16(<vscale x 16 x i16> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i32 = call <vscale x 4 x i32> @llvm.vector.reverse.nxv4i32(<vscale x 4 x i32> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv8i32 = call <vscale x 8 x i32> @llvm.vector.reverse.nxv8i32(<vscale x 8 x i32> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i64 = call <vscale x 2 x i64> @llvm.vector.reverse.nxv2i64(<vscale x 2 x i64> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv4i64 = call <vscale x 4 x i64> @llvm.vector.reverse.nxv4i64(<vscale x 4 x i64> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f16 = call <vscale x 2 x half> @llvm.vector.reverse.nxv2f16(<vscale x 2 x half> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4f16 = call <vscale x 4 x half> @llvm.vector.reverse.nxv4f16(<vscale x 4 x half> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8f16 = call <vscale x 8 x half> @llvm.vector.reverse.nxv8f16(<vscale x 8 x half> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16f16 = call <vscale x 16 x half> @llvm.vector.reverse.nxv16f16(<vscale x 16 x half> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f32 = call <vscale x 2 x float> @llvm.vector.reverse.nxv2f32(<vscale x 2 x float> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4f32 = call <vscale x 4 x float> @llvm.vector.reverse.nxv4f32(<vscale x 4 x float> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv8f32 = call <vscale x 8 x float> @llvm.vector.reverse.nxv8f32(<vscale x 8 x float> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f64 = call <vscale x 2 x double> @llvm.vector.reverse.nxv2f64(<vscale x 2 x double> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv4f64 = call <vscale x 4 x double> @llvm.vector.reverse.nxv4f64(<vscale x 4 x double> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2bf16 = call <vscale x 2 x bfloat> @llvm.vector.reverse.nxv2bf16(<vscale x 2 x bfloat> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4bf16 = call <vscale x 4 x bfloat> @llvm.vector.reverse.nxv4bf16(<vscale x 4 x bfloat> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8bf16 = call <vscale x 8 x bfloat> @llvm.vector.reverse.nxv8bf16(<vscale x 8 x bfloat> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16bf16 = call <vscale x 16 x bfloat> @llvm.vector.reverse.nxv16bf16(<vscale x 16 x bfloat> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv16i1 = call <vscale x 16 x i1> @llvm.vector.reverse.nxv16i1(<vscale x 16 x i1> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8i1 = call <vscale x 8 x i1> @llvm.vector.reverse.nxv8i1(<vscale x 8 x i1> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i1 = call <vscale x 4 x i1> @llvm.vector.reverse.nxv4i1(<vscale x 4 x i1> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i1 = call <vscale x 2 x i1> @llvm.vector.reverse.nxv2i1(<vscale x 2 x i1> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv16i8 = call <vscale x 16 x i8> @llvm.vector.reverse.nxv16i8(<vscale x 16 x i8> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv32i8 = call <vscale x 32 x i8> @llvm.vector.reverse.nxv32i8(<vscale x 32 x i8> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i16 = call <vscale x 2 x i16> @llvm.vector.reverse.nxv2i16(<vscale x 2 x i16> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i16 = call <vscale x 4 x i16> @llvm.vector.reverse.nxv4i16(<vscale x 4 x i16> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8i16 = call <vscale x 8 x i16> @llvm.vector.reverse.nxv8i16(<vscale x 8 x i16> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16i16 = call <vscale x 16 x i16> @llvm.vector.reverse.nxv16i16(<vscale x 16 x i16> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i32 = call <vscale x 4 x i32> @llvm.vector.reverse.nxv4i32(<vscale x 4 x i32> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv8i32 = call <vscale x 8 x i32> @llvm.vector.reverse.nxv8i32(<vscale x 8 x i32> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i64 = call <vscale x 2 x i64> @llvm.vector.reverse.nxv2i64(<vscale x 2 x i64> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv4i64 = call <vscale x 4 x i64> @llvm.vector.reverse.nxv4i64(<vscale x 4 x i64> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f16 = call <vscale x 2 x half> @llvm.vector.reverse.nxv2f16(<vscale x 2 x half> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4f16 = call <vscale x 4 x half> @llvm.vector.reverse.nxv4f16(<vscale x 4 x half> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8f16 = call <vscale x 8 x half> @llvm.vector.reverse.nxv8f16(<vscale x 8 x half> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16f16 = call <vscale x 16 x half> @llvm.vector.reverse.nxv16f16(<vscale x 16 x half> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f32 = call <vscale x 2 x float> @llvm.vector.reverse.nxv2f32(<vscale x 2 x float> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4f32 = call <vscale x 4 x float> @llvm.vector.reverse.nxv4f32(<vscale x 4 x float> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv8f32 = call <vscale x 8 x float> @llvm.vector.reverse.nxv8f32(<vscale x 8 x float> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f64 = call <vscale x 2 x double> @llvm.vector.reverse.nxv2f64(<vscale x 2 x double> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv4f64 = call <vscale x 4 x double> @llvm.vector.reverse.nxv4f64(<vscale x 4 x double> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2bf16 = call <vscale x 2 x bfloat> @llvm.vector.reverse.nxv2bf16(<vscale x 2 x bfloat> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4bf16 = call <vscale x 4 x bfloat> @llvm.vector.reverse.nxv4bf16(<vscale x 4 x bfloat> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8bf16 = call <vscale x 8 x bfloat> @llvm.vector.reverse.nxv8bf16(<vscale x 8 x bfloat> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16bf16 = call <vscale x 16 x bfloat> @llvm.vector.reverse.nxv16bf16(<vscale x 16 x bfloat> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv16i1 = call <vscale x 16 x i1> @llvm.vector.reverse.nxv16i1(<vscale x 16 x i1> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8i1 = call <vscale x 8 x i1> @llvm.vector.reverse.nxv8i1(<vscale x 8 x i1> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i1 = call <vscale x 4 x i1> @llvm.vector.reverse.nxv4i1(<vscale x 4 x i1> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i1 = call <vscale x 2 x i1> @llvm.vector.reverse.nxv2i1(<vscale x 2 x i1> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-VSCALE-2-LABEL: 'vector_reverse' -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv16i8 = call <vscale x 16 x i8> @llvm.vector.reverse.nxv16i8(<vscale x 16 x i8> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv32i8 = call <vscale x 32 x i8> @llvm.vector.reverse.nxv32i8(<vscale x 32 x i8> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i16 = call <vscale x 2 x i16> @llvm.vector.reverse.nxv2i16(<vscale x 2 x i16> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i16 = call <vscale x 4 x i16> @llvm.vector.reverse.nxv4i16(<vscale x 4 x i16> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8i16 = call <vscale x 8 x i16> @llvm.vector.reverse.nxv8i16(<vscale x 8 x i16> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16i16 = call <vscale x 16 x i16> @llvm.vector.reverse.nxv16i16(<vscale x 16 x i16> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i32 = call <vscale x 4 x i32> @llvm.vector.reverse.nxv4i32(<vscale x 4 x i32> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv8i32 = call <vscale x 8 x i32> @llvm.vector.reverse.nxv8i32(<vscale x 8 x i32> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i64 = call <vscale x 2 x i64> @llvm.vector.reverse.nxv2i64(<vscale x 2 x i64> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv4i64 = call <vscale x 4 x i64> @llvm.vector.reverse.nxv4i64(<vscale x 4 x i64> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f16 = call <vscale x 2 x half> @llvm.vector.reverse.nxv2f16(<vscale x 2 x half> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4f16 = call <vscale x 4 x half> @llvm.vector.reverse.nxv4f16(<vscale x 4 x half> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8f16 = call <vscale x 8 x half> @llvm.vector.reverse.nxv8f16(<vscale x 8 x half> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16f16 = call <vscale x 16 x half> @llvm.vector.reverse.nxv16f16(<vscale x 16 x half> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f32 = call <vscale x 2 x float> @llvm.vector.reverse.nxv2f32(<vscale x 2 x float> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4f32 = call <vscale x 4 x float> @llvm.vector.reverse.nxv4f32(<vscale x 4 x float> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv8f32 = call <vscale x 8 x float> @llvm.vector.reverse.nxv8f32(<vscale x 8 x float> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f64 = call <vscale x 2 x double> @llvm.vector.reverse.nxv2f64(<vscale x 2 x double> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv4f64 = call <vscale x 4 x double> @llvm.vector.reverse.nxv4f64(<vscale x 4 x double> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2bf16 = call <vscale x 2 x bfloat> @llvm.vector.reverse.nxv2bf16(<vscale x 2 x bfloat> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4bf16 = call <vscale x 4 x bfloat> @llvm.vector.reverse.nxv4bf16(<vscale x 4 x bfloat> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8bf16 = call <vscale x 8 x bfloat> @llvm.vector.reverse.nxv8bf16(<vscale x 8 x bfloat> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16bf16 = call <vscale x 16 x bfloat> @llvm.vector.reverse.nxv16bf16(<vscale x 16 x bfloat> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv16i1 = call <vscale x 16 x i1> @llvm.vector.reverse.nxv16i1(<vscale x 16 x i1> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8i1 = call <vscale x 8 x i1> @llvm.vector.reverse.nxv8i1(<vscale x 8 x i1> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i1 = call <vscale x 4 x i1> @llvm.vector.reverse.nxv4i1(<vscale x 4 x i1> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i1 = call <vscale x 2 x i1> @llvm.vector.reverse.nxv2i1(<vscale x 2 x i1> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv16i8 = call <vscale x 16 x i8> @llvm.vector.reverse.nxv16i8(<vscale x 16 x i8> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv32i8 = call <vscale x 32 x i8> @llvm.vector.reverse.nxv32i8(<vscale x 32 x i8> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i16 = call <vscale x 2 x i16> @llvm.vector.reverse.nxv2i16(<vscale x 2 x i16> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i16 = call <vscale x 4 x i16> @llvm.vector.reverse.nxv4i16(<vscale x 4 x i16> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8i16 = call <vscale x 8 x i16> @llvm.vector.reverse.nxv8i16(<vscale x 8 x i16> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16i16 = call <vscale x 16 x i16> @llvm.vector.reverse.nxv16i16(<vscale x 16 x i16> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i32 = call <vscale x 4 x i32> @llvm.vector.reverse.nxv4i32(<vscale x 4 x i32> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv8i32 = call <vscale x 8 x i32> @llvm.vector.reverse.nxv8i32(<vscale x 8 x i32> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i64 = call <vscale x 2 x i64> @llvm.vector.reverse.nxv2i64(<vscale x 2 x i64> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv4i64 = call <vscale x 4 x i64> @llvm.vector.reverse.nxv4i64(<vscale x 4 x i64> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f16 = call <vscale x 2 x half> @llvm.vector.reverse.nxv2f16(<vscale x 2 x half> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4f16 = call <vscale x 4 x half> @llvm.vector.reverse.nxv4f16(<vscale x 4 x half> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8f16 = call <vscale x 8 x half> @llvm.vector.reverse.nxv8f16(<vscale x 8 x half> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16f16 = call <vscale x 16 x half> @llvm.vector.reverse.nxv16f16(<vscale x 16 x half> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f32 = call <vscale x 2 x float> @llvm.vector.reverse.nxv2f32(<vscale x 2 x float> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4f32 = call <vscale x 4 x float> @llvm.vector.reverse.nxv4f32(<vscale x 4 x float> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv8f32 = call <vscale x 8 x float> @llvm.vector.reverse.nxv8f32(<vscale x 8 x float> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f64 = call <vscale x 2 x double> @llvm.vector.reverse.nxv2f64(<vscale x 2 x double> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv4f64 = call <vscale x 4 x double> @llvm.vector.reverse.nxv4f64(<vscale x 4 x double> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2bf16 = call <vscale x 2 x bfloat> @llvm.vector.reverse.nxv2bf16(<vscale x 2 x bfloat> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4bf16 = call <vscale x 4 x bfloat> @llvm.vector.reverse.nxv4bf16(<vscale x 4 x bfloat> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8bf16 = call <vscale x 8 x bfloat> @llvm.vector.reverse.nxv8bf16(<vscale x 8 x bfloat> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16bf16 = call <vscale x 16 x bfloat> @llvm.vector.reverse.nxv16bf16(<vscale x 16 x bfloat> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv16i1 = call <vscale x 16 x i1> @llvm.vector.reverse.nxv16i1(<vscale x 16 x i1> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8i1 = call <vscale x 8 x i1> @llvm.vector.reverse.nxv8i1(<vscale x 8 x i1> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i1 = call <vscale x 4 x i1> @llvm.vector.reverse.nxv4i1(<vscale x 4 x i1> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i1 = call <vscale x 2 x i1> @llvm.vector.reverse.nxv2i1(<vscale x 2 x i1> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; TYPE_BASED_ONLY-LABEL: 'vector_reverse' -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv16i8 = call <vscale x 16 x i8> @llvm.vector.reverse.nxv16i8(<vscale x 16 x i8> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv32i8 = call <vscale x 32 x i8> @llvm.vector.reverse.nxv32i8(<vscale x 32 x i8> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i16 = call <vscale x 2 x i16> @llvm.vector.reverse.nxv2i16(<vscale x 2 x i16> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i16 = call <vscale x 4 x i16> @llvm.vector.reverse.nxv4i16(<vscale x 4 x i16> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8i16 = call <vscale x 8 x i16> @llvm.vector.reverse.nxv8i16(<vscale x 8 x i16> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16i16 = call <vscale x 16 x i16> @llvm.vector.reverse.nxv16i16(<vscale x 16 x i16> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i32 = call <vscale x 4 x i32> @llvm.vector.reverse.nxv4i32(<vscale x 4 x i32> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv8i32 = call <vscale x 8 x i32> @llvm.vector.reverse.nxv8i32(<vscale x 8 x i32> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i64 = call <vscale x 2 x i64> @llvm.vector.reverse.nxv2i64(<vscale x 2 x i64> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv4i64 = call <vscale x 4 x i64> @llvm.vector.reverse.nxv4i64(<vscale x 4 x i64> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f16 = call <vscale x 2 x half> @llvm.vector.reverse.nxv2f16(<vscale x 2 x half> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4f16 = call <vscale x 4 x half> @llvm.vector.reverse.nxv4f16(<vscale x 4 x half> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8f16 = call <vscale x 8 x half> @llvm.vector.reverse.nxv8f16(<vscale x 8 x half> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16f16 = call <vscale x 16 x half> @llvm.vector.reverse.nxv16f16(<vscale x 16 x half> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f32 = call <vscale x 2 x float> @llvm.vector.reverse.nxv2f32(<vscale x 2 x float> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4f32 = call <vscale x 4 x float> @llvm.vector.reverse.nxv4f32(<vscale x 4 x float> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv8f32 = call <vscale x 8 x float> @llvm.vector.reverse.nxv8f32(<vscale x 8 x float> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f64 = call <vscale x 2 x double> @llvm.vector.reverse.nxv2f64(<vscale x 2 x double> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv4f64 = call <vscale x 4 x double> @llvm.vector.reverse.nxv4f64(<vscale x 4 x double> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2bf16 = call <vscale x 2 x bfloat> @llvm.vector.reverse.nxv2bf16(<vscale x 2 x bfloat> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4bf16 = call <vscale x 4 x bfloat> @llvm.vector.reverse.nxv4bf16(<vscale x 4 x bfloat> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8bf16 = call <vscale x 8 x bfloat> @llvm.vector.reverse.nxv8bf16(<vscale x 8 x bfloat> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16bf16 = call <vscale x 16 x bfloat> @llvm.vector.reverse.nxv16bf16(<vscale x 16 x bfloat> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv16i1 = call <vscale x 16 x i1> @llvm.vector.reverse.nxv16i1(<vscale x 16 x i1> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8i1 = call <vscale x 8 x i1> @llvm.vector.reverse.nxv8i1(<vscale x 8 x i1> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i1 = call <vscale x 4 x i1> @llvm.vector.reverse.nxv4i1(<vscale x 4 x i1> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i1 = call <vscale x 2 x i1> @llvm.vector.reverse.nxv2i1(<vscale x 2 x i1> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv16i8 = call <vscale x 16 x i8> @llvm.vector.reverse.nxv16i8(<vscale x 16 x i8> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv32i8 = call <vscale x 32 x i8> @llvm.vector.reverse.nxv32i8(<vscale x 32 x i8> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i16 = call <vscale x 2 x i16> @llvm.vector.reverse.nxv2i16(<vscale x 2 x i16> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i16 = call <vscale x 4 x i16> @llvm.vector.reverse.nxv4i16(<vscale x 4 x i16> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8i16 = call <vscale x 8 x i16> @llvm.vector.reverse.nxv8i16(<vscale x 8 x i16> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16i16 = call <vscale x 16 x i16> @llvm.vector.reverse.nxv16i16(<vscale x 16 x i16> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i32 = call <vscale x 4 x i32> @llvm.vector.reverse.nxv4i32(<vscale x 4 x i32> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv8i32 = call <vscale x 8 x i32> @llvm.vector.reverse.nxv8i32(<vscale x 8 x i32> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i64 = call <vscale x 2 x i64> @llvm.vector.reverse.nxv2i64(<vscale x 2 x i64> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv4i64 = call <vscale x 4 x i64> @llvm.vector.reverse.nxv4i64(<vscale x 4 x i64> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f16 = call <vscale x 2 x half> @llvm.vector.reverse.nxv2f16(<vscale x 2 x half> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4f16 = call <vscale x 4 x half> @llvm.vector.reverse.nxv4f16(<vscale x 4 x half> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8f16 = call <vscale x 8 x half> @llvm.vector.reverse.nxv8f16(<vscale x 8 x half> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16f16 = call <vscale x 16 x half> @llvm.vector.reverse.nxv16f16(<vscale x 16 x half> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f32 = call <vscale x 2 x float> @llvm.vector.reverse.nxv2f32(<vscale x 2 x float> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4f32 = call <vscale x 4 x float> @llvm.vector.reverse.nxv4f32(<vscale x 4 x float> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv8f32 = call <vscale x 8 x float> @llvm.vector.reverse.nxv8f32(<vscale x 8 x float> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2f64 = call <vscale x 2 x double> @llvm.vector.reverse.nxv2f64(<vscale x 2 x double> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv4f64 = call <vscale x 4 x double> @llvm.vector.reverse.nxv4f64(<vscale x 4 x double> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2bf16 = call <vscale x 2 x bfloat> @llvm.vector.reverse.nxv2bf16(<vscale x 2 x bfloat> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4bf16 = call <vscale x 4 x bfloat> @llvm.vector.reverse.nxv4bf16(<vscale x 4 x bfloat> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8bf16 = call <vscale x 8 x bfloat> @llvm.vector.reverse.nxv8bf16(<vscale x 8 x bfloat> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %reverse_nxv16bf16 = call <vscale x 16 x bfloat> @llvm.vector.reverse.nxv16bf16(<vscale x 16 x bfloat> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv16i1 = call <vscale x 16 x i1> @llvm.vector.reverse.nxv16i1(<vscale x 16 x i1> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv8i1 = call <vscale x 8 x i1> @llvm.vector.reverse.nxv8i1(<vscale x 8 x i1> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv4i1 = call <vscale x 4 x i1> @llvm.vector.reverse.nxv4i1(<vscale x 4 x i1> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %reverse_nxv2i1 = call <vscale x 2 x i1> @llvm.vector.reverse.nxv2i1(<vscale x 2 x i1> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %reverse_nxv16i8 = call <vscale x 16 x i8> @llvm.vector.reverse.nxv16i8(<vscale x 16 x i8> undef) - %reverse_nxv32i8 = call <vscale x 32 x i8> @llvm.vector.reverse.nxv32i8(<vscale x 32 x i8> undef) - %reverse_nxv2i16 = call <vscale x 2 x i16> @llvm.vector.reverse.nxv2i16(<vscale x 2 x i16> undef) - %reverse_nxv4i16 = call <vscale x 4 x i16> @llvm.vector.reverse.nxv4i16(<vscale x 4 x i16> undef) - %reverse_nxv8i16 = call <vscale x 8 x i16> @llvm.vector.reverse.nxv8i16(<vscale x 8 x i16> undef) - %reverse_nxv16i16 = call <vscale x 16 x i16> @llvm.vector.reverse.nxv16i16(<vscale x 16 x i16> undef) - %reverse_nxv4i32 = call <vscale x 4 x i32> @llvm.vector.reverse.nxv4i32(<vscale x 4 x i32> undef) - %reverse_nxv8i32 = call <vscale x 8 x i32> @llvm.vector.reverse.nxv8i32(<vscale x 8 x i32> undef) - %reverse_nxv2i64 = call <vscale x 2 x i64> @llvm.vector.reverse.nxv2i64(<vscale x 2 x i64> undef) - %reverse_nxv4i64 = call <vscale x 4 x i64> @llvm.vector.reverse.nxv4i64(<vscale x 4 x i64> undef) - %reverse_nxv2f16 = call <vscale x 2 x half> @llvm.vector.reverse.nxv2f16(<vscale x 2 x half> undef) - %reverse_nxv4f16 = call <vscale x 4 x half> @llvm.vector.reverse.nxv4f16(<vscale x 4 x half> undef) - %reverse_nxv8f16 = call <vscale x 8 x half> @llvm.vector.reverse.nxv8f16(<vscale x 8 x half> undef) - %reverse_nxv16f16 = call <vscale x 16 x half> @llvm.vector.reverse.nxv16f16(<vscale x 16 x half> undef) - %reverse_nxv2f32 = call <vscale x 2 x float> @llvm.vector.reverse.nxv2f32(<vscale x 2 x float> undef) - %reverse_nxv4f32 = call <vscale x 4 x float> @llvm.vector.reverse.nxv4f32(<vscale x 4 x float> undef) - %reverse_nxv8f32 = call <vscale x 8 x float> @llvm.vector.reverse.nxv8f32(<vscale x 8 x float> undef) - %reverse_nxv2f64 = call <vscale x 2 x double> @llvm.vector.reverse.nxv2f64(<vscale x 2 x double> undef) - %reverse_nxv4f64 = call <vscale x 4 x double> @llvm.vector.reverse.nxv4f64(<vscale x 4 x double> undef) - %reverse_nxv2bf16 = call <vscale x 2 x bfloat> @llvm.vector.reverse.nxv2bf16(<vscale x 2 x bfloat> undef) - %reverse_nxv4bf16 = call <vscale x 4 x bfloat> @llvm.vector.reverse.nxv4bf16(<vscale x 4 x bfloat> undef) - %reverse_nxv8bf16 = call <vscale x 8 x bfloat> @llvm.vector.reverse.nxv8bf16(<vscale x 8 x bfloat> undef) - %reverse_nxv16bf16 = call <vscale x 16 x bfloat> @llvm.vector.reverse.nxv16bf16(<vscale x 16 x bfloat> undef) - %reverse_nxv16i1 = call <vscale x 16 x i1> @llvm.vector.reverse.nxv16i1(<vscale x 16 x i1> undef) - %reverse_nxv8i1 = call <vscale x 8 x i1> @llvm.vector.reverse.nxv8i1(<vscale x 8 x i1> undef) - %reverse_nxv4i1 = call <vscale x 4 x i1> @llvm.vector.reverse.nxv4i1(<vscale x 4 x i1> undef) - %reverse_nxv2i1 = call <vscale x 2 x i1> @llvm.vector.reverse.nxv2i1(<vscale x 2 x i1> undef) + %reverse_nxv16i8 = call <vscale x 16 x i8> @llvm.vector.reverse.nxv16i8(<vscale x 16 x i8> poison) + %reverse_nxv32i8 = call <vscale x 32 x i8> @llvm.vector.reverse.nxv32i8(<vscale x 32 x i8> poison) + %reverse_nxv2i16 = call <vscale x 2 x i16> @llvm.vector.reverse.nxv2i16(<vscale x 2 x i16> poison) + %reverse_nxv4i16 = call <vscale x 4 x i16> @llvm.vector.reverse.nxv4i16(<vscale x 4 x i16> poison) + %reverse_nxv8i16 = call <vscale x 8 x i16> @llvm.vector.reverse.nxv8i16(<vscale x 8 x i16> poison) + %reverse_nxv16i16 = call <vscale x 16 x i16> @llvm.vector.reverse.nxv16i16(<vscale x 16 x i16> poison) + %reverse_nxv4i32 = call <vscale x 4 x i32> @llvm.vector.reverse.nxv4i32(<vscale x 4 x i32> poison) + %reverse_nxv8i32 = call <vscale x 8 x i32> @llvm.vector.reverse.nxv8i32(<vscale x 8 x i32> poison) + %reverse_nxv2i64 = call <vscale x 2 x i64> @llvm.vector.reverse.nxv2i64(<vscale x 2 x i64> poison) + %reverse_nxv4i64 = call <vscale x 4 x i64> @llvm.vector.reverse.nxv4i64(<vscale x 4 x i64> poison) + %reverse_nxv2f16 = call <vscale x 2 x half> @llvm.vector.reverse.nxv2f16(<vscale x 2 x half> poison) + %reverse_nxv4f16 = call <vscale x 4 x half> @llvm.vector.reverse.nxv4f16(<vscale x 4 x half> poison) + %reverse_nxv8f16 = call <vscale x 8 x half> @llvm.vector.reverse.nxv8f16(<vscale x 8 x half> poison) + %reverse_nxv16f16 = call <vscale x 16 x half> @llvm.vector.reverse.nxv16f16(<vscale x 16 x half> poison) + %reverse_nxv2f32 = call <vscale x 2 x float> @llvm.vector.reverse.nxv2f32(<vscale x 2 x float> poison) + %reverse_nxv4f32 = call <vscale x 4 x float> @llvm.vector.reverse.nxv4f32(<vscale x 4 x float> poison) + %reverse_nxv8f32 = call <vscale x 8 x float> @llvm.vector.reverse.nxv8f32(<vscale x 8 x float> poison) + %reverse_nxv2f64 = call <vscale x 2 x double> @llvm.vector.reverse.nxv2f64(<vscale x 2 x double> poison) + %reverse_nxv4f64 = call <vscale x 4 x double> @llvm.vector.reverse.nxv4f64(<vscale x 4 x double> poison) + %reverse_nxv2bf16 = call <vscale x 2 x bfloat> @llvm.vector.reverse.nxv2bf16(<vscale x 2 x bfloat> poison) + %reverse_nxv4bf16 = call <vscale x 4 x bfloat> @llvm.vector.reverse.nxv4bf16(<vscale x 4 x bfloat> poison) + %reverse_nxv8bf16 = call <vscale x 8 x bfloat> @llvm.vector.reverse.nxv8bf16(<vscale x 8 x bfloat> poison) + %reverse_nxv16bf16 = call <vscale x 16 x bfloat> @llvm.vector.reverse.nxv16bf16(<vscale x 16 x bfloat> poison) + %reverse_nxv16i1 = call <vscale x 16 x i1> @llvm.vector.reverse.nxv16i1(<vscale x 16 x i1> poison) + %reverse_nxv8i1 = call <vscale x 8 x i1> @llvm.vector.reverse.nxv8i1(<vscale x 8 x i1> poison) + %reverse_nxv4i1 = call <vscale x 4 x i1> @llvm.vector.reverse.nxv4i1(<vscale x 4 x i1> poison) + %reverse_nxv2i1 = call <vscale x 2 x i1> @llvm.vector.reverse.nxv2i1(<vscale x 2 x i1> poison) ret void } declare <vscale x 16 x i8> @llvm.vector.reverse.nxv16i8(<vscale x 16 x i8>) @@ -912,158 +912,158 @@ declare <vscale x 4 x double> @llvm.vector.splice.nxv4f64(<vscale x 4 x double>, define void @get_lane_mask() #0 { ; CHECK-VSCALE-1-LABEL: 'get_lane_mask' -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i64 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv8i1_i64 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv4i1_i64 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv2i1_i64 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i32 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv8i1_i32 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv4i1_i32 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv2i1_i32 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %mask_nxv32i1_i64 = call <vscale x 32 x i1> @llvm.get.active.lane.mask.nxv32i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i16 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i16(i16 undef, i16 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 32 for: %mask_v16i1_i64 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 16 for: %mask_v8i1_i64 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 8 for: %mask_v4i1_i64 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 4 for: %mask_v2i1_i64 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 32 for: %mask_v16i1_i32 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 16 for: %mask_v8i1_i32 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 8 for: %mask_v4i1_i32 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 4 for: %mask_v2i1_i32 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 48 for: %mask_v32i1_i64 = call <32 x i1> @llvm.get.active.lane.mask.v32i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 6 for: %mask_v16i1_i16 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i16(i16 undef, i16 undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i64 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv8i1_i64 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv4i1_i64 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv2i1_i64 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i32 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv8i1_i32 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv4i1_i32 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv2i1_i32 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 2 for: %mask_nxv32i1_i64 = call <vscale x 32 x i1> @llvm.get.active.lane.mask.nxv32i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i16 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i16(i16 poison, i16 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 32 for: %mask_v16i1_i64 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 16 for: %mask_v8i1_i64 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 8 for: %mask_v4i1_i64 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 4 for: %mask_v2i1_i64 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 32 for: %mask_v16i1_i32 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 16 for: %mask_v8i1_i32 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 8 for: %mask_v4i1_i32 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 4 for: %mask_v2i1_i32 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 48 for: %mask_v32i1_i64 = call <32 x i1> @llvm.get.active.lane.mask.v32i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 6 for: %mask_v16i1_i16 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i16(i16 poison, i16 poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-VSCALE-2-LABEL: 'get_lane_mask' -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i64 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv8i1_i64 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv4i1_i64 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv2i1_i64 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i32 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv8i1_i32 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv4i1_i32 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv2i1_i32 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %mask_nxv32i1_i64 = call <vscale x 32 x i1> @llvm.get.active.lane.mask.nxv32i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i16 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i16(i16 undef, i16 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 32 for: %mask_v16i1_i64 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 16 for: %mask_v8i1_i64 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 8 for: %mask_v4i1_i64 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 4 for: %mask_v2i1_i64 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 32 for: %mask_v16i1_i32 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 16 for: %mask_v8i1_i32 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 8 for: %mask_v4i1_i32 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 4 for: %mask_v2i1_i32 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 undef, i32 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 48 for: %mask_v32i1_i64 = call <32 x i1> @llvm.get.active.lane.mask.v32i1.i64(i64 undef, i64 undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 6 for: %mask_v16i1_i16 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i16(i16 undef, i16 undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i64 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv8i1_i64 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv4i1_i64 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv2i1_i64 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i32 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv8i1_i32 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv4i1_i32 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv2i1_i32 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 2 for: %mask_nxv32i1_i64 = call <vscale x 32 x i1> @llvm.get.active.lane.mask.nxv32i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i16 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i16(i16 poison, i16 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 32 for: %mask_v16i1_i64 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 16 for: %mask_v8i1_i64 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 8 for: %mask_v4i1_i64 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 4 for: %mask_v2i1_i64 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 32 for: %mask_v16i1_i32 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 16 for: %mask_v8i1_i32 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 8 for: %mask_v4i1_i32 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 4 for: %mask_v2i1_i32 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 poison, i32 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 48 for: %mask_v32i1_i64 = call <32 x i1> @llvm.get.active.lane.mask.v32i1.i64(i64 poison, i64 poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 6 for: %mask_v16i1_i16 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i16(i16 poison, i16 poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; TYPE_BASED_ONLY-LABEL: 'get_lane_mask' -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i64 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i64(i64 undef, i64 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv8i1_i64 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i64(i64 undef, i64 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv4i1_i64 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i64(i64 undef, i64 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv2i1_i64 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i64(i64 undef, i64 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i32 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i32(i32 undef, i32 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv8i1_i32 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i32(i32 undef, i32 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv4i1_i32 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 undef, i32 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv2i1_i32 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i32(i32 undef, i32 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %mask_nxv32i1_i64 = call <vscale x 32 x i1> @llvm.get.active.lane.mask.nxv32i1.i64(i64 undef, i64 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i16 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i16(i16 undef, i16 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 32 for: %mask_v16i1_i64 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i64(i64 undef, i64 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 16 for: %mask_v8i1_i64 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i64(i64 undef, i64 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 8 for: %mask_v4i1_i64 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i64(i64 undef, i64 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 4 for: %mask_v2i1_i64 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i64(i64 undef, i64 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 32 for: %mask_v16i1_i32 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i32(i32 undef, i32 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 16 for: %mask_v8i1_i32 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i32(i32 undef, i32 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 8 for: %mask_v4i1_i32 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 undef, i32 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 4 for: %mask_v2i1_i32 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 undef, i32 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 48 for: %mask_v32i1_i64 = call <32 x i1> @llvm.get.active.lane.mask.v32i1.i64(i64 undef, i64 undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 6 for: %mask_v16i1_i16 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i16(i16 undef, i16 undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i64 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i64(i64 poison, i64 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv8i1_i64 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i64(i64 poison, i64 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv4i1_i64 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i64(i64 poison, i64 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv2i1_i64 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i64(i64 poison, i64 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i32 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i32(i32 poison, i32 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv8i1_i32 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i32(i32 poison, i32 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv4i1_i32 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 poison, i32 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv2i1_i32 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i32(i32 poison, i32 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 2 for: %mask_nxv32i1_i64 = call <vscale x 32 x i1> @llvm.get.active.lane.mask.nxv32i1.i64(i64 poison, i64 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 1 for: %mask_nxv16i1_i16 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i16(i16 poison, i16 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 32 for: %mask_v16i1_i64 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i64(i64 poison, i64 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 16 for: %mask_v8i1_i64 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i64(i64 poison, i64 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 8 for: %mask_v4i1_i64 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i64(i64 poison, i64 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 4 for: %mask_v2i1_i64 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i64(i64 poison, i64 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 32 for: %mask_v16i1_i32 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i32(i32 poison, i32 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 16 for: %mask_v8i1_i32 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i32(i32 poison, i32 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 8 for: %mask_v4i1_i32 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 poison, i32 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 4 for: %mask_v2i1_i32 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 poison, i32 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 48 for: %mask_v32i1_i64 = call <32 x i1> @llvm.get.active.lane.mask.v32i1.i64(i64 poison, i64 poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 6 for: %mask_v16i1_i16 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i16(i16 poison, i16 poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %mask_nxv16i1_i64 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i64(i64 undef, i64 undef) - %mask_nxv8i1_i64 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i64(i64 undef, i64 undef) - %mask_nxv4i1_i64 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i64(i64 undef, i64 undef) - %mask_nxv2i1_i64 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i64(i64 undef, i64 undef) + %mask_nxv16i1_i64 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i64(i64 poison, i64 poison) + %mask_nxv8i1_i64 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i64(i64 poison, i64 poison) + %mask_nxv4i1_i64 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i64(i64 poison, i64 poison) + %mask_nxv2i1_i64 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i64(i64 poison, i64 poison) - %mask_nxv16i1_i32 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i32(i32 undef, i32 undef) - %mask_nxv8i1_i32 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i32(i32 undef, i32 undef) - %mask_nxv4i1_i32 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 undef, i32 undef) - %mask_nxv2i1_i32 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i32(i32 undef, i32 undef) + %mask_nxv16i1_i32 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i32(i32 poison, i32 poison) + %mask_nxv8i1_i32 = call <vscale x 8 x i1> @llvm.get.active.lane.mask.nxv8i1.i32(i32 poison, i32 poison) + %mask_nxv4i1_i32 = call <vscale x 4 x i1> @llvm.get.active.lane.mask.nxv4i1.i32(i32 poison, i32 poison) + %mask_nxv2i1_i32 = call <vscale x 2 x i1> @llvm.get.active.lane.mask.nxv2i1.i32(i32 poison, i32 poison) - %mask_nxv32i1_i64 = call <vscale x 32 x i1> @llvm.get.active.lane.mask.nxv32i1.i64(i64 undef, i64 undef) - %mask_nxv16i1_i16 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i16(i16 undef, i16 undef) + %mask_nxv32i1_i64 = call <vscale x 32 x i1> @llvm.get.active.lane.mask.nxv32i1.i64(i64 poison, i64 poison) + %mask_nxv16i1_i16 = call <vscale x 16 x i1> @llvm.get.active.lane.mask.nxv16i1.i16(i16 poison, i16 poison) - %mask_v16i1_i64 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i64(i64 undef, i64 undef) - %mask_v8i1_i64 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i64(i64 undef, i64 undef) - %mask_v4i1_i64 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i64(i64 undef, i64 undef) - %mask_v2i1_i64 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i64(i64 undef, i64 undef) + %mask_v16i1_i64 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i64(i64 poison, i64 poison) + %mask_v8i1_i64 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i64(i64 poison, i64 poison) + %mask_v4i1_i64 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i64(i64 poison, i64 poison) + %mask_v2i1_i64 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i64(i64 poison, i64 poison) - %mask_v16i1_i32 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i32(i32 undef, i32 undef) - %mask_v8i1_i32 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i32(i32 undef, i32 undef) - %mask_v4i1_i32 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 undef, i32 undef) - %mask_v2i1_i32 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 undef, i32 undef) + %mask_v16i1_i32 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i32(i32 poison, i32 poison) + %mask_v8i1_i32 = call <8 x i1> @llvm.get.active.lane.mask.v8i1.i32(i32 poison, i32 poison) + %mask_v4i1_i32 = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 poison, i32 poison) + %mask_v2i1_i32 = call <2 x i1> @llvm.get.active.lane.mask.v2i1.i32(i32 poison, i32 poison) - %mask_v32i1_i64 = call <32 x i1> @llvm.get.active.lane.mask.v32i1.i64(i64 undef, i64 undef) - %mask_v16i1_i16 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i16(i16 undef, i16 undef) + %mask_v32i1_i64 = call <32 x i1> @llvm.get.active.lane.mask.v32i1.i64(i64 poison, i64 poison) + %mask_v16i1_i16 = call <16 x i1> @llvm.get.active.lane.mask.v16i1.i16(i16 poison, i16 poison) ret void } define void @fshr() #0 { ; CHECK-VSCALE-1-LABEL: 'fshr' -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %1 = call <vscale x 16 x i8> @llvm.fshr.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i8> undef, <vscale x 16 x i8> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %2 = call <vscale x 8 x i16> @llvm.fshr.nxv8i16(<vscale x 8 x i16> undef, <vscale x 8 x i16> undef, <vscale x 8 x i16> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %3 = call <vscale x 4 x i32> @llvm.fshr.nxv4i32(<vscale x 4 x i32> undef, <vscale x 4 x i32> undef, <vscale x 4 x i32> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %4 = call <vscale x 2 x i64> @llvm.fshr.nxv2i64(<vscale x 2 x i64> undef, <vscale x 2 x i64> undef, <vscale x 2 x i64> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %1 = call <vscale x 16 x i8> @llvm.fshr.nxv16i8(<vscale x 16 x i8> poison, <vscale x 16 x i8> poison, <vscale x 16 x i8> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %2 = call <vscale x 8 x i16> @llvm.fshr.nxv8i16(<vscale x 8 x i16> poison, <vscale x 8 x i16> poison, <vscale x 8 x i16> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %3 = call <vscale x 4 x i32> @llvm.fshr.nxv4i32(<vscale x 4 x i32> poison, <vscale x 4 x i32> poison, <vscale x 4 x i32> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %4 = call <vscale x 2 x i64> @llvm.fshr.nxv2i64(<vscale x 2 x i64> poison, <vscale x 2 x i64> poison, <vscale x 2 x i64> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-VSCALE-2-LABEL: 'fshr' -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %1 = call <vscale x 16 x i8> @llvm.fshr.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i8> undef, <vscale x 16 x i8> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %2 = call <vscale x 8 x i16> @llvm.fshr.nxv8i16(<vscale x 8 x i16> undef, <vscale x 8 x i16> undef, <vscale x 8 x i16> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %3 = call <vscale x 4 x i32> @llvm.fshr.nxv4i32(<vscale x 4 x i32> undef, <vscale x 4 x i32> undef, <vscale x 4 x i32> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %4 = call <vscale x 2 x i64> @llvm.fshr.nxv2i64(<vscale x 2 x i64> undef, <vscale x 2 x i64> undef, <vscale x 2 x i64> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %1 = call <vscale x 16 x i8> @llvm.fshr.nxv16i8(<vscale x 16 x i8> poison, <vscale x 16 x i8> poison, <vscale x 16 x i8> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %2 = call <vscale x 8 x i16> @llvm.fshr.nxv8i16(<vscale x 8 x i16> poison, <vscale x 8 x i16> poison, <vscale x 8 x i16> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %3 = call <vscale x 4 x i32> @llvm.fshr.nxv4i32(<vscale x 4 x i32> poison, <vscale x 4 x i32> poison, <vscale x 4 x i32> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %4 = call <vscale x 2 x i64> @llvm.fshr.nxv2i64(<vscale x 2 x i64> poison, <vscale x 2 x i64> poison, <vscale x 2 x i64> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; TYPE_BASED_ONLY-LABEL: 'fshr' -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %1 = call <vscale x 16 x i8> @llvm.fshr.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i8> undef, <vscale x 16 x i8> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %2 = call <vscale x 8 x i16> @llvm.fshr.nxv8i16(<vscale x 8 x i16> undef, <vscale x 8 x i16> undef, <vscale x 8 x i16> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %3 = call <vscale x 4 x i32> @llvm.fshr.nxv4i32(<vscale x 4 x i32> undef, <vscale x 4 x i32> undef, <vscale x 4 x i32> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %4 = call <vscale x 2 x i64> @llvm.fshr.nxv2i64(<vscale x 2 x i64> undef, <vscale x 2 x i64> undef, <vscale x 2 x i64> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %1 = call <vscale x 16 x i8> @llvm.fshr.nxv16i8(<vscale x 16 x i8> poison, <vscale x 16 x i8> poison, <vscale x 16 x i8> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %2 = call <vscale x 8 x i16> @llvm.fshr.nxv8i16(<vscale x 8 x i16> poison, <vscale x 8 x i16> poison, <vscale x 8 x i16> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %3 = call <vscale x 4 x i32> @llvm.fshr.nxv4i32(<vscale x 4 x i32> poison, <vscale x 4 x i32> poison, <vscale x 4 x i32> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %4 = call <vscale x 2 x i64> @llvm.fshr.nxv2i64(<vscale x 2 x i64> poison, <vscale x 2 x i64> poison, <vscale x 2 x i64> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - call <vscale x 16 x i8> @llvm.fshr.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i8> undef, <vscale x 16 x i8> undef) - call <vscale x 8 x i16> @llvm.fshr.nxv8i16(<vscale x 8 x i16> undef, <vscale x 8 x i16> undef, <vscale x 8 x i16> undef) - call <vscale x 4 x i32> @llvm.fshr.nxv4i32(<vscale x 4 x i32> undef, <vscale x 4 x i32> undef, <vscale x 4 x i32> undef) - call <vscale x 2 x i64> @llvm.fshr.nxv2i64(<vscale x 2 x i64> undef, <vscale x 2 x i64> undef, <vscale x 2 x i64> undef) + call <vscale x 16 x i8> @llvm.fshr.nxv16i8(<vscale x 16 x i8> poison, <vscale x 16 x i8> poison, <vscale x 16 x i8> poison) + call <vscale x 8 x i16> @llvm.fshr.nxv8i16(<vscale x 8 x i16> poison, <vscale x 8 x i16> poison, <vscale x 8 x i16> poison) + call <vscale x 4 x i32> @llvm.fshr.nxv4i32(<vscale x 4 x i32> poison, <vscale x 4 x i32> poison, <vscale x 4 x i32> poison) + call <vscale x 2 x i64> @llvm.fshr.nxv2i64(<vscale x 2 x i64> poison, <vscale x 2 x i64> poison, <vscale x 2 x i64> poison) ret void } define void @fshl() #0 { ; CHECK-VSCALE-1-LABEL: 'fshl' -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %1 = call <vscale x 16 x i8> @llvm.fshl.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i8> undef, <vscale x 16 x i8> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %2 = call <vscale x 8 x i16> @llvm.fshl.nxv8i16(<vscale x 8 x i16> undef, <vscale x 8 x i16> undef, <vscale x 8 x i16> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %3 = call <vscale x 4 x i32> @llvm.fshl.nxv4i32(<vscale x 4 x i32> undef, <vscale x 4 x i32> undef, <vscale x 4 x i32> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %4 = call <vscale x 2 x i64> @llvm.fshl.nxv2i64(<vscale x 2 x i64> undef, <vscale x 2 x i64> undef, <vscale x 2 x i64> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %1 = call <vscale x 16 x i8> @llvm.fshl.nxv16i8(<vscale x 16 x i8> poison, <vscale x 16 x i8> poison, <vscale x 16 x i8> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %2 = call <vscale x 8 x i16> @llvm.fshl.nxv8i16(<vscale x 8 x i16> poison, <vscale x 8 x i16> poison, <vscale x 8 x i16> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %3 = call <vscale x 4 x i32> @llvm.fshl.nxv4i32(<vscale x 4 x i32> poison, <vscale x 4 x i32> poison, <vscale x 4 x i32> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 5 for: %4 = call <vscale x 2 x i64> @llvm.fshl.nxv2i64(<vscale x 2 x i64> poison, <vscale x 2 x i64> poison, <vscale x 2 x i64> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-VSCALE-2-LABEL: 'fshl' -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %1 = call <vscale x 16 x i8> @llvm.fshl.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i8> undef, <vscale x 16 x i8> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %2 = call <vscale x 8 x i16> @llvm.fshl.nxv8i16(<vscale x 8 x i16> undef, <vscale x 8 x i16> undef, <vscale x 8 x i16> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %3 = call <vscale x 4 x i32> @llvm.fshl.nxv4i32(<vscale x 4 x i32> undef, <vscale x 4 x i32> undef, <vscale x 4 x i32> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %4 = call <vscale x 2 x i64> @llvm.fshl.nxv2i64(<vscale x 2 x i64> undef, <vscale x 2 x i64> undef, <vscale x 2 x i64> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %1 = call <vscale x 16 x i8> @llvm.fshl.nxv16i8(<vscale x 16 x i8> poison, <vscale x 16 x i8> poison, <vscale x 16 x i8> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %2 = call <vscale x 8 x i16> @llvm.fshl.nxv8i16(<vscale x 8 x i16> poison, <vscale x 8 x i16> poison, <vscale x 8 x i16> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %3 = call <vscale x 4 x i32> @llvm.fshl.nxv4i32(<vscale x 4 x i32> poison, <vscale x 4 x i32> poison, <vscale x 4 x i32> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 5 for: %4 = call <vscale x 2 x i64> @llvm.fshl.nxv2i64(<vscale x 2 x i64> poison, <vscale x 2 x i64> poison, <vscale x 2 x i64> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; TYPE_BASED_ONLY-LABEL: 'fshl' -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %1 = call <vscale x 16 x i8> @llvm.fshl.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i8> undef, <vscale x 16 x i8> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %2 = call <vscale x 8 x i16> @llvm.fshl.nxv8i16(<vscale x 8 x i16> undef, <vscale x 8 x i16> undef, <vscale x 8 x i16> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %3 = call <vscale x 4 x i32> @llvm.fshl.nxv4i32(<vscale x 4 x i32> undef, <vscale x 4 x i32> undef, <vscale x 4 x i32> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %4 = call <vscale x 2 x i64> @llvm.fshl.nxv2i64(<vscale x 2 x i64> undef, <vscale x 2 x i64> undef, <vscale x 2 x i64> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %1 = call <vscale x 16 x i8> @llvm.fshl.nxv16i8(<vscale x 16 x i8> poison, <vscale x 16 x i8> poison, <vscale x 16 x i8> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %2 = call <vscale x 8 x i16> @llvm.fshl.nxv8i16(<vscale x 8 x i16> poison, <vscale x 8 x i16> poison, <vscale x 8 x i16> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %3 = call <vscale x 4 x i32> @llvm.fshl.nxv4i32(<vscale x 4 x i32> poison, <vscale x 4 x i32> poison, <vscale x 4 x i32> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 7 for: %4 = call <vscale x 2 x i64> @llvm.fshl.nxv2i64(<vscale x 2 x i64> poison, <vscale x 2 x i64> poison, <vscale x 2 x i64> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - call <vscale x 16 x i8> @llvm.fshl.nxv16i8(<vscale x 16 x i8> undef, <vscale x 16 x i8> undef, <vscale x 16 x i8> undef) - call <vscale x 8 x i16> @llvm.fshl.nxv8i16(<vscale x 8 x i16> undef, <vscale x 8 x i16> undef, <vscale x 8 x i16> undef) - call <vscale x 4 x i32> @llvm.fshl.nxv4i32(<vscale x 4 x i32> undef, <vscale x 4 x i32> undef, <vscale x 4 x i32> undef) - call <vscale x 2 x i64> @llvm.fshl.nxv2i64(<vscale x 2 x i64> undef, <vscale x 2 x i64> undef, <vscale x 2 x i64> undef) + call <vscale x 16 x i8> @llvm.fshl.nxv16i8(<vscale x 16 x i8> poison, <vscale x 16 x i8> poison, <vscale x 16 x i8> poison) + call <vscale x 8 x i16> @llvm.fshl.nxv8i16(<vscale x 8 x i16> poison, <vscale x 8 x i16> poison, <vscale x 8 x i16> poison) + call <vscale x 4 x i32> @llvm.fshl.nxv4i32(<vscale x 4 x i32> poison, <vscale x 4 x i32> poison, <vscale x 4 x i32> poison) + call <vscale x 2 x i64> @llvm.fshl.nxv2i64(<vscale x 2 x i64> poison, <vscale x 2 x i64> poison, <vscale x 2 x i64> poison) ret void } @@ -1362,48 +1362,48 @@ define void @histogram_nxv4i64(<vscale x 4 x ptr> %buckets, <vscale x 4 x i1> %m define void @match() #3 { ; CHECK-VSCALE-1-LABEL: 'match' -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 4 for: %match_nxv16i8_v16i8 = call <vscale x 16 x i1> @llvm.experimental.vector.match.nxv16i8.v16i8(<vscale x 16 x i8> undef, <16 x i8> undef, <vscale x 16 x i1> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 4 for: %match_nxv8i16_v8i16 = call <vscale x 8 x i1> @llvm.experimental.vector.match.nxv8i16.v8i16(<vscale x 8 x i16> undef, <8 x i16> undef, <vscale x 8 x i1> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:29 CodeSize:21 Lat:29 SizeLat:29 for: %match_nxv4i32_v4i32 = call <vscale x 4 x i1> @llvm.experimental.vector.match.nxv4i32.v4i32(<vscale x 4 x i32> undef, <4 x i32> undef, <vscale x 4 x i1> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:15 CodeSize:11 Lat:15 SizeLat:15 for: %match_nxv2i64_v2i64 = call <vscale x 2 x i1> @llvm.experimental.vector.match.nxv2i64.v2i64(<vscale x 2 x i64> undef, <2 x i64> undef, <vscale x 2 x i1> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 14 for: %match_v16i8_v16i8 = call <16 x i1> @llvm.experimental.vector.match.v16i8.v16i8(<16 x i8> undef, <16 x i8> undef, <16 x i1> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 14 for: %match_v8i16_v8i16 = call <8 x i1> @llvm.experimental.vector.match.v8i16.v8i16(<8 x i16> undef, <8 x i16> undef, <8 x i1> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:29 CodeSize:21 Lat:29 SizeLat:29 for: %match_v4i32_v4i32 = call <4 x i1> @llvm.experimental.vector.match.v4i32.v4i32(<4 x i32> undef, <4 x i32> undef, <4 x i1> undef) -; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:15 CodeSize:11 Lat:15 SizeLat:15 for: %match_v2i64_v2i64 = call <2 x i1> @llvm.experimental.vector.match.v2i64.v2i64(<2 x i64> undef, <2 x i64> undef, <2 x i1> undef) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 4 for: %match_nxv16i8_v16i8 = call <vscale x 16 x i1> @llvm.experimental.vector.match.nxv16i8.v16i8(<vscale x 16 x i8> poison, <16 x i8> poison, <vscale x 16 x i1> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 4 for: %match_nxv8i16_v8i16 = call <vscale x 8 x i1> @llvm.experimental.vector.match.nxv8i16.v8i16(<vscale x 8 x i16> poison, <8 x i16> poison, <vscale x 8 x i1> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:29 CodeSize:21 Lat:29 SizeLat:29 for: %match_nxv4i32_v4i32 = call <vscale x 4 x i1> @llvm.experimental.vector.match.nxv4i32.v4i32(<vscale x 4 x i32> poison, <4 x i32> poison, <vscale x 4 x i1> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:15 CodeSize:11 Lat:15 SizeLat:15 for: %match_nxv2i64_v2i64 = call <vscale x 2 x i1> @llvm.experimental.vector.match.nxv2i64.v2i64(<vscale x 2 x i64> poison, <2 x i64> poison, <vscale x 2 x i1> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 14 for: %match_v16i8_v16i8 = call <16 x i1> @llvm.experimental.vector.match.v16i8.v16i8(<16 x i8> poison, <16 x i8> poison, <16 x i1> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of 14 for: %match_v8i16_v8i16 = call <8 x i1> @llvm.experimental.vector.match.v8i16.v8i16(<8 x i16> poison, <8 x i16> poison, <8 x i1> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:29 CodeSize:21 Lat:29 SizeLat:29 for: %match_v4i32_v4i32 = call <4 x i1> @llvm.experimental.vector.match.v4i32.v4i32(<4 x i32> poison, <4 x i32> poison, <4 x i1> poison) +; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:15 CodeSize:11 Lat:15 SizeLat:15 for: %match_v2i64_v2i64 = call <2 x i1> @llvm.experimental.vector.match.v2i64.v2i64(<2 x i64> poison, <2 x i64> poison, <2 x i1> poison) ; CHECK-VSCALE-1-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; CHECK-VSCALE-2-LABEL: 'match' -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 4 for: %match_nxv16i8_v16i8 = call <vscale x 16 x i1> @llvm.experimental.vector.match.nxv16i8.v16i8(<vscale x 16 x i8> undef, <16 x i8> undef, <vscale x 16 x i1> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 4 for: %match_nxv8i16_v8i16 = call <vscale x 8 x i1> @llvm.experimental.vector.match.nxv8i16.v8i16(<vscale x 8 x i16> undef, <8 x i16> undef, <vscale x 8 x i1> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:29 CodeSize:21 Lat:29 SizeLat:29 for: %match_nxv4i32_v4i32 = call <vscale x 4 x i1> @llvm.experimental.vector.match.nxv4i32.v4i32(<vscale x 4 x i32> undef, <4 x i32> undef, <vscale x 4 x i1> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:15 CodeSize:11 Lat:15 SizeLat:15 for: %match_nxv2i64_v2i64 = call <vscale x 2 x i1> @llvm.experimental.vector.match.nxv2i64.v2i64(<vscale x 2 x i64> undef, <2 x i64> undef, <vscale x 2 x i1> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 14 for: %match_v16i8_v16i8 = call <16 x i1> @llvm.experimental.vector.match.v16i8.v16i8(<16 x i8> undef, <16 x i8> undef, <16 x i1> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 14 for: %match_v8i16_v8i16 = call <8 x i1> @llvm.experimental.vector.match.v8i16.v8i16(<8 x i16> undef, <8 x i16> undef, <8 x i1> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:29 CodeSize:21 Lat:29 SizeLat:29 for: %match_v4i32_v4i32 = call <4 x i1> @llvm.experimental.vector.match.v4i32.v4i32(<4 x i32> undef, <4 x i32> undef, <4 x i1> undef) -; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:15 CodeSize:11 Lat:15 SizeLat:15 for: %match_v2i64_v2i64 = call <2 x i1> @llvm.experimental.vector.match.v2i64.v2i64(<2 x i64> undef, <2 x i64> undef, <2 x i1> undef) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 4 for: %match_nxv16i8_v16i8 = call <vscale x 16 x i1> @llvm.experimental.vector.match.nxv16i8.v16i8(<vscale x 16 x i8> poison, <16 x i8> poison, <vscale x 16 x i1> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 4 for: %match_nxv8i16_v8i16 = call <vscale x 8 x i1> @llvm.experimental.vector.match.nxv8i16.v8i16(<vscale x 8 x i16> poison, <8 x i16> poison, <vscale x 8 x i1> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:29 CodeSize:21 Lat:29 SizeLat:29 for: %match_nxv4i32_v4i32 = call <vscale x 4 x i1> @llvm.experimental.vector.match.nxv4i32.v4i32(<vscale x 4 x i32> poison, <4 x i32> poison, <vscale x 4 x i1> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:15 CodeSize:11 Lat:15 SizeLat:15 for: %match_nxv2i64_v2i64 = call <vscale x 2 x i1> @llvm.experimental.vector.match.nxv2i64.v2i64(<vscale x 2 x i64> poison, <2 x i64> poison, <vscale x 2 x i1> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 14 for: %match_v16i8_v16i8 = call <16 x i1> @llvm.experimental.vector.match.v16i8.v16i8(<16 x i8> poison, <16 x i8> poison, <16 x i1> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of 14 for: %match_v8i16_v8i16 = call <8 x i1> @llvm.experimental.vector.match.v8i16.v8i16(<8 x i16> poison, <8 x i16> poison, <8 x i1> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:29 CodeSize:21 Lat:29 SizeLat:29 for: %match_v4i32_v4i32 = call <4 x i1> @llvm.experimental.vector.match.v4i32.v4i32(<4 x i32> poison, <4 x i32> poison, <4 x i1> poison) +; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:15 CodeSize:11 Lat:15 SizeLat:15 for: %match_v2i64_v2i64 = call <2 x i1> @llvm.experimental.vector.match.v2i64.v2i64(<2 x i64> poison, <2 x i64> poison, <2 x i1> poison) ; CHECK-VSCALE-2-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; ; TYPE_BASED_ONLY-LABEL: 'match' -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 4 for: %match_nxv16i8_v16i8 = call <vscale x 16 x i1> @llvm.experimental.vector.match.nxv16i8.v16i8(<vscale x 16 x i8> undef, <16 x i8> undef, <vscale x 16 x i1> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 4 for: %match_nxv8i16_v8i16 = call <vscale x 8 x i1> @llvm.experimental.vector.match.nxv8i16.v8i16(<vscale x 8 x i16> undef, <8 x i16> undef, <vscale x 8 x i1> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:29 CodeSize:21 Lat:29 SizeLat:29 for: %match_nxv4i32_v4i32 = call <vscale x 4 x i1> @llvm.experimental.vector.match.nxv4i32.v4i32(<vscale x 4 x i32> undef, <4 x i32> undef, <vscale x 4 x i1> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:15 CodeSize:11 Lat:15 SizeLat:15 for: %match_nxv2i64_v2i64 = call <vscale x 2 x i1> @llvm.experimental.vector.match.nxv2i64.v2i64(<vscale x 2 x i64> undef, <2 x i64> undef, <vscale x 2 x i1> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 14 for: %match_v16i8_v16i8 = call <16 x i1> @llvm.experimental.vector.match.v16i8.v16i8(<16 x i8> undef, <16 x i8> undef, <16 x i1> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 14 for: %match_v8i16_v8i16 = call <8 x i1> @llvm.experimental.vector.match.v8i16.v8i16(<8 x i16> undef, <8 x i16> undef, <8 x i1> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:29 CodeSize:21 Lat:29 SizeLat:29 for: %match_v4i32_v4i32 = call <4 x i1> @llvm.experimental.vector.match.v4i32.v4i32(<4 x i32> undef, <4 x i32> undef, <4 x i1> undef) -; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:15 CodeSize:11 Lat:15 SizeLat:15 for: %match_v2i64_v2i64 = call <2 x i1> @llvm.experimental.vector.match.v2i64.v2i64(<2 x i64> undef, <2 x i64> undef, <2 x i1> undef) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 4 for: %match_nxv16i8_v16i8 = call <vscale x 16 x i1> @llvm.experimental.vector.match.nxv16i8.v16i8(<vscale x 16 x i8> poison, <16 x i8> poison, <vscale x 16 x i1> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 4 for: %match_nxv8i16_v8i16 = call <vscale x 8 x i1> @llvm.experimental.vector.match.nxv8i16.v8i16(<vscale x 8 x i16> poison, <8 x i16> poison, <vscale x 8 x i1> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:29 CodeSize:21 Lat:29 SizeLat:29 for: %match_nxv4i32_v4i32 = call <vscale x 4 x i1> @llvm.experimental.vector.match.nxv4i32.v4i32(<vscale x 4 x i32> poison, <4 x i32> poison, <vscale x 4 x i1> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:15 CodeSize:11 Lat:15 SizeLat:15 for: %match_nxv2i64_v2i64 = call <vscale x 2 x i1> @llvm.experimental.vector.match.nxv2i64.v2i64(<vscale x 2 x i64> poison, <2 x i64> poison, <vscale x 2 x i1> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 14 for: %match_v16i8_v16i8 = call <16 x i1> @llvm.experimental.vector.match.v16i8.v16i8(<16 x i8> poison, <16 x i8> poison, <16 x i1> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of 14 for: %match_v8i16_v8i16 = call <8 x i1> @llvm.experimental.vector.match.v8i16.v8i16(<8 x i16> poison, <8 x i16> poison, <8 x i1> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:29 CodeSize:21 Lat:29 SizeLat:29 for: %match_v4i32_v4i32 = call <4 x i1> @llvm.experimental.vector.match.v4i32.v4i32(<4 x i32> poison, <4 x i32> poison, <4 x i1> poison) +; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:15 CodeSize:11 Lat:15 SizeLat:15 for: %match_v2i64_v2i64 = call <2 x i1> @llvm.experimental.vector.match.v2i64.v2i64(<2 x i64> poison, <2 x i64> poison, <2 x i1> poison) ; TYPE_BASED_ONLY-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %match_nxv16i8_v16i8 = call <vscale x 16 x i1> @llvm.experimental.vector.match.nxv16i8.v16i8(<vscale x 16 x i8> undef, <16 x i8> undef, <vscale x 16 x i1> undef) - %match_nxv8i16_v8i16 = call <vscale x 8 x i1> @llvm.experimental.vector.match.nxv8i16.v8i16(<vscale x 8 x i16> undef, <8 x i16> undef, <vscale x 8 x i1> undef) - %match_nxv4i32_v4i32 = call <vscale x 4 x i1> @llvm.experimental.vector.match.nxv4i32.v4i32(<vscale x 4 x i32> undef, <4 x i32> undef, <vscale x 4 x i1> undef) - %match_nxv2i64_v2i64 = call <vscale x 2 x i1> @llvm.experimental.vector.match.nxv2i64.v2i64(<vscale x 2 x i64> undef, <2 x i64> undef, <vscale x 2 x i1> undef) + %match_nxv16i8_v16i8 = call <vscale x 16 x i1> @llvm.experimental.vector.match.nxv16i8.v16i8(<vscale x 16 x i8> poison, <16 x i8> poison, <vscale x 16 x i1> poison) + %match_nxv8i16_v8i16 = call <vscale x 8 x i1> @llvm.experimental.vector.match.nxv8i16.v8i16(<vscale x 8 x i16> poison, <8 x i16> poison, <vscale x 8 x i1> poison) + %match_nxv4i32_v4i32 = call <vscale x 4 x i1> @llvm.experimental.vector.match.nxv4i32.v4i32(<vscale x 4 x i32> poison, <4 x i32> poison, <vscale x 4 x i1> poison) + %match_nxv2i64_v2i64 = call <vscale x 2 x i1> @llvm.experimental.vector.match.nxv2i64.v2i64(<vscale x 2 x i64> poison, <2 x i64> poison, <vscale x 2 x i1> poison) - %match_v16i8_v16i8 = call <16 x i1> @llvm.experimental.vector.match.v16i8.v16i8(<16 x i8> undef, <16 x i8> undef, <16 x i1> undef) - %match_v8i16_v8i16 = call <8 x i1> @llvm.experimental.vector.match.v8i16.v8i16(<8 x i16> undef, <8 x i16> undef, <8 x i1> undef) - %match_v4i32_v4i32 = call <4 x i1> @llvm.experimental.vector.match.v4i32.v4i32(<4 x i32> undef, <4 x i32> undef, <4 x i1> undef) - %match_v2i64_v2i64 = call <2 x i1> @llvm.experimental.vector.match.v2i64.v2i64(<2 x i64> undef, <2 x i64> undef, <2 x i1> undef) + %match_v16i8_v16i8 = call <16 x i1> @llvm.experimental.vector.match.v16i8.v16i8(<16 x i8> poison, <16 x i8> poison, <16 x i1> poison) + %match_v8i16_v8i16 = call <8 x i1> @llvm.experimental.vector.match.v8i16.v8i16(<8 x i16> poison, <8 x i16> poison, <8 x i1> poison) + %match_v4i32_v4i32 = call <4 x i1> @llvm.experimental.vector.match.v4i32.v4i32(<4 x i32> poison, <4 x i32> poison, <4 x i1> poison) + %match_v2i64_v2i64 = call <2 x i1> @llvm.experimental.vector.match.v2i64.v2i64(<2 x i64> poison, <2 x i64> poison, <2 x i1> poison) ret void } diff --git a/llvm/test/Analysis/CostModel/AArch64/sve-trunc.ll b/llvm/test/Analysis/CostModel/AArch64/sve-trunc.ll index 397b737..f7d3719 100644 --- a/llvm/test/Analysis/CostModel/AArch64/sve-trunc.ll +++ b/llvm/test/Analysis/CostModel/AArch64/sve-trunc.ll @@ -5,82 +5,82 @@ target datalayout = "e-m:e-i8:8:32-i16:16:32-i64:64-i128:128-n32:64-S128" define void @sve_truncs() { ; CHECK-LABEL: 'sve_truncs' -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv2i8_to_i1 = trunc <vscale x 2 x i8> undef to <vscale x 2 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv2i16_to_i1 = trunc <vscale x 2 x i16> undef to <vscale x 2 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv2i32_to_i1 = trunc <vscale x 2 x i32> undef to <vscale x 2 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv2i64_to_i1 = trunc <vscale x 2 x i64> undef to <vscale x 2 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv4i8_to_i1 = trunc <vscale x 4 x i8> undef to <vscale x 4 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv4i16_to_i1 = trunc <vscale x 4 x i16> undef to <vscale x 4 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv4i32_to_i1 = trunc <vscale x 4 x i32> undef to <vscale x 4 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv4i64_to_i1 = trunc <vscale x 4 x i64> undef to <vscale x 4 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv8i8_to_i1 = trunc <vscale x 8 x i8> undef to <vscale x 8 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv8i16_to_i1 = trunc <vscale x 8 x i16> undef to <vscale x 8 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv8i32_to_i1 = trunc <vscale x 8 x i32> undef to <vscale x 8 x i1> -; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv8i64_to_i1 = trunc <vscale x 8 x i64> undef to <vscale x 8 x i1> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv2i16_to_i8 = trunc <vscale x 2 x i16> undef to <vscale x 2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv2i32_to_i8 = trunc <vscale x 2 x i32> undef to <vscale x 2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv2i64_to_i8 = trunc <vscale x 2 x i64> undef to <vscale x 2 x i8> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv2i32_to_i16 = trunc <vscale x 2 x i32> undef to <vscale x 2 x i16> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv2i64_to_i16 = trunc <vscale x 2 x i64> undef to <vscale x 2 x i16> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv2i64_to_i32 = trunc <vscale x 2 x i64> undef to <vscale x 2 x i32> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv4i16_to_i8 = trunc <vscale x 4 x i16> undef to <vscale x 4 x i8> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv4i32_to_i8 = trunc <vscale x 4 x i32> undef to <vscale x 4 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %trunc_nxv4i64_to_i8 = trunc <vscale x 4 x i64> undef to <vscale x 4 x i8> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv4i32_to_i16 = trunc <vscale x 4 x i32> undef to <vscale x 4 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %trunc_nxv4i64_to_i16 = trunc <vscale x 4 x i64> undef to <vscale x 4 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %trunc_nxv4i64_to_i32 = trunc <vscale x 4 x i64> undef to <vscale x 4 x i32> -; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv8i16_to_i8 = trunc <vscale x 8 x i16> undef to <vscale x 8 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %trunc_nxv8i32_to_i8 = trunc <vscale x 8 x i32> undef to <vscale x 8 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv8i64_to_i8 = trunc <vscale x 8 x i64> undef to <vscale x 8 x i8> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %trunc_nxv8i32_to_i16 = trunc <vscale x 8 x i32> undef to <vscale x 8 x i16> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv8i64_to_i16 = trunc <vscale x 8 x i64> undef to <vscale x 8 x i16> -; CHECK-NEXT: Cost Model: Found costs of 1 for: %trunc_nxv16i16_to_i8 = trunc <vscale x 16 x i16> undef to <vscale x 16 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv16i32_to_i8 = trunc <vscale x 16 x i32> undef to <vscale x 16 x i8> -; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv16i64_to_i8 = trunc <vscale x 16 x i64> undef to <vscale x 16 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv2i8_to_i1 = trunc <vscale x 2 x i8> poison to <vscale x 2 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv2i16_to_i1 = trunc <vscale x 2 x i16> poison to <vscale x 2 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv2i32_to_i1 = trunc <vscale x 2 x i32> poison to <vscale x 2 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv2i64_to_i1 = trunc <vscale x 2 x i64> poison to <vscale x 2 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv4i8_to_i1 = trunc <vscale x 4 x i8> poison to <vscale x 4 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv4i16_to_i1 = trunc <vscale x 4 x i16> poison to <vscale x 4 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv4i32_to_i1 = trunc <vscale x 4 x i32> poison to <vscale x 4 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv4i64_to_i1 = trunc <vscale x 4 x i64> poison to <vscale x 4 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv8i8_to_i1 = trunc <vscale x 8 x i8> poison to <vscale x 8 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:2 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv8i16_to_i1 = trunc <vscale x 8 x i16> poison to <vscale x 8 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:5 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv8i32_to_i1 = trunc <vscale x 8 x i32> poison to <vscale x 8 x i1> +; CHECK-NEXT: Cost Model: Found costs of RThru:11 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv8i64_to_i1 = trunc <vscale x 8 x i64> poison to <vscale x 8 x i1> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv2i16_to_i8 = trunc <vscale x 2 x i16> poison to <vscale x 2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv2i32_to_i8 = trunc <vscale x 2 x i32> poison to <vscale x 2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv2i64_to_i8 = trunc <vscale x 2 x i64> poison to <vscale x 2 x i8> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv2i32_to_i16 = trunc <vscale x 2 x i32> poison to <vscale x 2 x i16> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv2i64_to_i16 = trunc <vscale x 2 x i64> poison to <vscale x 2 x i16> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv2i64_to_i32 = trunc <vscale x 2 x i64> poison to <vscale x 2 x i32> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv4i16_to_i8 = trunc <vscale x 4 x i16> poison to <vscale x 4 x i8> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv4i32_to_i8 = trunc <vscale x 4 x i32> poison to <vscale x 4 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %trunc_nxv4i64_to_i8 = trunc <vscale x 4 x i64> poison to <vscale x 4 x i8> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv4i32_to_i16 = trunc <vscale x 4 x i32> poison to <vscale x 4 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %trunc_nxv4i64_to_i16 = trunc <vscale x 4 x i64> poison to <vscale x 4 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %trunc_nxv4i64_to_i32 = trunc <vscale x 4 x i64> poison to <vscale x 4 x i32> +; CHECK-NEXT: Cost Model: Found costs of 0 for: %trunc_nxv8i16_to_i8 = trunc <vscale x 8 x i16> poison to <vscale x 8 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %trunc_nxv8i32_to_i8 = trunc <vscale x 8 x i32> poison to <vscale x 8 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv8i64_to_i8 = trunc <vscale x 8 x i64> poison to <vscale x 8 x i8> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %trunc_nxv8i32_to_i16 = trunc <vscale x 8 x i32> poison to <vscale x 8 x i16> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv8i64_to_i16 = trunc <vscale x 8 x i64> poison to <vscale x 8 x i16> +; CHECK-NEXT: Cost Model: Found costs of 1 for: %trunc_nxv16i16_to_i8 = trunc <vscale x 16 x i16> poison to <vscale x 16 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:3 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv16i32_to_i8 = trunc <vscale x 16 x i32> poison to <vscale x 16 x i8> +; CHECK-NEXT: Cost Model: Found costs of RThru:7 CodeSize:1 Lat:1 SizeLat:1 for: %trunc_nxv16i64_to_i8 = trunc <vscale x 16 x i64> poison to <vscale x 16 x i8> ; CHECK-NEXT: Cost Model: Found costs of RThru:0 CodeSize:1 Lat:1 SizeLat:1 for: ret void ; - %trunc_nxv2i8_to_i1 = trunc <vscale x 2 x i8> undef to <vscale x 2 x i1> - %trunc_nxv2i16_to_i1 = trunc <vscale x 2 x i16> undef to <vscale x 2 x i1> - %trunc_nxv2i32_to_i1 = trunc <vscale x 2 x i32> undef to <vscale x 2 x i1> - %trunc_nxv2i64_to_i1 = trunc <vscale x 2 x i64> undef to <vscale x 2 x i1> + %trunc_nxv2i8_to_i1 = trunc <vscale x 2 x i8> poison to <vscale x 2 x i1> + %trunc_nxv2i16_to_i1 = trunc <vscale x 2 x i16> poison to <vscale x 2 x i1> + %trunc_nxv2i32_to_i1 = trunc <vscale x 2 x i32> poison to <vscale x 2 x i1> + %trunc_nxv2i64_to_i1 = trunc <vscale x 2 x i64> poison to <vscale x 2 x i1> - %trunc_nxv4i8_to_i1 = trunc <vscale x 4 x i8> undef to <vscale x 4 x i1> - %trunc_nxv4i16_to_i1 = trunc <vscale x 4 x i16> undef to <vscale x 4 x i1> - %trunc_nxv4i32_to_i1 = trunc <vscale x 4 x i32> undef to <vscale x 4 x i1> - %trunc_nxv4i64_to_i1 = trunc <vscale x 4 x i64> undef to <vscale x 4 x i1> + %trunc_nxv4i8_to_i1 = trunc <vscale x 4 x i8> poison to <vscale x 4 x i1> + %trunc_nxv4i16_to_i1 = trunc <vscale x 4 x i16> poison to <vscale x 4 x i1> + %trunc_nxv4i32_to_i1 = trunc <vscale x 4 x i32> poison to <vscale x 4 x i1> + %trunc_nxv4i64_to_i1 = trunc <vscale x 4 x i64> poison to <vscale x 4 x i1> - %trunc_nxv8i8_to_i1 = trunc <vscale x 8 x i8> undef to <vscale x 8 x i1> - %trunc_nxv8i16_to_i1 = trunc <vscale x 8 x i16> undef to <vscale x 8 x i1> - %trunc_nxv8i32_to_i1 = trunc <vscale x 8 x i32> undef to <vscale x 8 x i1> - %trunc_nxv8i64_to_i1 = trunc <vscale x 8 x i64> undef to <vscale x 8 x i1> + %trunc_nxv8i8_to_i1 = trunc <vscale x 8 x i8> poison to <vscale x 8 x i1> + %trunc_nxv8i16_to_i1 = trunc <vscale x 8 x i16> poison to <vscale x 8 x i1> + %trunc_nxv8i32_to_i1 = trunc <vscale x 8 x i32> poison to <vscale x 8 x i1> + %trunc_nxv8i64_to_i1 = trunc <vscale x 8 x i64> poison to <vscale x 8 x i1> ; Truncates to unpacked or legal types with vscale x 2 elements - %trunc_nxv2i16_to_i8 = trunc <vscale x 2 x i16> undef to <vscale x 2 x i8> - %trunc_nxv2i32_to_i8 = trunc <vscale x 2 x i32> undef to <vscale x 2 x i8> - %trunc_nxv2i64_to_i8 = trunc <vscale x 2 x i64> undef to <vscale x 2 x i8> - %trunc_nxv2i32_to_i16 = trunc <vscale x 2 x i32> undef to <vscale x 2 x i16> - %trunc_nxv2i64_to_i16 = trunc <vscale x 2 x i64> undef to <vscale x 2 x i16> - %trunc_nxv2i64_to_i32 = trunc <vscale x 2 x i64> undef to <vscale x 2 x i32> + %trunc_nxv2i16_to_i8 = trunc <vscale x 2 x i16> poison to <vscale x 2 x i8> + %trunc_nxv2i32_to_i8 = trunc <vscale x 2 x i32> poison to <vscale x 2 x i8> + %trunc_nxv2i64_to_i8 = trunc <vscale x 2 x i64> poison to <vscale x 2 x i8> + %trunc_nxv2i32_to_i16 = trunc <vscale x 2 x i32> poison to <vscale x 2 x i16> + %trunc_nxv2i64_to_i16 = trunc <vscale x 2 x i64> poison to <vscale x 2 x i16> + %trunc_nxv2i64_to_i32 = trunc <vscale x 2 x i64> poison to <vscale x 2 x i32> ; Truncates to unpacked or legal with vscale x 4 elements - %trunc_nxv4i16_to_i8 = trunc <vscale x 4 x i16> undef to <vscale x 4 x i8> - %trunc_nxv4i32_to_i8 = trunc <vscale x 4 x i32> undef to <vscale x 4 x i8> - %trunc_nxv4i64_to_i8 = trunc <vscale x 4 x i64> undef to <vscale x 4 x i8> - %trunc_nxv4i32_to_i16 = trunc <vscale x 4 x i32> undef to <vscale x 4 x i16> - %trunc_nxv4i64_to_i16 = trunc <vscale x 4 x i64> undef to <vscale x 4 x i16> - %trunc_nxv4i64_to_i32 = trunc <vscale x 4 x i64> undef to <vscale x 4 x i32> + %trunc_nxv4i16_to_i8 = trunc <vscale x 4 x i16> poison to <vscale x 4 x i8> + %trunc_nxv4i32_to_i8 = trunc <vscale x 4 x i32> poison to <vscale x 4 x i8> + %trunc_nxv4i64_to_i8 = trunc <vscale x 4 x i64> poison to <vscale x 4 x i8> + %trunc_nxv4i32_to_i16 = trunc <vscale x 4 x i32> poison to <vscale x 4 x i16> + %trunc_nxv4i64_to_i16 = trunc <vscale x 4 x i64> poison to <vscale x 4 x i16> + %trunc_nxv4i64_to_i32 = trunc <vscale x 4 x i64> poison to <vscale x 4 x i32> ; Truncates to unpacked or legal with vscale x 8 elements - %trunc_nxv8i16_to_i8 = trunc <vscale x 8 x i16> undef to <vscale x 8 x i8> - %trunc_nxv8i32_to_i8 = trunc <vscale x 8 x i32> undef to <vscale x 8 x i8> - %trunc_nxv8i64_to_i8 = trunc <vscale x 8 x i64> undef to <vscale x 8 x i8> - %trunc_nxv8i32_to_i16 = trunc <vscale x 8 x i32> undef to <vscale x 8 x i16> - %trunc_nxv8i64_to_i16 = trunc <vscale x 8 x i64> undef to <vscale x 8 x i16> + %trunc_nxv8i16_to_i8 = trunc <vscale x 8 x i16> poison to <vscale x 8 x i8> + %trunc_nxv8i32_to_i8 = trunc <vscale x 8 x i32> poison to <vscale x 8 x i8> + %trunc_nxv8i64_to_i8 = trunc <vscale x 8 x i64> poison to <vscale x 8 x i8> + %trunc_nxv8i32_to_i16 = trunc <vscale x 8 x i32> poison to <vscale x 8 x i16> + %trunc_nxv8i64_to_i16 = trunc <vscale x 8 x i64> poison to <vscale x 8 x i16> ; Truncates to unpacked or legal with vscale x 16 elements - %trunc_nxv16i16_to_i8 = trunc <vscale x 16 x i16> undef to <vscale x 16 x i8> - %trunc_nxv16i32_to_i8 = trunc <vscale x 16 x i32> undef to <vscale x 16 x i8> - %trunc_nxv16i64_to_i8 = trunc <vscale x 16 x i64> undef to <vscale x 16 x i8> + %trunc_nxv16i16_to_i8 = trunc <vscale x 16 x i16> poison to <vscale x 16 x i8> + %trunc_nxv16i32_to_i8 = trunc <vscale x 16 x i32> poison to <vscale x 16 x i8> + %trunc_nxv16i64_to_i8 = trunc <vscale x 16 x i64> poison to <vscale x 16 x i8> ret void } diff --git a/llvm/test/Analysis/ScalarEvolution/ptrtoaddr.ll b/llvm/test/Analysis/ScalarEvolution/ptrtoaddr.ll new file mode 100644 index 0000000..ebab9f0 --- /dev/null +++ b/llvm/test/Analysis/ScalarEvolution/ptrtoaddr.ll @@ -0,0 +1,135 @@ +; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py +; RUN: opt < %s --data-layout="e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-f80:128-n8:16:32:64-S128" -S -disable-output -disable-verify "-passes=print<scalar-evolution>" 2>&1 | FileCheck --check-prefixes=ALL,X64 %s +; RUN: opt < %s --data-layout="e-m:e-p:32:32-p270:32:32-p271:32:32-p272:64:64-f64:32:64-f80:32-n8:16:32-S128" -S -disable-output -disable-verify "-passes=print<scalar-evolution>" 2>&1 | FileCheck --check-prefixes=ALL,X32 %s + +declare void @useptr(ptr) + +define void @ptrtoaddr(ptr %in, ptr %out0, ptr %out1, ptr %out2, ptr %out3) { +; X64-LABEL: 'ptrtoaddr' +; X64-NEXT: Classifying expressions for: @ptrtoaddr +; X64-NEXT: %p0 = ptrtoaddr ptr %in to i64 +; X64-NEXT: --> %p0 U: full-set S: full-set +; X64-NEXT: %p1 = ptrtoaddr ptr %in to i32 +; X64-NEXT: --> %p1 U: full-set S: full-set +; X64-NEXT: %p2 = ptrtoaddr ptr %in to i16 +; X64-NEXT: --> %p2 U: full-set S: full-set +; X64-NEXT: %p3 = ptrtoaddr ptr %in to i128 +; X64-NEXT: --> %p3 U: full-set S: full-set +; X64-NEXT: Determining loop execution counts for: @ptrtoaddr +; +; X32-LABEL: 'ptrtoaddr' +; X32-NEXT: Classifying expressions for: @ptrtoaddr +; X32-NEXT: %p0 = ptrtoaddr ptr %in to i64 +; X32-NEXT: --> %p0 U: full-set S: full-set +; X32-NEXT: %p1 = ptrtoaddr ptr %in to i32 +; X32-NEXT: --> %p1 U: full-set S: full-set +; X32-NEXT: %p2 = ptrtoaddr ptr %in to i16 +; X32-NEXT: --> %p2 U: full-set S: full-set +; X32-NEXT: %p3 = ptrtoaddr ptr %in to i128 +; X32-NEXT: --> %p3 U: full-set S: full-set +; X32-NEXT: Determining loop execution counts for: @ptrtoaddr +; + %p0 = ptrtoaddr ptr %in to i64 + %p1 = ptrtoaddr ptr %in to i32 + %p2 = ptrtoaddr ptr %in to i16 + %p3 = ptrtoaddr ptr %in to i128 + store i64 %p0, ptr %out0 + store i32 %p1, ptr %out1 + store i16 %p2, ptr %out2 + store i128 %p3, ptr %out3 + ret void +} + +define void @ptrtoaddr_as1(ptr addrspace(1) %in, ptr %out0, ptr %out1, ptr %out2, ptr %out3) { +; X64-LABEL: 'ptrtoaddr_as1' +; X64-NEXT: Classifying expressions for: @ptrtoaddr_as1 +; X64-NEXT: %p0 = ptrtoaddr ptr addrspace(1) %in to i64 +; X64-NEXT: --> %p0 U: full-set S: full-set +; X64-NEXT: %p1 = ptrtoaddr ptr addrspace(1) %in to i32 +; X64-NEXT: --> %p1 U: full-set S: full-set +; X64-NEXT: %p2 = ptrtoaddr ptr addrspace(1) %in to i16 +; X64-NEXT: --> %p2 U: full-set S: full-set +; X64-NEXT: %p3 = ptrtoaddr ptr addrspace(1) %in to i128 +; X64-NEXT: --> %p3 U: full-set S: full-set +; X64-NEXT: Determining loop execution counts for: @ptrtoaddr_as1 +; +; X32-LABEL: 'ptrtoaddr_as1' +; X32-NEXT: Classifying expressions for: @ptrtoaddr_as1 +; X32-NEXT: %p0 = ptrtoaddr ptr addrspace(1) %in to i64 +; X32-NEXT: --> %p0 U: full-set S: full-set +; X32-NEXT: %p1 = ptrtoaddr ptr addrspace(1) %in to i32 +; X32-NEXT: --> %p1 U: full-set S: full-set +; X32-NEXT: %p2 = ptrtoaddr ptr addrspace(1) %in to i16 +; X32-NEXT: --> %p2 U: full-set S: full-set +; X32-NEXT: %p3 = ptrtoaddr ptr addrspace(1) %in to i128 +; X32-NEXT: --> %p3 U: full-set S: full-set +; X32-NEXT: Determining loop execution counts for: @ptrtoaddr_as1 +; + %p0 = ptrtoaddr ptr addrspace(1) %in to i64 + %p1 = ptrtoaddr ptr addrspace(1) %in to i32 + %p2 = ptrtoaddr ptr addrspace(1) %in to i16 + %p3 = ptrtoaddr ptr addrspace(1) %in to i128 + store i64 %p0, ptr %out0 + store i32 %p1, ptr %out1 + store i16 %p2, ptr %out2 + store i128 %p3, ptr %out3 + ret void +} + +define void @ptrtoaddr_of_bitcast(ptr %in, ptr %out0) { +; X64-LABEL: 'ptrtoaddr_of_bitcast' +; X64-NEXT: Classifying expressions for: @ptrtoaddr_of_bitcast +; X64-NEXT: %in_casted = bitcast ptr %in to ptr +; X64-NEXT: --> %in U: full-set S: full-set +; X64-NEXT: %p0 = ptrtoaddr ptr %in_casted to i64 +; X64-NEXT: --> %p0 U: full-set S: full-set +; X64-NEXT: Determining loop execution counts for: @ptrtoaddr_of_bitcast +; +; X32-LABEL: 'ptrtoaddr_of_bitcast' +; X32-NEXT: Classifying expressions for: @ptrtoaddr_of_bitcast +; X32-NEXT: %in_casted = bitcast ptr %in to ptr +; X32-NEXT: --> %in U: full-set S: full-set +; X32-NEXT: %p0 = ptrtoaddr ptr %in_casted to i64 +; X32-NEXT: --> %p0 U: full-set S: full-set +; X32-NEXT: Determining loop execution counts for: @ptrtoaddr_of_bitcast +; + %in_casted = bitcast ptr %in to ptr + %p0 = ptrtoaddr ptr %in_casted to i64 + store i64 %p0, ptr %out0 + ret void +} + +define void @ptrtoaddr_of_nullptr(ptr %out0) { +; ALL-LABEL: 'ptrtoaddr_of_nullptr' +; ALL-NEXT: Classifying expressions for: @ptrtoaddr_of_nullptr +; ALL-NEXT: %p0 = ptrtoaddr ptr null to i64 +; ALL-NEXT: --> %p0 U: full-set S: full-set +; ALL-NEXT: Determining loop execution counts for: @ptrtoaddr_of_nullptr +; + %p0 = ptrtoaddr ptr null to i64 + store i64 %p0, ptr %out0 + ret void +} + +define void @ptrtoaddr_of_gep(ptr %in, ptr %out0) { +; X64-LABEL: 'ptrtoaddr_of_gep' +; X64-NEXT: Classifying expressions for: @ptrtoaddr_of_gep +; X64-NEXT: %in_adj = getelementptr inbounds i8, ptr %in, i64 42 +; X64-NEXT: --> (42 + %in) U: full-set S: full-set +; X64-NEXT: %p0 = ptrtoaddr ptr %in_adj to i64 +; X64-NEXT: --> %p0 U: full-set S: full-set +; X64-NEXT: Determining loop execution counts for: @ptrtoaddr_of_gep +; +; X32-LABEL: 'ptrtoaddr_of_gep' +; X32-NEXT: Classifying expressions for: @ptrtoaddr_of_gep +; X32-NEXT: %in_adj = getelementptr inbounds i8, ptr %in, i64 42 +; X32-NEXT: --> (42 + %in) U: full-set S: full-set +; X32-NEXT: %p0 = ptrtoaddr ptr %in_adj to i64 +; X32-NEXT: --> %p0 U: full-set S: full-set +; X32-NEXT: Determining loop execution counts for: @ptrtoaddr_of_gep +; + %in_adj = getelementptr inbounds i8, ptr %in, i64 42 + %p0 = ptrtoaddr ptr %in_adj to i64 + store i64 %p0, ptr %out0 + ret void +} diff --git a/llvm/test/Analysis/ScalarEvolution/ptrtoint.ll b/llvm/test/Analysis/ScalarEvolution/ptrtoint.ll index acac2c9..0c1f37b 100644 --- a/llvm/test/Analysis/ScalarEvolution/ptrtoint.ll +++ b/llvm/test/Analysis/ScalarEvolution/ptrtoint.ll @@ -382,7 +382,7 @@ define void @pr46786_c26_char(ptr %arg, ptr %arg1, ptr %arg2) { ; X64-NEXT: %i9 = ptrtoint ptr %i7 to i64 ; X64-NEXT: --> {(ptrtoint ptr %arg to i64),+,1}<nuw><%bb6> U: full-set S: full-set Exits: (-1 + (ptrtoint ptr %arg1 to i64)) LoopDispositions: { %bb6: Computable } ; X64-NEXT: %i10 = sub i64 %i9, %i4 -; X64-NEXT: --> {0,+,1}<nuw><%bb6> U: full-set S: full-set Exits: (-1 + (-1 * (ptrtoint ptr %arg to i64)) + (ptrtoint ptr %arg1 to i64)) LoopDispositions: { %bb6: Computable } +; X64-NEXT: --> {0,+,1}<nuw><%bb6> U: [0,-1) S: [0,-1) Exits: (-1 + (-1 * (ptrtoint ptr %arg to i64)) + (ptrtoint ptr %arg1 to i64)) LoopDispositions: { %bb6: Computable } ; X64-NEXT: %i11 = getelementptr inbounds i8, ptr %arg2, i64 %i10 ; X64-NEXT: --> {%arg2,+,1}<nw><%bb6> U: full-set S: full-set Exits: (-1 + (-1 * (ptrtoint ptr %arg to i64)) + (ptrtoint ptr %arg1 to i64) + %arg2) LoopDispositions: { %bb6: Computable } ; X64-NEXT: %i12 = load i8, ptr %i11, align 1 @@ -393,7 +393,7 @@ define void @pr46786_c26_char(ptr %arg, ptr %arg1, ptr %arg2) { ; X64-NEXT: --> {(1 + %arg),+,1}<nuw><%bb6> U: full-set S: full-set Exits: ((-1 * (ptrtoint ptr %arg to i64)) + (ptrtoint ptr %arg1 to i64) + %arg) LoopDispositions: { %bb6: Computable } ; X64-NEXT: Determining loop execution counts for: @pr46786_c26_char ; X64-NEXT: Loop %bb6: backedge-taken count is (-1 + (-1 * (ptrtoint ptr %arg to i64)) + (ptrtoint ptr %arg1 to i64)) -; X64-NEXT: Loop %bb6: constant max backedge-taken count is i64 -1 +; X64-NEXT: Loop %bb6: constant max backedge-taken count is i64 -2 ; X64-NEXT: Loop %bb6: symbolic max backedge-taken count is (-1 + (-1 * (ptrtoint ptr %arg to i64)) + (ptrtoint ptr %arg1 to i64)) ; X64-NEXT: Loop %bb6: Trip multiple is 1 ; @@ -406,9 +406,9 @@ define void @pr46786_c26_char(ptr %arg, ptr %arg1, ptr %arg2) { ; X32-NEXT: %i8 = load i8, ptr %i7, align 1 ; X32-NEXT: --> %i8 U: full-set S: full-set Exits: <<Unknown>> LoopDispositions: { %bb6: Variant } ; X32-NEXT: %i9 = ptrtoint ptr %i7 to i64 -; X32-NEXT: --> {(zext i32 (ptrtoint ptr %arg to i32) to i64),+,1}<nuw><%bb6> U: [0,8589934591) S: [0,8589934591) Exits: ((zext i32 (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32)) to i64) + (zext i32 (ptrtoint ptr %arg to i32) to i64)) LoopDispositions: { %bb6: Computable } +; X32-NEXT: --> {(zext i32 (ptrtoint ptr %arg to i32) to i64),+,1}<nuw><%bb6> U: [0,8589934590) S: [0,8589934590) Exits: ((zext i32 (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32)) to i64) + (zext i32 (ptrtoint ptr %arg to i32) to i64)) LoopDispositions: { %bb6: Computable } ; X32-NEXT: %i10 = sub i64 %i9, %i4 -; X32-NEXT: --> {0,+,1}<nuw><%bb6> U: [0,4294967296) S: [0,4294967296) Exits: (zext i32 (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32)) to i64) LoopDispositions: { %bb6: Computable } +; X32-NEXT: --> {0,+,1}<nuw><%bb6> U: [0,4294967295) S: [0,4294967295) Exits: (zext i32 (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32)) to i64) LoopDispositions: { %bb6: Computable } ; X32-NEXT: %i11 = getelementptr inbounds i8, ptr %arg2, i64 %i10 ; X32-NEXT: --> {%arg2,+,1}<%bb6> U: full-set S: full-set Exits: (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32) + %arg2) LoopDispositions: { %bb6: Computable } ; X32-NEXT: %i12 = load i8, ptr %i11, align 1 @@ -419,7 +419,7 @@ define void @pr46786_c26_char(ptr %arg, ptr %arg1, ptr %arg2) { ; X32-NEXT: --> {(1 + %arg),+,1}<nuw><%bb6> U: full-set S: full-set Exits: ((-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32) + %arg) LoopDispositions: { %bb6: Computable } ; X32-NEXT: Determining loop execution counts for: @pr46786_c26_char ; X32-NEXT: Loop %bb6: backedge-taken count is (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32)) -; X32-NEXT: Loop %bb6: constant max backedge-taken count is i32 -1 +; X32-NEXT: Loop %bb6: constant max backedge-taken count is i32 -2 ; X32-NEXT: Loop %bb6: symbolic max backedge-taken count is (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32)) ; X32-NEXT: Loop %bb6: Trip multiple is 1 ; @@ -459,7 +459,7 @@ define void @pr46786_c26_char_cmp_ops_swapped(ptr %arg, ptr %arg1, ptr %arg2) { ; X64-NEXT: %i9 = ptrtoint ptr %i7 to i64 ; X64-NEXT: --> {(ptrtoint ptr %arg to i64),+,1}<nuw><%bb6> U: full-set S: full-set Exits: (-1 + (ptrtoint ptr %arg1 to i64)) LoopDispositions: { %bb6: Computable } ; X64-NEXT: %i10 = sub i64 %i9, %i4 -; X64-NEXT: --> {0,+,1}<nuw><%bb6> U: full-set S: full-set Exits: (-1 + (-1 * (ptrtoint ptr %arg to i64)) + (ptrtoint ptr %arg1 to i64)) LoopDispositions: { %bb6: Computable } +; X64-NEXT: --> {0,+,1}<nuw><%bb6> U: [0,-1) S: [0,-1) Exits: (-1 + (-1 * (ptrtoint ptr %arg to i64)) + (ptrtoint ptr %arg1 to i64)) LoopDispositions: { %bb6: Computable } ; X64-NEXT: %i11 = getelementptr inbounds i8, ptr %arg2, i64 %i10 ; X64-NEXT: --> {%arg2,+,1}<nw><%bb6> U: full-set S: full-set Exits: (-1 + (-1 * (ptrtoint ptr %arg to i64)) + (ptrtoint ptr %arg1 to i64) + %arg2) LoopDispositions: { %bb6: Computable } ; X64-NEXT: %i12 = load i8, ptr %i11, align 1 @@ -470,7 +470,7 @@ define void @pr46786_c26_char_cmp_ops_swapped(ptr %arg, ptr %arg1, ptr %arg2) { ; X64-NEXT: --> {(1 + %arg),+,1}<nuw><%bb6> U: full-set S: full-set Exits: ((-1 * (ptrtoint ptr %arg to i64)) + (ptrtoint ptr %arg1 to i64) + %arg) LoopDispositions: { %bb6: Computable } ; X64-NEXT: Determining loop execution counts for: @pr46786_c26_char_cmp_ops_swapped ; X64-NEXT: Loop %bb6: backedge-taken count is (-1 + (-1 * (ptrtoint ptr %arg to i64)) + (ptrtoint ptr %arg1 to i64)) -; X64-NEXT: Loop %bb6: constant max backedge-taken count is i64 -1 +; X64-NEXT: Loop %bb6: constant max backedge-taken count is i64 -2 ; X64-NEXT: Loop %bb6: symbolic max backedge-taken count is (-1 + (-1 * (ptrtoint ptr %arg to i64)) + (ptrtoint ptr %arg1 to i64)) ; X64-NEXT: Loop %bb6: Trip multiple is 1 ; @@ -483,9 +483,9 @@ define void @pr46786_c26_char_cmp_ops_swapped(ptr %arg, ptr %arg1, ptr %arg2) { ; X32-NEXT: %i8 = load i8, ptr %i7, align 1 ; X32-NEXT: --> %i8 U: full-set S: full-set Exits: <<Unknown>> LoopDispositions: { %bb6: Variant } ; X32-NEXT: %i9 = ptrtoint ptr %i7 to i64 -; X32-NEXT: --> {(zext i32 (ptrtoint ptr %arg to i32) to i64),+,1}<nuw><%bb6> U: [0,8589934591) S: [0,8589934591) Exits: ((zext i32 (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32)) to i64) + (zext i32 (ptrtoint ptr %arg to i32) to i64)) LoopDispositions: { %bb6: Computable } +; X32-NEXT: --> {(zext i32 (ptrtoint ptr %arg to i32) to i64),+,1}<nuw><%bb6> U: [0,8589934590) S: [0,8589934590) Exits: ((zext i32 (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32)) to i64) + (zext i32 (ptrtoint ptr %arg to i32) to i64)) LoopDispositions: { %bb6: Computable } ; X32-NEXT: %i10 = sub i64 %i9, %i4 -; X32-NEXT: --> {0,+,1}<nuw><%bb6> U: [0,4294967296) S: [0,4294967296) Exits: (zext i32 (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32)) to i64) LoopDispositions: { %bb6: Computable } +; X32-NEXT: --> {0,+,1}<nuw><%bb6> U: [0,4294967295) S: [0,4294967295) Exits: (zext i32 (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32)) to i64) LoopDispositions: { %bb6: Computable } ; X32-NEXT: %i11 = getelementptr inbounds i8, ptr %arg2, i64 %i10 ; X32-NEXT: --> {%arg2,+,1}<%bb6> U: full-set S: full-set Exits: (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32) + %arg2) LoopDispositions: { %bb6: Computable } ; X32-NEXT: %i12 = load i8, ptr %i11, align 1 @@ -496,7 +496,7 @@ define void @pr46786_c26_char_cmp_ops_swapped(ptr %arg, ptr %arg1, ptr %arg2) { ; X32-NEXT: --> {(1 + %arg),+,1}<nuw><%bb6> U: full-set S: full-set Exits: ((-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32) + %arg) LoopDispositions: { %bb6: Computable } ; X32-NEXT: Determining loop execution counts for: @pr46786_c26_char_cmp_ops_swapped ; X32-NEXT: Loop %bb6: backedge-taken count is (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32)) -; X32-NEXT: Loop %bb6: constant max backedge-taken count is i32 -1 +; X32-NEXT: Loop %bb6: constant max backedge-taken count is i32 -2 ; X32-NEXT: Loop %bb6: symbolic max backedge-taken count is (-1 + (-1 * (ptrtoint ptr %arg to i32)) + (ptrtoint ptr %arg1 to i32)) ; X32-NEXT: Loop %bb6: Trip multiple is 1 ; diff --git a/llvm/test/Analysis/ScalarEvolution/trip-multiple-guard-info.ll b/llvm/test/Analysis/ScalarEvolution/trip-multiple-guard-info.ll index 7ba422d..a477465c 100644 --- a/llvm/test/Analysis/ScalarEvolution/trip-multiple-guard-info.ll +++ b/llvm/test/Analysis/ScalarEvolution/trip-multiple-guard-info.ll @@ -578,22 +578,22 @@ define void @test_ptr_aligned_by_2_and_4_via_assumption(ptr %start, ptr %end) { ; CHECK-LABEL: 'test_ptr_aligned_by_2_and_4_via_assumption' ; CHECK-NEXT: Classifying expressions for: @test_ptr_aligned_by_2_and_4_via_assumption ; CHECK-NEXT: %iv = phi ptr [ %start, %entry ], [ %iv.next, %loop ] -; CHECK-NEXT: --> {%start,+,4}<%loop> U: full-set S: full-set Exits: <<Unknown>> LoopDispositions: { %loop: Computable } +; CHECK-NEXT: --> {%start,+,4}<%loop> U: [0,-1) S: [-9223372036854775808,9223372036854775807) Exits: <<Unknown>> LoopDispositions: { %loop: Computable } ; CHECK-NEXT: %iv.next = getelementptr i8, ptr %iv, i64 4 -; CHECK-NEXT: --> {(4 + %start),+,4}<%loop> U: full-set S: full-set Exits: <<Unknown>> LoopDispositions: { %loop: Computable } +; CHECK-NEXT: --> {(4 + %start),+,4}<%loop> U: [0,-1) S: [-9223372036854775808,9223372036854775807) Exits: <<Unknown>> LoopDispositions: { %loop: Computable } ; CHECK-NEXT: Determining loop execution counts for: @test_ptr_aligned_by_2_and_4_via_assumption ; CHECK-NEXT: Loop %loop: Unpredictable backedge-taken count. ; CHECK-NEXT: Loop %loop: Unpredictable constant max backedge-taken count. ; CHECK-NEXT: Loop %loop: Unpredictable symbolic max backedge-taken count. ; CHECK-NEXT: Loop %loop: Predicated backedge-taken count is ((-4 + (-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64)) /u 4) ; CHECK-NEXT: Predicates: -; CHECK-NEXT: Equal predicate: (zext i2 ((trunc i64 (ptrtoint ptr %end to i64) to i2) + (-1 * (trunc i64 (ptrtoint ptr %start to i64) to i2))) to i64) == 0 +; CHECK-NEXT: Equal predicate: (zext i2 (-1 * (trunc i64 (ptrtoint ptr %start to i64) to i2)) to i64) == 0 ; CHECK-NEXT: Loop %loop: Predicated constant max backedge-taken count is i64 4611686018427387903 ; CHECK-NEXT: Predicates: -; CHECK-NEXT: Equal predicate: (zext i2 ((trunc i64 (ptrtoint ptr %end to i64) to i2) + (-1 * (trunc i64 (ptrtoint ptr %start to i64) to i2))) to i64) == 0 +; CHECK-NEXT: Equal predicate: (zext i2 (-1 * (trunc i64 (ptrtoint ptr %start to i64) to i2)) to i64) == 0 ; CHECK-NEXT: Loop %loop: Predicated symbolic max backedge-taken count is ((-4 + (-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64)) /u 4) ; CHECK-NEXT: Predicates: -; CHECK-NEXT: Equal predicate: (zext i2 ((trunc i64 (ptrtoint ptr %end to i64) to i2) + (-1 * (trunc i64 (ptrtoint ptr %start to i64) to i2))) to i64) == 0 +; CHECK-NEXT: Equal predicate: (zext i2 (-1 * (trunc i64 (ptrtoint ptr %start to i64) to i2)) to i64) == 0 ; entry: call void @llvm.assume(i1 true) [ "align"(ptr %start, i64 2) ] @@ -615,9 +615,9 @@ define void @test_ptrs_aligned_by_4_via_assumption(ptr %start, ptr %end) { ; CHECK-LABEL: 'test_ptrs_aligned_by_4_via_assumption' ; CHECK-NEXT: Classifying expressions for: @test_ptrs_aligned_by_4_via_assumption ; CHECK-NEXT: %iv = phi ptr [ %start, %entry ], [ %iv.next, %loop ] -; CHECK-NEXT: --> {%start,+,4}<%loop> U: full-set S: full-set Exits: ((4 * ((-4 + (-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64)) /u 4))<nuw> + %start) LoopDispositions: { %loop: Computable } +; CHECK-NEXT: --> {%start,+,4}<%loop> U: [0,-3) S: [-9223372036854775808,9223372036854775805) Exits: (-4 + (-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64) + %start) LoopDispositions: { %loop: Computable } ; CHECK-NEXT: %iv.next = getelementptr i8, ptr %iv, i64 4 -; CHECK-NEXT: --> {(4 + %start),+,4}<%loop> U: full-set S: full-set Exits: (4 + (4 * ((-4 + (-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64)) /u 4))<nuw> + %start) LoopDispositions: { %loop: Computable } +; CHECK-NEXT: --> {(4 + %start),+,4}<%loop> U: [0,-3) S: [-9223372036854775808,9223372036854775805) Exits: ((-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64) + %start) LoopDispositions: { %loop: Computable } ; CHECK-NEXT: Determining loop execution counts for: @test_ptrs_aligned_by_4_via_assumption ; CHECK-NEXT: Loop %loop: backedge-taken count is ((-4 + (-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64)) /u 4) ; CHECK-NEXT: Loop %loop: constant max backedge-taken count is i64 4611686018427387903 @@ -644,9 +644,9 @@ define void @test_ptrs_aligned_by_8_via_assumption(ptr %start, ptr %end) { ; CHECK-LABEL: 'test_ptrs_aligned_by_8_via_assumption' ; CHECK-NEXT: Classifying expressions for: @test_ptrs_aligned_by_8_via_assumption ; CHECK-NEXT: %iv = phi ptr [ %start, %entry ], [ %iv.next, %loop ] -; CHECK-NEXT: --> {%start,+,4}<%loop> U: full-set S: full-set Exits: ((4 * ((-4 + (-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64)) /u 4))<nuw> + %start) LoopDispositions: { %loop: Computable } +; CHECK-NEXT: --> {%start,+,4}<%loop> U: [0,-3) S: [-9223372036854775808,9223372036854775805) Exits: (-4 + (-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64) + %start) LoopDispositions: { %loop: Computable } ; CHECK-NEXT: %iv.next = getelementptr i8, ptr %iv, i64 4 -; CHECK-NEXT: --> {(4 + %start),+,4}<%loop> U: full-set S: full-set Exits: (4 + (4 * ((-4 + (-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64)) /u 4))<nuw> + %start) LoopDispositions: { %loop: Computable } +; CHECK-NEXT: --> {(4 + %start)<nuw><nsw>,+,4}<%loop> U: [0,-3) S: [-9223372036854775808,9223372036854775805) Exits: ((-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64) + %start) LoopDispositions: { %loop: Computable } ; CHECK-NEXT: Determining loop execution counts for: @test_ptrs_aligned_by_8_via_assumption ; CHECK-NEXT: Loop %loop: backedge-taken count is ((-4 + (-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64)) /u 4) ; CHECK-NEXT: Loop %loop: constant max backedge-taken count is i64 4611686018427387903 @@ -677,22 +677,22 @@ define void @test_ptr_aligned_by_4_via_assumption_multiple_loop_predecessors(ptr ; CHECK-NEXT: %c = call i1 @cond() ; CHECK-NEXT: --> %c U: full-set S: full-set ; CHECK-NEXT: %iv = phi ptr [ %start, %then ], [ %start, %else ], [ %iv.next, %loop ] -; CHECK-NEXT: --> {%start,+,4}<%loop> U: full-set S: full-set Exits: <<Unknown>> LoopDispositions: { %loop: Computable } +; CHECK-NEXT: --> {%start,+,4}<%loop> U: [0,-1) S: [-9223372036854775808,9223372036854775807) Exits: <<Unknown>> LoopDispositions: { %loop: Computable } ; CHECK-NEXT: %iv.next = getelementptr i8, ptr %iv, i64 4 -; CHECK-NEXT: --> {(4 + %start),+,4}<%loop> U: full-set S: full-set Exits: <<Unknown>> LoopDispositions: { %loop: Computable } +; CHECK-NEXT: --> {(4 + %start),+,4}<%loop> U: [0,-1) S: [-9223372036854775808,9223372036854775807) Exits: <<Unknown>> LoopDispositions: { %loop: Computable } ; CHECK-NEXT: Determining loop execution counts for: @test_ptr_aligned_by_4_via_assumption_multiple_loop_predecessors ; CHECK-NEXT: Loop %loop: Unpredictable backedge-taken count. ; CHECK-NEXT: Loop %loop: Unpredictable constant max backedge-taken count. ; CHECK-NEXT: Loop %loop: Unpredictable symbolic max backedge-taken count. ; CHECK-NEXT: Loop %loop: Predicated backedge-taken count is ((-4 + (-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64)) /u 4) ; CHECK-NEXT: Predicates: -; CHECK-NEXT: Equal predicate: (zext i2 ((trunc i64 (ptrtoint ptr %end to i64) to i2) + (-1 * (trunc i64 (ptrtoint ptr %start to i64) to i2))) to i64) == 0 +; CHECK-NEXT: Equal predicate: (zext i2 (-1 * (trunc i64 (ptrtoint ptr %start to i64) to i2)) to i64) == 0 ; CHECK-NEXT: Loop %loop: Predicated constant max backedge-taken count is i64 4611686018427387903 ; CHECK-NEXT: Predicates: -; CHECK-NEXT: Equal predicate: (zext i2 ((trunc i64 (ptrtoint ptr %end to i64) to i2) + (-1 * (trunc i64 (ptrtoint ptr %start to i64) to i2))) to i64) == 0 +; CHECK-NEXT: Equal predicate: (zext i2 (-1 * (trunc i64 (ptrtoint ptr %start to i64) to i2)) to i64) == 0 ; CHECK-NEXT: Loop %loop: Predicated symbolic max backedge-taken count is ((-4 + (-1 * (ptrtoint ptr %start to i64)) + (ptrtoint ptr %end to i64)) /u 4) ; CHECK-NEXT: Predicates: -; CHECK-NEXT: Equal predicate: (zext i2 ((trunc i64 (ptrtoint ptr %end to i64) to i2) + (-1 * (trunc i64 (ptrtoint ptr %start to i64) to i2))) to i64) == 0 +; CHECK-NEXT: Equal predicate: (zext i2 (-1 * (trunc i64 (ptrtoint ptr %start to i64) to i2)) to i64) == 0 ; entry: call void @llvm.assume(i1 true) [ "align"(ptr %start, i64 2) ] diff --git a/llvm/test/Analysis/TypeBasedAliasAnalysis/intrinsics.ll b/llvm/test/Analysis/TypeBasedAliasAnalysis/intrinsics.ll index 7ec674a..dc4a72e 100644 --- a/llvm/test/Analysis/TypeBasedAliasAnalysis/intrinsics.ll +++ b/llvm/test/Analysis/TypeBasedAliasAnalysis/intrinsics.ll @@ -1,3 +1,4 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 6 ; RUN: opt -aa-pipeline=tbaa,basic-aa -passes=gvn -S < %s | FileCheck %s target datalayout = "e-p:32:32:32-i1:8:32-i8:8:32-i16:16:32-i32:32:32-i64:32:32-f32:32:32-f64:32:32-v64:32:64-v128:32:128-a0:0:32-n32" @@ -5,12 +6,15 @@ target datalayout = "e-p:32:32:32-i1:8:32-i8:8:32-i16:16:32-i32:32:32-i64:32:32- ; TBAA should prove that these calls don't interfere, since they are ; IntrArgReadMem and have TBAA metadata. -; CHECK: define <8 x i16> @test0(ptr %p, ptr %q, <8 x i16> %y, <8 x i1> %m, <8 x i16> %pt) { -; CHECK-NEXT: entry: -; CHECK-NEXT: %a = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr %p, i32 16, <8 x i1> %m, <8 x i16> %pt) [[NUW:#[0-9]+]] -; CHECK-NEXT: call void @llvm.masked.store.v8i16.p0(<8 x i16> %y, ptr %q, i32 16, <8 x i1> %m) -; CHECK-NEXT: %c = add <8 x i16> %a, %a define <8 x i16> @test0(ptr %p, ptr %q, <8 x i16> %y, <8 x i1> %m, <8 x i16> %pt) { +; CHECK-LABEL: define <8 x i16> @test0( +; CHECK-SAME: ptr [[P:%.*]], ptr [[Q:%.*]], <8 x i16> [[Y:%.*]], <8 x i1> [[M:%.*]], <8 x i16> [[PT:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[A:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[P]], i32 16, <8 x i1> [[M]], <8 x i16> [[PT]]) #[[ATTR2:[0-9]+]], !tbaa [[B_TBAA0:![0-9]+]] +; CHECK-NEXT: call void @llvm.masked.store.v8i16.p0(<8 x i16> [[Y]], ptr [[Q]], i32 16, <8 x i1> [[M]]), !tbaa [[A_TBAA3:![0-9]+]] +; CHECK-NEXT: [[C:%.*]] = add <8 x i16> [[A]], [[A]] +; CHECK-NEXT: ret <8 x i16> [[C]] +; entry: %a = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr %p, i32 16, <8 x i1> %m, <8 x i16> %pt) nounwind, !tbaa !2 call void @llvm.masked.store.v8i16.p0(<8 x i16> %y, ptr %q, i32 16, <8 x i1> %m), !tbaa !1 @@ -24,10 +28,16 @@ declare void @llvm.masked.store.v8i16.p0(<8 x i16>, ptr, i32, <8 x i1>) nounwind ; CHECK: attributes #0 = { nocallback nofree nosync nounwind willreturn memory(argmem: read) } ; CHECK: attributes #1 = { nocallback nofree nosync nounwind willreturn memory(argmem: write) } -; CHECK: attributes [[NUW]] = { nounwind } !0 = !{!"tbaa root"} !1 = !{!3, !3, i64 0} !2 = !{!4, !4, i64 0} !3 = !{!"A", !0} !4 = !{!"B", !0} +;. +; CHECK: [[B_TBAA0]] = !{[[META1:![0-9]+]], [[META1]], i64 0} +; CHECK: [[META1]] = !{!"B", [[META2:![0-9]+]]} +; CHECK: [[META2]] = !{!"tbaa root"} +; CHECK: [[A_TBAA3]] = !{[[META4:![0-9]+]], [[META4]], i64 0} +; CHECK: [[META4]] = !{!"A", [[META2]]} +;. diff --git a/llvm/test/Assembler/autoupgrade-lifetime-intrinsics.ll b/llvm/test/Assembler/autoupgrade-lifetime-intrinsics.ll index 377c002..49174d2 100644 --- a/llvm/test/Assembler/autoupgrade-lifetime-intrinsics.ll +++ b/llvm/test/Assembler/autoupgrade-lifetime-intrinsics.ll @@ -56,6 +56,45 @@ define void @remove_unanalyzable(ptr %p) { ret void } +define void @no_declaration() { +; CHECK-LABEL: define void @no_declaration() { +; CHECK-NEXT: [[A:%.*]] = alloca i8, align 1, addrspace(2) +; CHECK-NEXT: call void @llvm.lifetime.start.p2(ptr addrspace(2) [[A]]) +; CHECK-NEXT: call void @llvm.lifetime.end.p2(ptr addrspace(2) [[A]]) +; CHECK-NEXT: ret void +; + %a = alloca i8, addrspace(2) + call void @llvm.lifetime.start.p2(i64 1, ptr addrspace(2) %a) + call void @llvm.lifetime.end.p2(i64 1, ptr addrspace(2) %a) + ret void +} + +define void @no_suffix1() { +; CHECK-LABEL: define void @no_suffix1() { +; CHECK-NEXT: [[A:%.*]] = alloca i8, align 1, addrspace(3) +; CHECK-NEXT: call void @llvm.lifetime.start.p3(ptr addrspace(3) [[A]]) +; CHECK-NEXT: call void @llvm.lifetime.end.p3(ptr addrspace(3) [[A]]) +; CHECK-NEXT: ret void +; + %a = alloca i8, addrspace(3) + call void @llvm.lifetime.start(i64 1, ptr addrspace(3) %a) + call void @llvm.lifetime.end(i64 1, ptr addrspace(3) %a) + ret void +} + +define void @no_suffix2() { +; CHECK-LABEL: define void @no_suffix2() { +; CHECK-NEXT: [[A:%.*]] = alloca i8, align 1, addrspace(4) +; CHECK-NEXT: call void @llvm.lifetime.start.p4(ptr addrspace(4) [[A]]) +; CHECK-NEXT: call void @llvm.lifetime.end.p4(ptr addrspace(4) [[A]]) +; CHECK-NEXT: ret void +; + %a = alloca i8, addrspace(4) + call void @llvm.lifetime.start(i64 1, ptr addrspace(4) %a) + call void @llvm.lifetime.end(i64 1, ptr addrspace(4) %a) + ret void +} + declare void @llvm.lifetime.start.p0(i64, ptr) declare void @llvm.lifetime.end.p0(i64, ptr) declare void @llvm.lifetime.start.p1(i64, ptr addrspace(1)) diff --git a/llvm/test/Assembler/autoupgrade-wasm-intrinsics.ll b/llvm/test/Assembler/autoupgrade-wasm-intrinsics.ll index 012fa1d..e54efa4 100644 --- a/llvm/test/Assembler/autoupgrade-wasm-intrinsics.ll +++ b/llvm/test/Assembler/autoupgrade-wasm-intrinsics.ll @@ -46,7 +46,10 @@ define <4 x float> @test_fms(<4 x float> %a, <4 x float> %b, <4 x float> %c) { ret <4 x float> %res } -declare <16 x i8> @llvm.wasm.laneselect.v16i8(<16 x i8>, <16 x i8>, <16 x i8>) +; This declaration is intentionally omitted to check that intrinsic upgrade +; also works without a declaration. +; declare <16 x i8> @llvm.wasm.laneselect.v16i8(<16 x i8>, <16 x i8>, <16 x i8>) + declare <8 x i16> @llvm.wasm.dot.i8x16.i7x16.signed(<16 x i8>, <16 x i8>) declare <4 x i32> @llvm.wasm.dot.i8x16.i7x16.add.signed(<16 x i8>, <16 x i8>, <4 x i32>) declare <4 x float> @llvm.wasm.fma.v4f32(<4 x float>, <4 x float>, <4 x float>) diff --git a/llvm/test/Assembler/dicompileunit-invalid-language-version.ll b/llvm/test/Assembler/dicompileunit-invalid-language-version.ll new file mode 100644 index 0000000..b3794ac --- /dev/null +++ b/llvm/test/Assembler/dicompileunit-invalid-language-version.ll @@ -0,0 +1,25 @@ +; RUN: split-file %s %t +; RUN: not llvm-as < %t/dw_lang_with_version.ll -disable-output 2>&1 | FileCheck %s --check-prefix=WRONG-ATTR +; RUN: not llvm-as < %t/overflow.ll -disable-output 2>&1 | FileCheck %s --check-prefix=OVERFLOW +; RUN: not llvm-as < %t/version_without_name.ll -disable-output 2>&1 | FileCheck %s --check-prefix=NO-NAME +; RUN: not llvm-as < %t/negative.ll -disable-output 2>&1 | FileCheck %s --check-prefix=NEGATIVE + +; WRONG-ATTR: error: 'sourceLanguageVersion' requires an associated 'sourceLanguageName' on !DICompileUnit +; OVERFLOW: error: value for 'sourceLanguageVersion' too large, limit is 4294967295 +; NEGATIVE: error: expected unsigned integer +; NO-NAME: error: missing one of 'language' or 'sourceLanguageName', required for !DICompileUnit + +;--- dw_lang_with_version.ll +!0 = distinct !DICompileUnit(language: DW_LANG_C, sourceLanguageVersion: 1, + file: !DIFile(filename: "", directory: "")) + +;--- overflow.ll +!0 = distinct !DICompileUnit(sourceLanguageName: DW_LNAME_C, sourceLanguageVersion: 4294967298) + +;--- negative.ll +!0 = distinct !DICompileUnit(sourceLanguageName: DW_LNAME_C, sourceLanguageVersion: -1, + file: !DIFile(filename: "", directory: "")) + +;--- version_without_name.ll +!0 = distinct !DICompileUnit(sourceLanguageVersion: 1, + file: !DIFile(filename: "", directory: "")) diff --git a/llvm/test/Assembler/implicit-intrinsic-declaration-invalid3.ll b/llvm/test/Assembler/implicit-intrinsic-declaration-invalid3.ll index ad5a96a..4caee57 100644 --- a/llvm/test/Assembler/implicit-intrinsic-declaration-invalid3.ll +++ b/llvm/test/Assembler/implicit-intrinsic-declaration-invalid3.ll @@ -2,7 +2,7 @@ ; Use of unknown intrinsic without declaration should be rejected. -; CHECK: error: use of undefined value '@llvm.foobar' +; CHECK: error: unknown intrinsic 'llvm.foobar' define void @test() { call i8 @llvm.foobar(i8 0, i16 1) ret void diff --git a/llvm/test/Bitcode/Inputs/compile-unit-no-versioned-language.bc b/llvm/test/Bitcode/Inputs/compile-unit-no-versioned-language.bc Binary files differnew file mode 100644 index 0000000..461a34d0 --- /dev/null +++ b/llvm/test/Bitcode/Inputs/compile-unit-no-versioned-language.bc diff --git a/llvm/test/Bitcode/dwarf-source-language-version.ll b/llvm/test/Bitcode/dwarf-source-language-version.ll new file mode 100644 index 0000000..311afd5 --- /dev/null +++ b/llvm/test/Bitcode/dwarf-source-language-version.ll @@ -0,0 +1,17 @@ +; RUN: llvm-as < %s | llvm-dis | llvm-as | llvm-dis | FileCheck %s --implicit-check-not "sourceLanguageVersion: 0" + +; CHECK: sourceLanguageVersion: 120 + +source_filename = "cu.cpp" +target triple = "arm64-apple-macosx" + +!llvm.dbg.cu = !{!0, !5} +!llvm.module.flags = !{!3, !4} + +!0 = distinct !DICompileUnit(sourceLanguageName: DW_LNAME_ObjC_plus_plus, sourceLanguageVersion: 120, file: !1, producer: "handwritten", isOptimized: false, runtimeVersion: 0, emissionKind: FullDebug, enums: !2, globals: !2, splitDebugInlining: false, nameTableKind: Apple, sysroot: "/") +!1 = !DIFile(filename: "cu.cpp", directory: "/tmp") +!2 = !{} +!3 = !{i32 7, !"Dwarf Version", i32 5} +!4 = !{i32 2, !"Debug Info Version", i32 3} +!5 = distinct !DICompileUnit(sourceLanguageName: DW_LNAME_ObjC_plus_plus, sourceLanguageVersion: 0, file: !6, producer: "handwritten", isOptimized: false, runtimeVersion: 0, emissionKind: FullDebug, enums: !2, globals: !2, splitDebugInlining: false, nameTableKind: Apple, sysroot: "/") +!6 = !DIFile(filename: "cu2.cpp", directory: "/tmp") diff --git a/llvm/test/Bitcode/upgrade-DICompileUnit-no-versioned-language.test b/llvm/test/Bitcode/upgrade-DICompileUnit-no-versioned-language.test new file mode 100644 index 0000000..9475f9b --- /dev/null +++ b/llvm/test/Bitcode/upgrade-DICompileUnit-no-versioned-language.test @@ -0,0 +1,21 @@ +; Test loading metadata which was not aware of versioned language names. +; +; RUN: llvm-dis -o - %p/Inputs/compile-unit-no-versioned-language.bc \ +; RUN: | FileCheck %s --implicit-check-not "sourceLanguageName" --implicit-check-not "sourceLanguageVersion" + +; Input bitcode file was compiled from following source on +; LLVM commit `fc22b58c25963ece6b041cadbdc931c2338955e4`: +; +; source_filename = "cu.cpp" +; target triple = "arm64-apple-macosx" +; +; !llvm.dbg.cu = !{!0} +; !llvm.module.flags = !{!3, !4} +; +; !0 = distinct !DICompileUnit(language: DW_LANG_ObjC, file: !1, producer: "handwritten", isOptimized: false, runtimeVersion: 0, emissionKind: FullDebug, enums: !2, globals: !2, splitDebugInlining: false, nameTableKind: Apple, sysroot: "/") +; !1 = !DIFile(filename: "cu.cpp", directory: "/tmp") +; !2 = !{} +; !3 = !{i32 7, !"Dwarf Version", i32 5} +; !4 = !{i32 2, !"Debug Info Version", i32 3} + +; CHECK: distinct !DICompileUnit(language: DW_LANG_ObjC, diff --git a/llvm/test/CodeGen/AArch64/GlobalISel/legalize-vector-compress.mir b/llvm/test/CodeGen/AArch64/GlobalISel/legalize-vector-compress.mir index cc75774..c2bf95c 100644 --- a/llvm/test/CodeGen/AArch64/GlobalISel/legalize-vector-compress.mir +++ b/llvm/test/CodeGen/AArch64/GlobalISel/legalize-vector-compress.mir @@ -15,8 +15,9 @@ body: | ; CHECK-NEXT: [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 0 ; CHECK-NEXT: [[C1:%[0-9]+]]:_(s64) = G_CONSTANT i64 0 ; CHECK-NEXT: [[EVEC:%[0-9]+]]:_(s32) = G_EXTRACT_VECTOR_ELT [[COPY]](<4 x s32>), [[C1]](s64) + ; CHECK-NEXT: [[COPY2:%[0-9]+]]:_(s64) = COPY [[C1]](s64) ; CHECK-NEXT: [[C2:%[0-9]+]]:_(s64) = G_CONSTANT i64 4 - ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s64) = G_MUL [[C1]], [[C2]] + ; CHECK-NEXT: [[MUL:%[0-9]+]]:_(s64) = G_MUL [[COPY2]], [[C2]] ; CHECK-NEXT: [[PTR_ADD:%[0-9]+]]:_(p0) = G_PTR_ADD [[FRAME_INDEX]], [[MUL]](s64) ; CHECK-NEXT: G_STORE [[EVEC]](s32), [[PTR_ADD]](p0) :: (store (s32)) ; CHECK-NEXT: [[EVEC1:%[0-9]+]]:_(s16) = G_EXTRACT_VECTOR_ELT [[COPY1]](<4 x s16>), [[C1]](s64) @@ -91,7 +92,8 @@ body: | ; CHECK-NEXT: [[LOAD:%[0-9]+]]:_(s32) = G_LOAD [[PTR_ADD]](p0) :: (load (s32)) ; CHECK-NEXT: [[C3:%[0-9]+]]:_(s64) = G_CONSTANT i64 0 ; CHECK-NEXT: [[EVEC:%[0-9]+]]:_(s32) = G_EXTRACT_VECTOR_ELT [[COPY]](<4 x s32>), [[C3]](s64) - ; CHECK-NEXT: [[MUL1:%[0-9]+]]:_(s64) = G_MUL [[C3]], [[C2]] + ; CHECK-NEXT: [[COPY3:%[0-9]+]]:_(s64) = COPY [[C3]](s64) + ; CHECK-NEXT: [[MUL1:%[0-9]+]]:_(s64) = G_MUL [[COPY3]], [[C2]] ; CHECK-NEXT: [[PTR_ADD1:%[0-9]+]]:_(p0) = G_PTR_ADD [[FRAME_INDEX]], [[MUL1]](s64) ; CHECK-NEXT: G_STORE [[EVEC]](s32), [[PTR_ADD1]](p0) :: (store (s32)) ; CHECK-NEXT: [[EVEC1:%[0-9]+]]:_(s16) = G_EXTRACT_VECTOR_ELT [[COPY1]](<4 x s16>), [[C3]](s64) diff --git a/llvm/test/CodeGen/AArch64/aarch64-matmul.ll b/llvm/test/CodeGen/AArch64/aarch64-matmul.ll index 649d0a9..e7e9ee7 100644 --- a/llvm/test/CodeGen/AArch64/aarch64-matmul.ll +++ b/llvm/test/CodeGen/AArch64/aarch64-matmul.ll @@ -1,41 +1,54 @@ -; RUN: llc -mtriple=aarch64-none-linux-gnu -mattr=+neon,+i8mm < %s -o -| FileCheck %s +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 6 +; RUN: llc -mtriple aarch64-none-linux-gnu -mattr=+neon,+i8mm < %s | FileCheck %s +; RUN: llc -mtriple aarch64-none-linux-gnu -mattr=+neon,+i8mm -global-isel < %s | FileCheck %s define <4 x i32> @smmla.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) { +; CHECK-LABEL: smmla.v4i32.v16i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: smmla v0.4s, v1.16b, v2.16b +; CHECK-NEXT: ret entry: -; CHECK-LABEL: smmla.v4i32.v16i8 -; CHECK: smmla v0.4s, v1.16b, v2.16b %vmmla1.i = tail call <4 x i32> @llvm.aarch64.neon.smmla.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) ret <4 x i32> %vmmla1.i } define <4 x i32> @ummla.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) { +; CHECK-LABEL: ummla.v4i32.v16i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: ummla v0.4s, v1.16b, v2.16b +; CHECK-NEXT: ret entry: -; CHECK-LABEL: ummla.v4i32.v16i8 -; CHECK: ummla v0.4s, v1.16b, v2.16b %vmmla1.i = tail call <4 x i32> @llvm.aarch64.neon.ummla.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) ret <4 x i32> %vmmla1.i } define <4 x i32> @usmmla.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) { +; CHECK-LABEL: usmmla.v4i32.v16i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: usmmla v0.4s, v1.16b, v2.16b +; CHECK-NEXT: ret entry: -; CHECK-LABEL: usmmla.v4i32.v16i8 -; CHECK: usmmla v0.4s, v1.16b, v2.16b %vusmmla1.i = tail call <4 x i32> @llvm.aarch64.neon.usmmla.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) #3 ret <4 x i32> %vusmmla1.i } define <2 x i32> @usdot.v2i32.v8i8(<2 x i32> %r, <8 x i8> %a, <8 x i8> %b) { +; CHECK-LABEL: usdot.v2i32.v8i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: usdot v0.2s, v1.8b, v2.8b +; CHECK-NEXT: ret entry: -; CHECK-LABEL: usdot.v2i32.v8i8 -; CHECK: usdot v0.2s, v1.8b, v2.8b %vusdot1.i = tail call <2 x i32> @llvm.aarch64.neon.usdot.v2i32.v8i8(<2 x i32> %r, <8 x i8> %a, <8 x i8> %b) ret <2 x i32> %vusdot1.i } define <2 x i32> @usdot_lane.v2i32.v8i8(<2 x i32> %r, <8 x i8> %a, <8 x i8> %b) { +; CHECK-LABEL: usdot_lane.v2i32.v8i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: // kill: def $d2 killed $d2 def $q2 +; CHECK-NEXT: usdot v0.2s, v1.8b, v2.4b[0] +; CHECK-NEXT: ret entry: -; CHECK-LABEL: usdot_lane.v2i32.v8i8 -; CHECK: usdot v0.2s, v1.8b, v2.4b[0] %0 = bitcast <8 x i8> %b to <2 x i32> %shuffle = shufflevector <2 x i32> %0, <2 x i32> undef, <2 x i32> zeroinitializer %1 = bitcast <2 x i32> %shuffle to <8 x i8> @@ -44,9 +57,12 @@ entry: } define <2 x i32> @sudot_lane.v2i32.v8i8(<2 x i32> %r, <8 x i8> %a, <8 x i8> %b) { +; CHECK-LABEL: sudot_lane.v2i32.v8i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: // kill: def $d2 killed $d2 def $q2 +; CHECK-NEXT: sudot v0.2s, v1.8b, v2.4b[0] +; CHECK-NEXT: ret entry: -; CHECK-LABEL: sudot_lane.v2i32.v8i8 -; CHECK: sudot v0.2s, v1.8b, v2.4b[0] %0 = bitcast <8 x i8> %b to <2 x i32> %shuffle = shufflevector <2 x i32> %0, <2 x i32> undef, <2 x i32> zeroinitializer %1 = bitcast <2 x i32> %shuffle to <8 x i8> @@ -55,9 +71,11 @@ entry: } define <2 x i32> @usdot_lane.v2i32.v16i8(<2 x i32> %r, <8 x i8> %a, <16 x i8> %b) { +; CHECK-LABEL: usdot_lane.v2i32.v16i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: usdot v0.2s, v1.8b, v2.4b[0] +; CHECK-NEXT: ret entry: -; CHECK-LABEL: usdot_lane.v2i32.v16i8 -; CHECK: usdot v0.2s, v1.8b, v2.4b[0] %0 = bitcast <16 x i8> %b to <4 x i32> %shuffle = shufflevector <4 x i32> %0, <4 x i32> undef, <2 x i32> zeroinitializer %1 = bitcast <2 x i32> %shuffle to <8 x i8> @@ -66,9 +84,11 @@ entry: } define <2 x i32> @sudot_lane.v2i32.v16i8(<2 x i32> %r, <8 x i8> %a, <16 x i8> %b) { +; CHECK-LABEL: sudot_lane.v2i32.v16i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: sudot v0.2s, v1.8b, v2.4b[0] +; CHECK-NEXT: ret entry: -; CHECK-LABEL: sudot_lane.v2i32.v16i8 -; CHECK: sudot v0.2s, v1.8b, v2.4b[0] %0 = bitcast <16 x i8> %b to <4 x i32> %shuffle = shufflevector <4 x i32> %0, <4 x i32> undef, <2 x i32> zeroinitializer %1 = bitcast <2 x i32> %shuffle to <8 x i8> @@ -77,17 +97,22 @@ entry: } define <4 x i32> @usdot.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) { +; CHECK-LABEL: usdot.v4i32.v16i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: usdot v0.4s, v1.16b, v2.16b +; CHECK-NEXT: ret entry: -; CHECK-LABEL: usdot.v4i32.v16i8 -; CHECK: usdot v0.4s, v1.16b, v2.16b %vusdot1.i = tail call <4 x i32> @llvm.aarch64.neon.usdot.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) #3 ret <4 x i32> %vusdot1.i } define <4 x i32> @usdot_lane.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <8 x i8> %b) { +; CHECK-LABEL: usdot_lane.v4i32.v16i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: // kill: def $d2 killed $d2 def $q2 +; CHECK-NEXT: usdot v0.4s, v1.16b, v2.4b[0] +; CHECK-NEXT: ret entry: -; CHECK-LABEL: usdot_lane.v4i32.v16i8 -; CHECK: usdot v0.4s, v1.16b, v2.4b[0] %0 = bitcast <8 x i8> %b to <2 x i32> %shuffle = shufflevector <2 x i32> %0, <2 x i32> undef, <4 x i32> zeroinitializer %1 = bitcast <4 x i32> %shuffle to <16 x i8> @@ -96,9 +121,12 @@ entry: } define <4 x i32> @sudot_lane.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <8 x i8> %b) { +; CHECK-LABEL: sudot_lane.v4i32.v16i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: // kill: def $d2 killed $d2 def $q2 +; CHECK-NEXT: sudot v0.4s, v1.16b, v2.4b[0] +; CHECK-NEXT: ret entry: -; CHECK-LABEL: sudot_lane.v4i32.v16i8 -; CHECK: sudot v0.4s, v1.16b, v2.4b[0] %0 = bitcast <8 x i8> %b to <2 x i32> %shuffle = shufflevector <2 x i32> %0, <2 x i32> undef, <4 x i32> zeroinitializer %1 = bitcast <4 x i32> %shuffle to <16 x i8> @@ -107,9 +135,11 @@ entry: } define <4 x i32> @usdot_laneq.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) { +; CHECK-LABEL: usdot_laneq.v4i32.v16i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: usdot v0.4s, v1.16b, v2.4b[0] +; CHECK-NEXT: ret entry: -; CHECK-LABEL: usdot_laneq.v4i32.v16i8 -; CHECK: usdot v0.4s, v1.16b, v2.4b[0] %0 = bitcast <16 x i8> %b to <4 x i32> %shuffle = shufflevector <4 x i32> %0, <4 x i32> undef, <4 x i32> zeroinitializer %1 = bitcast <4 x i32> %shuffle to <16 x i8> @@ -118,9 +148,11 @@ entry: } define <4 x i32> @sudot_laneq.v4i32.v16i8(<4 x i32> %r, <16 x i8> %a, <16 x i8> %b) { +; CHECK-LABEL: sudot_laneq.v4i32.v16i8: +; CHECK: // %bb.0: // %entry +; CHECK-NEXT: sudot v0.4s, v1.16b, v2.4b[0] +; CHECK-NEXT: ret entry: -; CHECK-LABEL: sudot_laneq.v4i32.v16i8 -; CHECK: sudot v0.4s, v1.16b, v2.4b[0] %0 = bitcast <16 x i8> %b to <4 x i32> %shuffle = shufflevector <4 x i32> %0, <4 x i32> undef, <4 x i32> zeroinitializer %1 = bitcast <4 x i32> %shuffle to <16 x i8> @@ -133,4 +165,3 @@ declare <4 x i32> @llvm.aarch64.neon.ummla.v4i32.v16i8(<4 x i32>, <16 x i8>, <16 declare <4 x i32> @llvm.aarch64.neon.usmmla.v4i32.v16i8(<4 x i32>, <16 x i8>, <16 x i8>) #2 declare <2 x i32> @llvm.aarch64.neon.usdot.v2i32.v8i8(<2 x i32>, <8 x i8>, <8 x i8>) #2 declare <4 x i32> @llvm.aarch64.neon.usdot.v4i32.v16i8(<4 x i32>, <16 x i8>, <16 x i8>) #2 - diff --git a/llvm/test/CodeGen/AArch64/aarch64-post-coalescer.mir b/llvm/test/CodeGen/AArch64/aarch64-post-coalescer.mir new file mode 100644 index 0000000..6540160 --- /dev/null +++ b/llvm/test/CodeGen/AArch64/aarch64-post-coalescer.mir @@ -0,0 +1,16 @@ +# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 6 +# RUN: llc -mtriple=aarch64 -mattr=+sme -run-pass=aarch64-post-coalescer-pass -o - %s | FileCheck %s + +--- +name: foo +machineFunctionInfo: + hasStreamingModeChanges: true +body: | + bb.0.entry: + ; CHECK-LABEL: name: foo + ; CHECK: $d0 = COPY undef %0:fpr64 + ; CHECK-NEXT: FAKE_USE implicit $d0 + %1:fpr64 = COALESCER_BARRIER_FPR64 undef %1 + $d0 = COPY %1 + FAKE_USE implicit $d0 +... diff --git a/llvm/test/CodeGen/AArch64/machine-sme-abi-find-insert-pt.mir b/llvm/test/CodeGen/AArch64/machine-sme-abi-find-insert-pt.mir new file mode 100644 index 0000000..3f174a6 --- /dev/null +++ b/llvm/test/CodeGen/AArch64/machine-sme-abi-find-insert-pt.mir @@ -0,0 +1,227 @@ +# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 6 +# RUN: llc -mtriple=aarch64 -mattr=+sve -mattr=+sme -run-pass=aarch64-machine-sme-abi -verify-machineinstrs %s -o - | FileCheck %s + +--- | + ; Test moving a state change to be before a $nzcv def + define void @move_before_nzcv_def() "aarch64_inout_za" { ret void } + + ; Test moving a state change to a point where $x0 is live + define void @move_to_x0_live() "aarch64_inout_za" { ret void } + + ; Test we don't move before a previous state change. + define void @do_not_move_before_prior_state_change() "aarch64_za_state_agnostic" { ret void } + + ; Test we don't move into a call sequence. + define void @do_not_move_into_call() "aarch64_inout_za" { ret void } + + declare void @clobber() + declare void @inout_call() "aarch64_inout_za" +... +--- +name: move_before_nzcv_def +tracksRegLiveness: true +isSSA: true +noVRegs: false +body: | + bb.0: + + ; CHECK-LABEL: name: move_before_nzcv_def + ; CHECK: [[RDSVLI_XI:%[0-9]+]]:gpr64 = RDSVLI_XI 1, implicit $vg + ; CHECK-NEXT: [[COPY:%[0-9]+]]:gpr64 = COPY $sp + ; CHECK-NEXT: [[MSUBXrrr:%[0-9]+]]:gpr64 = MSUBXrrr [[RDSVLI_XI]], [[RDSVLI_XI]], [[COPY]] + ; CHECK-NEXT: $sp = COPY [[MSUBXrrr]] + ; CHECK-NEXT: STPXi [[MSUBXrrr]], [[RDSVLI_XI]], %stack.0, 0 + ; CHECK-NEXT: [[ADDXri:%[0-9]+]]:gpr64sp = ADDXri %stack.0, 0, 0 + ; CHECK-NEXT: [[COPY1:%[0-9]+]]:gpr64 = COPY [[ADDXri]] + ; CHECK-NEXT: MSR 56965, [[COPY1]] + ; CHECK-NEXT: ADJCALLSTACKDOWN 0, 0, implicit-def dead $sp, implicit $sp + ; CHECK-NEXT: RequiresZASavePseudo + ; CHECK-NEXT: BL @clobber, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp + ; CHECK-NEXT: ADJCALLSTACKUP 0, 0, implicit-def dead $sp, implicit $sp + ; CHECK-NEXT: MSRpstatesvcrImm1 2, 1, implicit-def $nzcv + ; CHECK-NEXT: [[MRS:%[0-9]+]]:gpr64 = MRS 56965, implicit-def $nzcv + ; CHECK-NEXT: $x0 = ADDXri %stack.0, 0, 0 + ; CHECK-NEXT: RestoreZAPseudo [[MRS]], $x0, &__arm_tpidr2_restore, csr_aarch64_sme_abi_support_routines_preservemost_from_x0 + ; CHECK-NEXT: MSR 56965, $xzr + ; CHECK-NEXT: $nzcv = IMPLICIT_DEF + ; CHECK-NEXT: $zab0 = IMPLICIT_DEF + ; CHECK-NEXT: FAKE_USE $nzcv + ; CHECK-NEXT: RET_ReallyLR + ADJCALLSTACKDOWN 0, 0, implicit-def dead $sp, implicit $sp + RequiresZASavePseudo + BL @clobber, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp + ADJCALLSTACKUP 0, 0, implicit-def dead $sp, implicit $sp + + $nzcv = IMPLICIT_DEF + $zab0 = IMPLICIT_DEF + FAKE_USE $nzcv + + RET_ReallyLR +... +--- +name: move_to_x0_live +tracksRegLiveness: true +isSSA: true +noVRegs: false +body: | + bb.0: + + ; CHECK-LABEL: name: move_to_x0_live + ; CHECK: [[RDSVLI_XI:%[0-9]+]]:gpr64 = RDSVLI_XI 1, implicit $vg + ; CHECK-NEXT: [[COPY:%[0-9]+]]:gpr64 = COPY $sp + ; CHECK-NEXT: [[MSUBXrrr:%[0-9]+]]:gpr64 = MSUBXrrr [[RDSVLI_XI]], [[RDSVLI_XI]], [[COPY]] + ; CHECK-NEXT: $sp = COPY [[MSUBXrrr]] + ; CHECK-NEXT: STPXi [[MSUBXrrr]], [[RDSVLI_XI]], %stack.0, 0 + ; CHECK-NEXT: [[ADDXri:%[0-9]+]]:gpr64sp = ADDXri %stack.0, 0, 0 + ; CHECK-NEXT: [[COPY1:%[0-9]+]]:gpr64 = COPY [[ADDXri]] + ; CHECK-NEXT: MSR 56965, [[COPY1]] + ; CHECK-NEXT: ADJCALLSTACKDOWN 0, 0, implicit-def dead $sp, implicit $sp + ; CHECK-NEXT: RequiresZASavePseudo + ; CHECK-NEXT: BL @clobber, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp + ; CHECK-NEXT: ADJCALLSTACKUP 0, 0, implicit-def dead $sp, implicit $sp + ; CHECK-NEXT: $x0 = IMPLICIT_DEF + ; CHECK-NEXT: [[COPY2:%[0-9]+]]:gpr64 = COPY $x0 + ; CHECK-NEXT: MSRpstatesvcrImm1 2, 1, implicit-def $nzcv + ; CHECK-NEXT: [[MRS:%[0-9]+]]:gpr64 = MRS 56965, implicit-def $nzcv + ; CHECK-NEXT: $x0 = ADDXri %stack.0, 0, 0 + ; CHECK-NEXT: RestoreZAPseudo [[MRS]], $x0, &__arm_tpidr2_restore, csr_aarch64_sme_abi_support_routines_preservemost_from_x0 + ; CHECK-NEXT: MSR 56965, $xzr + ; CHECK-NEXT: $x0 = COPY [[COPY2]] + ; CHECK-NEXT: $nzcv = IMPLICIT_DEF + ; CHECK-NEXT: FAKE_USE $x0 + ; CHECK-NEXT: $zab0 = IMPLICIT_DEF + ; CHECK-NEXT: FAKE_USE $nzcv + ; CHECK-NEXT: RET_ReallyLR + ADJCALLSTACKDOWN 0, 0, implicit-def dead $sp, implicit $sp + RequiresZASavePseudo + BL @clobber, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp + ADJCALLSTACKUP 0, 0, implicit-def dead $sp, implicit $sp + + $x0 = IMPLICIT_DEF + + $nzcv = IMPLICIT_DEF + FAKE_USE $x0 + + $zab0 = IMPLICIT_DEF + FAKE_USE $nzcv + + RET_ReallyLR +... +--- +name: do_not_move_before_prior_state_change +tracksRegLiveness: true +isSSA: true +noVRegs: false +body: | + ; CHECK-LABEL: name: do_not_move_before_prior_state_change + ; CHECK: bb.0: + ; CHECK-NEXT: successors: %bb.1(0x40000000), %bb.2(0x40000000) + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: BL &__arm_sme_state_size, csr_aarch64_sme_abi_support_routines_preservemost_from_x1, implicit-def $lr, implicit $sp, implicit-def $x0 + ; CHECK-NEXT: [[COPY:%[0-9]+]]:gpr64 = COPY $x0 + ; CHECK-NEXT: $sp = SUBXrx64 $sp, [[COPY]], 24 + ; CHECK-NEXT: [[COPY1:%[0-9]+]]:gpr64 = COPY $sp + ; CHECK-NEXT: $nzcv = IMPLICIT_DEF + ; CHECK-NEXT: $zab0 = IMPLICIT_DEF + ; CHECK-NEXT: [[MRS:%[0-9]+]]:gpr64 = MRS 55824, implicit-def $nzcv, implicit $nzcv + ; CHECK-NEXT: $x0 = COPY [[COPY1]] + ; CHECK-NEXT: BL &__arm_sme_save, csr_aarch64_sme_abi_support_routines_preservemost_from_x1, implicit-def $lr, implicit $sp, implicit $x0 + ; CHECK-NEXT: MSR 55824, [[MRS]], implicit-def $nzcv + ; CHECK-NEXT: Bcc 2, %bb.1, implicit $nzcv + ; CHECK-NEXT: B %bb.2 + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: bb.1: + ; CHECK-NEXT: liveins: $nzcv + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: FAKE_USE $nzcv + ; CHECK-NEXT: ADJCALLSTACKDOWN 0, 0, implicit-def dead $sp, implicit $sp + ; CHECK-NEXT: RequiresZASavePseudo + ; CHECK-NEXT: BL @clobber, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp + ; CHECK-NEXT: ADJCALLSTACKUP 0, 0, implicit-def dead $sp, implicit $sp + ; CHECK-NEXT: $x0 = COPY [[COPY1]] + ; CHECK-NEXT: BL &__arm_sme_restore, csr_aarch64_sme_abi_support_routines_preservemost_from_x1, implicit-def $lr, implicit $sp, implicit $x0 + ; CHECK-NEXT: RET_ReallyLR + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: bb.2: + ; CHECK-NEXT: ADJCALLSTACKDOWN 0, 0, implicit-def dead $sp, implicit $sp + ; CHECK-NEXT: RequiresZASavePseudo + ; CHECK-NEXT: BL @clobber, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp + ; CHECK-NEXT: ADJCALLSTACKUP 0, 0, implicit-def dead $sp, implicit $sp + ; CHECK-NEXT: $x0 = COPY [[COPY1]] + ; CHECK-NEXT: BL &__arm_sme_restore, csr_aarch64_sme_abi_support_routines_preservemost_from_x1, implicit-def $lr, implicit $sp, implicit $x0 + ; CHECK-NEXT: RET_ReallyLR + bb.0: + successors: %bb.1, %bb.2 + + ; The insertion point can move before the $nzcv def (as that would require + ; moving before a $zab0 def -- that requires the ACTIVE state). + $nzcv = IMPLICIT_DEF + $zab0 = IMPLICIT_DEF + Bcc 2, %bb.1, implicit $nzcv + B %bb.2 + ; bb.1 and bb.2 both require ZA saved on entry (to force bb.0's exit bundle to + ; pick the LOCAL_SAVED state). + bb.1: + liveins: $nzcv + FAKE_USE $nzcv + + ADJCALLSTACKDOWN 0, 0, implicit-def dead $sp, implicit $sp + RequiresZASavePseudo + BL @clobber, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp + ADJCALLSTACKUP 0, 0, implicit-def dead $sp, implicit $sp + + RET_ReallyLR + bb.2: + ADJCALLSTACKDOWN 0, 0, implicit-def dead $sp, implicit $sp + RequiresZASavePseudo + BL @clobber, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp + ADJCALLSTACKUP 0, 0, implicit-def dead $sp, implicit $sp + + RET_ReallyLR +... +--- +name: do_not_move_into_call +tracksRegLiveness: true +isSSA: true +noVRegs: false +body: | + bb.0: + + ; CHECK-LABEL: name: do_not_move_into_call + ; CHECK: [[RDSVLI_XI:%[0-9]+]]:gpr64 = RDSVLI_XI 1, implicit $vg + ; CHECK-NEXT: [[COPY:%[0-9]+]]:gpr64 = COPY $sp + ; CHECK-NEXT: [[MSUBXrrr:%[0-9]+]]:gpr64 = MSUBXrrr [[RDSVLI_XI]], [[RDSVLI_XI]], [[COPY]] + ; CHECK-NEXT: $sp = COPY [[MSUBXrrr]] + ; CHECK-NEXT: STPXi [[MSUBXrrr]], [[RDSVLI_XI]], %stack.0, 0 + ; CHECK-NEXT: [[ADDXri:%[0-9]+]]:gpr64sp = ADDXri %stack.0, 0, 0 + ; CHECK-NEXT: [[COPY1:%[0-9]+]]:gpr64 = COPY [[ADDXri]] + ; CHECK-NEXT: MSR 56965, [[COPY1]] + ; CHECK-NEXT: ADJCALLSTACKDOWN 0, 0, implicit-def dead $sp, implicit $sp + ; CHECK-NEXT: RequiresZASavePseudo + ; CHECK-NEXT: BL @clobber, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp + ; CHECK-NEXT: $nzcv = IMPLICIT_DEF + ; CHECK-NEXT: ADJCALLSTACKUP 0, 0, implicit-def dead $sp, implicit $sp + ; CHECK-NEXT: [[MRS:%[0-9]+]]:gpr64 = MRS 55824, implicit-def $nzcv, implicit $nzcv + ; CHECK-NEXT: MSRpstatesvcrImm1 2, 1, implicit-def $nzcv + ; CHECK-NEXT: [[MRS1:%[0-9]+]]:gpr64 = MRS 56965, implicit-def $nzcv + ; CHECK-NEXT: $x0 = ADDXri %stack.0, 0, 0 + ; CHECK-NEXT: RestoreZAPseudo [[MRS1]], $x0, &__arm_tpidr2_restore, csr_aarch64_sme_abi_support_routines_preservemost_from_x0 + ; CHECK-NEXT: MSR 56965, $xzr + ; CHECK-NEXT: MSR 55824, [[MRS]], implicit-def $nzcv + ; CHECK-NEXT: $zab0 = IMPLICIT_DEF + ; CHECK-NEXT: FAKE_USE $nzcv + ; CHECK-NEXT: RET_ReallyLR + ADJCALLSTACKDOWN 0, 0, implicit-def dead $sp, implicit $sp + RequiresZASavePseudo + BL @clobber, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp + + ; This is artificial test where NZCV is def'd inside a call, so we can't + ; move the insert point before it's definition. + $nzcv = IMPLICIT_DEF + ADJCALLSTACKUP 0, 0, implicit-def dead $sp, implicit $sp + + $zab0 = IMPLICIT_DEF + FAKE_USE $nzcv + + RET_ReallyLR +... diff --git a/llvm/test/CodeGen/AArch64/mir-yaml-has-streaming-mode-changes.ll b/llvm/test/CodeGen/AArch64/mir-yaml-has-streaming-mode-changes.ll new file mode 100644 index 0000000..8f1fe5c --- /dev/null +++ b/llvm/test/CodeGen/AArch64/mir-yaml-has-streaming-mode-changes.ll @@ -0,0 +1,13 @@ +; RUN: llc -mtriple=aarch64 -mattr=+sme -stop-after=aarch64-isel < %s | FileCheck %s + +target triple = "aarch64" + +declare void @foo() "aarch64_pstate_sm_enabled" + +define dso_local void @bar() local_unnamed_addr { +; CHECK-LABEL: name: bar +; CHECK: hasStreamingModeChanges: true +entry: + tail call void @foo() "aarch64_pstate_sm_enabled" + ret void +} diff --git a/llvm/test/CodeGen/AArch64/sme-agnostic-za.ll b/llvm/test/CodeGen/AArch64/sme-agnostic-za.ll index e3007a3..e4f9efa 100644 --- a/llvm/test/CodeGen/AArch64/sme-agnostic-za.ll +++ b/llvm/test/CodeGen/AArch64/sme-agnostic-za.ll @@ -391,11 +391,9 @@ define void @agnostic_za_buffer_alloc_with_stack_probes() nounwind "aarch64_za_s ; CHECK-NEWLOWERING-NEXT: sub x19, x8, x0 ; CHECK-NEWLOWERING-NEXT: .LBB7_1: // =>This Inner Loop Header: Depth=1 ; CHECK-NEWLOWERING-NEXT: sub sp, sp, #16, lsl #12 // =65536 -; CHECK-NEWLOWERING-NEXT: cmp sp, x19 ; CHECK-NEWLOWERING-NEXT: mov x0, x19 -; CHECK-NEWLOWERING-NEXT: mrs x8, NZCV ; CHECK-NEWLOWERING-NEXT: bl __arm_sme_save -; CHECK-NEWLOWERING-NEXT: msr NZCV, x8 +; CHECK-NEWLOWERING-NEXT: cmp sp, x19 ; CHECK-NEWLOWERING-NEXT: b.le .LBB7_3 ; CHECK-NEWLOWERING-NEXT: // %bb.2: // in Loop: Header=BB7_1 Depth=1 ; CHECK-NEWLOWERING-NEXT: mov x0, x19 diff --git a/llvm/test/CodeGen/AArch64/sme-lazy-sve-nzcv-live.mir b/llvm/test/CodeGen/AArch64/sme-lazy-sve-nzcv-live.mir index 18764d5..9f33c06 100644 --- a/llvm/test/CodeGen/AArch64/sme-lazy-sve-nzcv-live.mir +++ b/llvm/test/CodeGen/AArch64/sme-lazy-sve-nzcv-live.mir @@ -62,14 +62,12 @@ body: | ; CHECK-NEXT: RequiresZASavePseudo ; CHECK-NEXT: BL @clobber, csr_aarch64_aapcs, implicit-def dead $lr, implicit $sp, implicit-def $sp ; CHECK-NEXT: ADJCALLSTACKUP 0, 0, implicit-def dead $sp, implicit $sp - ; CHECK-NEXT: [[SUBSWri:%[0-9]+]]:gpr32 = SUBSWri [[COPY1]], 101, 0, implicit-def $nzcv - ; CHECK-NEXT: [[MRS:%[0-9]+]]:gpr64 = MRS 55824, implicit-def $nzcv, implicit $nzcv ; CHECK-NEXT: MSRpstatesvcrImm1 2, 1, implicit-def $nzcv ; CHECK-NEXT: [[MRS1:%[0-9]+]]:gpr64 = MRS 56965, implicit-def $nzcv ; CHECK-NEXT: $x0 = ADDXri %stack.0, 0, 0 ; CHECK-NEXT: RestoreZAPseudo [[MRS1]], $x0, &__arm_tpidr2_restore, csr_aarch64_sme_abi_support_routines_preservemost_from_x0 ; CHECK-NEXT: MSR 56965, $xzr - ; CHECK-NEXT: MSR 55824, [[MRS]], implicit-def $nzcv + ; CHECK-NEXT: [[SUBSWri:%[0-9]+]]:gpr32 = SUBSWri [[COPY1]], 101, 0, implicit-def $nzcv ; CHECK-NEXT: Bcc 11, %bb.2, implicit $nzcv ; CHECK-NEXT: B %bb.1 ; CHECK-NEXT: {{ $}} @@ -116,16 +114,14 @@ body: | # CHECK-ASM-LABEL: cmp_branch # CHECK-ASM: msr TPIDR2_EL0, x10 # CHECK-ASM-NEXT: bl clobber -# CHECK-ASM-NEXT: cmp w20, #101 -# CHECK-ASM-NEXT: mrs x8, NZCV # CHECK-ASM-NEXT: smstart za -# CHECK-ASM-NEXT: mrs x9, TPIDR2_EL0 +# CHECK-ASM-NEXT: mrs x8, TPIDR2_EL0 # CHECK-ASM-NEXT: sub x0, x29, #16 -# CHECK-ASM-NEXT: cbnz x9, .LBB0_2 +# CHECK-ASM-NEXT: cbnz x8, .LBB0_2 # CHECK-ASM: bl __arm_tpidr2_restore # CHECK-ASM-NEXT: .LBB0_2: +# CHECK-ASM-NEXT: cmp w20, #101 # CHECK-ASM-NEXT: msr TPIDR2_EL0, xzr -# CHECK-ASM-NEXT: msr NZCV, x8 # CHECK-ASM-NEXT: b.lt .LBB0_4 # CHECK-ASM: bl inout_call # CHECK-ASM-NEXT: .LBB0_4: diff --git a/llvm/test/CodeGen/ARM/call-graph-section-addrtaken.ll b/llvm/test/CodeGen/ARM/call-graph-section-addrtaken.ll index a2d6ca9..972a470 100644 --- a/llvm/test/CodeGen/ARM/call-graph-section-addrtaken.ll +++ b/llvm/test/CodeGen/ARM/call-graph-section-addrtaken.ll @@ -27,7 +27,7 @@ entry: !1 = !{i64 0, !"_ZTSFivE.generalized"} !2 = !{i64 0, !"_ZTSFviE.generalized"} -; CHECK: .section .callgraph,"o",%progbits,.text +; CHECK: .section .llvm.callgraph,"o",%progbits,.text ;; Version ; CHECK-NEXT: .byte 0 ;; Flags -- Potential indirect target so LSB is set to 1. Other bits are 0. diff --git a/llvm/test/CodeGen/ARM/call-graph-section-assembly.ll b/llvm/test/CodeGen/ARM/call-graph-section-assembly.ll index bf5249e..ec8d5b8 100644 --- a/llvm/test/CodeGen/ARM/call-graph-section-assembly.ll +++ b/llvm/test/CodeGen/ARM/call-graph-section-assembly.ll @@ -1,8 +1,8 @@ ;; Test if temporary labels are generated for each indirect callsite. -;; Test if the .callgraph section contains the MD5 hash of callees' type (type id) +;; Test if the .llvm.callgraph section contains the MD5 hash of callees' type (type id) ;; is correctly paired with its corresponding temporary label generated for indirect ;; call sites annotated with !callee_type metadata. -;; Test if the .callgraph section contains unique direct callees. +;; Test if the .llvm.callgraph section contains unique direct callees. ; RUN: llc -mtriple=arm-unknown-linux --call-graph-section -o - < %s | FileCheck %s @@ -36,7 +36,7 @@ entry: !4 = !{!5} !5 = !{i64 0, !"_ZTSFPvS_E.generalized"} -; CHECK: .section .callgraph,"o",%progbits,.text +; CHECK: .section .llvm.callgraph,"o",%progbits,.text ;; Version ; CHECK-NEXT: .byte 0 ;; Flags diff --git a/llvm/test/CodeGen/ARM/call-graph-section-tailcall.ll b/llvm/test/CodeGen/ARM/call-graph-section-tailcall.ll index d577603..8036004 100644 --- a/llvm/test/CodeGen/ARM/call-graph-section-tailcall.ll +++ b/llvm/test/CodeGen/ARM/call-graph-section-tailcall.ll @@ -1,7 +1,7 @@ -;; Tests that we store the type identifiers in .callgraph section of the object file for tailcalls. +;; Tests that we store the type identifiers in .llvm.callgraph section of the object file for tailcalls. ; RUN: llc -mtriple=arm-unknown-linux --call-graph-section -filetype=obj -o - < %s | \ -; RUN: llvm-readelf -x .callgraph - | FileCheck %s +; RUN: llvm-readelf -x .llvm.callgraph - | FileCheck %s define i32 @check_tailcall(ptr %func, i8 %x) !type !0 { entry: @@ -27,7 +27,7 @@ declare !type !2 i32 @bar(i8 signext) !2 = !{i64 0, !"_ZTSFicE.generalized"} !3 = !{i64 0, !"_ZTSFiiE.generalized"} -; CHECK: Hex dump of section '.callgraph': +; CHECK: Hex dump of section '.llvm.callgraph': ; CHECK-NEXT: 0x00000000 00050000 00008e19 0b7f3326 e3000154 ; CHECK-NEXT: 0x00000010 86bc5981 4b8e3000 05100000 00a150b8 ;; Verify that the type id 0x308e4b8159bc8654 is in section. diff --git a/llvm/test/CodeGen/ARM/call-graph-section.ll b/llvm/test/CodeGen/ARM/call-graph-section.ll index 928a1067..167cc6f 100644 --- a/llvm/test/CodeGen/ARM/call-graph-section.ll +++ b/llvm/test/CodeGen/ARM/call-graph-section.ll @@ -1,7 +1,7 @@ -;; Tests that we store the type identifiers in .callgraph section of the object file. +;; Tests that we store the type identifiers in .llvm.callgraph section of the object file. ; RUN: llc -mtriple=arm-unknown-linux --call-graph-section -filetype=obj -o - < %s | \ -; RUN: llvm-readelf -x .callgraph - | FileCheck %s +; RUN: llvm-readelf -x .llvm.callgraph - | FileCheck %s declare !type !0 void @foo() @@ -31,7 +31,7 @@ entry: ;; Make sure following type IDs are in call graph section ;; 0x5eecb3e2444f731f, 0x814b8e305486bc59, 0xf897fd777ade6814 -; CHECK: Hex dump of section '.callgraph': +; CHECK: Hex dump of section '.llvm.callgraph': ; CHECK-NEXT: 0x00000000 00050000 00000000 00000000 00000324 ; CHECK-NEXT: 0x00000010 44f731f5 eecb3e54 86bc5981 4b8e307a ; CHECK-NEXT: 0x00000020 de6814f8 97fd77 diff --git a/llvm/test/CodeGen/ARM/nnan-fsub.ll b/llvm/test/CodeGen/ARM/nnan-fsub.ll index 0183908..78dd36f 100644 --- a/llvm/test/CodeGen/ARM/nnan-fsub.ll +++ b/llvm/test/CodeGen/ARM/nnan-fsub.ll @@ -1,18 +1,22 @@ -; RUN: llc -mcpu=cortex-a9 < %s | FileCheck -check-prefix=SAFE %s -; RUN: llc -mcpu=cortex-a9 --enable-no-nans-fp-math < %s | FileCheck -check-prefix=FAST %s +; RUN: llc -mcpu=cortex-a9 < %s | FileCheck %s target triple = "armv7-apple-ios" -; SAFE: test -; FAST: test +; CHECK-LABEL: test define float @test(float %x, float %y) { entry: -; SAFE: vmul.f32 -; SAFE: vsub.f32 -; FAST: mov r0, #0 +; CHECK: vmul.f32 +; CHECK-NEXT: vsub.f32 %0 = fmul float %x, %y %1 = fsub float %0, %0 ret float %1 } - +; CHECK-LABEL: test_nnan +define float @test_nnan(float %x, float %y) { +entry: +; CHECK: mov r0, #0 + %0 = fmul float %x, %y + %1 = fsub nnan float %0, %0 + ret float %1 +} diff --git a/llvm/test/CodeGen/DirectX/CBufferAccess/memcpy.ll b/llvm/test/CodeGen/DirectX/CBufferAccess/memcpy.ll index a78fdd5..f1486f97 100644 --- a/llvm/test/CodeGen/DirectX/CBufferAccess/memcpy.ll +++ b/llvm/test/CodeGen/DirectX/CBufferAccess/memcpy.ll @@ -74,7 +74,7 @@ entry: ; CHECK: [[UPTO1:%.*]] = insertelement <3 x double> [[UPTO0]], double [[Y]], i32 1 ; CHECK: [[UPTO2:%.*]] = insertelement <3 x double> [[UPTO1]], double [[Z]], i32 2 ; CHECK: [[DEST:%.*]] = getelementptr inbounds i8, ptr [[A2_COPY:%.*]], i32 0 -; CHECK: store <3 x double> [[UPTO2]], ptr [[DEST]], align 32 +; CHECK: store <3 x double> [[UPTO2]], ptr [[DEST]], align 8 ; CHECK: [[LOAD:%.*]] = call { double, double } @llvm.dx.resource.load.cbufferrow.2.{{.*}}(target("dx.CBuffer", {{.*}})) [[CB]], i32 5) ; CHECK: [[X:%.*]] = extractvalue { double, double } [[LOAD]], 0 ; CHECK: [[Y:%.*]] = extractvalue { double, double } [[LOAD]], 1 @@ -83,9 +83,9 @@ entry: ; CHECK: [[UPTO0:%.*]] = insertelement <3 x double> poison, double [[X]], i32 0 ; CHECK: [[UPTO1:%.*]] = insertelement <3 x double> [[UPTO0]], double [[Y]], i32 1 ; CHECK: [[UPTO2:%.*]] = insertelement <3 x double> [[UPTO1]], double [[Z]], i32 2 -; CHECK: [[DEST:%.*]] = getelementptr inbounds i8, ptr [[A2_COPY]], i32 32 -; CHECK: store <3 x double> [[UPTO2]], ptr [[DEST]], align 32 - call void @llvm.memcpy.p0.p2.i32(ptr align 32 %a2.copy, ptr addrspace(2) align 32 @a2, i32 64, i1 false) +; CHECK: [[DEST:%.*]] = getelementptr inbounds i8, ptr [[A2_COPY]], i32 24 +; CHECK: store <3 x double> [[UPTO2]], ptr [[DEST]], align 8 + call void @llvm.memcpy.p0.p2.i32(ptr align 32 %a2.copy, ptr addrspace(2) align 32 @a2, i32 48, i1 false) ; CHECK: [[CB:%.*]] = load target("dx.CBuffer", {{.*}})), ptr @CB.cb, align 4 ; CHECK: [[LOAD:%.*]] = call { half, half, half, half, half, half, half, half } @llvm.dx.resource.load.cbufferrow.8.{{.*}}(target("dx.CBuffer", {{.*}})) [[CB]], i32 7) diff --git a/llvm/test/CodeGen/DirectX/ContainerData/PSVResources.ll b/llvm/test/CodeGen/DirectX/ContainerData/PSVResources.ll index bea0310..70224fc 100644 --- a/llvm/test/CodeGen/DirectX/ContainerData/PSVResources.ll +++ b/llvm/test/CodeGen/DirectX/ContainerData/PSVResources.ll @@ -94,6 +94,18 @@ define void @main() #0 { %uav2_2 = call target("dx.TypedBuffer", <4 x float>, 1, 0, 0) @llvm.dx.resource.handlefrombinding.tdx.TypedBuffer_f32_1_0( i32 4, i32 0, i32 10, i32 5, ptr null) + + ; RWBuffer<float4> UnboundedArray[] : register(u10, space5) +; CHECK: - Type: UAVTyped +; CHECK: Space: 5 +; CHECK: LowerBound: 10 +; CHECK: UpperBound: 4294967295 +; CHECK: Kind: TypedBuffer +; CHECK: Flags: +; CHECK: UsedByAtomic64: false + ; RWBuffer<float4> Buf = BufferArray[100]; + %uav3 = call target("dx.TypedBuffer", <4 x float>, 1, 0, 0) + @llvm.dx.resource.handlefrombinding(i32 5, i32 10, i32 -1, i32 100, ptr null) ret void } diff --git a/llvm/test/CodeGen/DirectX/Metadata/cbuffer_metadata.ll b/llvm/test/CodeGen/DirectX/Metadata/cbuffer_metadata.ll index 7ba2ed2..f1d28e2 100644 --- a/llvm/test/CodeGen/DirectX/Metadata/cbuffer_metadata.ll +++ b/llvm/test/CodeGen/DirectX/Metadata/cbuffer_metadata.ll @@ -19,11 +19,11 @@ target triple = "dxil-pc-shadermodel6.6-compute" ; PRINT:; Resource Bindings: ; PRINT-NEXT:; -; PRINT-NEXT:; Name Type Format Dim ID HLSL Bind Count -; PRINT-NEXT:; ------------------------------ ---------- ------- ----------- ------- -------------- ------ -; PRINT-NEXT:; CB1 cbuffer NA NA CB0 cb0 1 -; PRINT-NEXT:; CB2 cbuffer NA NA CB1 cb1 1 -; PRINT-NEXT:; MyConstants cbuffer NA NA CB2 cb5,space15 1 +; PRINT-NEXT:; Name Type Format Dim ID HLSL Bind Count +; PRINT-NEXT:; ---- +; PRINT-NEXT:; CB1 cbuffer NA NA CB0 cb0 1 +; PRINT-NEXT:; CB2 cbuffer NA NA CB1 cb1 1 +; PRINT-NEXT:; MyConstants cbuffer NA NA CB2 cb5,space15 1 define void @test() #0 { diff --git a/llvm/test/CodeGen/LoongArch/lasx/vselect.ll b/llvm/test/CodeGen/LoongArch/lasx/vselect.ll index bf31ccb..559cc53 100644 --- a/llvm/test/CodeGen/LoongArch/lasx/vselect.ll +++ b/llvm/test/CodeGen/LoongArch/lasx/vselect.ll @@ -32,6 +32,40 @@ define void @select_v32i8(ptr %res, ptr %a0, ptr %a1) nounwind { ret void } +define void @select_v32i8_1(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v32i8_1: +; CHECK: # %bb.0: +; CHECK-NEXT: xvld $xr0, $a1, 0 +; CHECK-NEXT: xvld $xr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI2_0) +; CHECK-NEXT: xvld $xr2, $a1, %pc_lo12(.LCPI2_0) +; CHECK-NEXT: xvbitsel.v $xr0, $xr1, $xr0, $xr2 +; CHECK-NEXT: xvst $xr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <32 x i8>, ptr %a0 + %v1 = load <32 x i8>, ptr %a1 + %sel = select <32 x i1> <i1 true, i1 true, i1 true, i1 true, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false>, <32 x i8> %v0, <32 x i8> %v1 + store <32 x i8> %sel, ptr %res + ret void +} + +define void @select_v32i8_2(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v32i8_2: +; CHECK: # %bb.0: +; CHECK-NEXT: xvld $xr0, $a1, 0 +; CHECK-NEXT: xvld $xr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI3_0) +; CHECK-NEXT: xvld $xr2, $a1, %pc_lo12(.LCPI3_0) +; CHECK-NEXT: xvbitsel.v $xr0, $xr1, $xr0, $xr2 +; CHECK-NEXT: xvst $xr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <32 x i8>, ptr %a0 + %v1 = load <32 x i8>, ptr %a1 + %sel = select <32 x i1> <i1 false, i1 true, i1 true, i1 true, i1 false, i1 true, i1 true, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true>, <32 x i8> %v0, <32 x i8> %v1 + store <32 x i8> %sel, ptr %res + ret void +} + define void @select_v16i16(ptr %res, ptr %a0, ptr %a1) nounwind { ; CHECK-LABEL: select_v16i16: ; CHECK: # %bb.0: @@ -49,6 +83,40 @@ define void @select_v16i16(ptr %res, ptr %a0, ptr %a1) nounwind { ret void } +define void @select_v16i16_1(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v16i16_1: +; CHECK: # %bb.0: +; CHECK-NEXT: xvld $xr0, $a1, 0 +; CHECK-NEXT: xvld $xr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI5_0) +; CHECK-NEXT: xvld $xr2, $a1, %pc_lo12(.LCPI5_0) +; CHECK-NEXT: xvbitsel.v $xr0, $xr1, $xr0, $xr2 +; CHECK-NEXT: xvst $xr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <16 x i16>, ptr %a0 + %v1 = load <16 x i16>, ptr %a1 + %sel = select <16 x i1> <i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>, <16 x i16> %v0, <16 x i16> %v1 + store <16 x i16> %sel, ptr %res + ret void +} + +define void @select_v16i16_2(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v16i16_2: +; CHECK: # %bb.0: +; CHECK-NEXT: xvld $xr0, $a1, 0 +; CHECK-NEXT: xvld $xr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI6_0) +; CHECK-NEXT: xvld $xr2, $a1, %pc_lo12(.LCPI6_0) +; CHECK-NEXT: xvbitsel.v $xr0, $xr1, $xr0, $xr2 +; CHECK-NEXT: xvst $xr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <16 x i16>, ptr %a0 + %v1 = load <16 x i16>, ptr %a1 + %sel = select <16 x i1> <i1 false, i1 true, i1 true, i1 true, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 false, i1 true, i1 true, i1 true, i1 false>, <16 x i16> %v0, <16 x i16> %v1 + store <16 x i16> %sel, ptr %res + ret void +} + define void @select_v8i32(ptr %res, ptr %a0, ptr %a1) nounwind { ; CHECK-LABEL: select_v8i32: ; CHECK: # %bb.0: @@ -65,19 +133,70 @@ define void @select_v8i32(ptr %res, ptr %a0, ptr %a1) nounwind { ret void } +define void @select_v8i32_1(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v8i32_1: +; CHECK: # %bb.0: +; CHECK-NEXT: xvld $xr0, $a1, 0 +; CHECK-NEXT: xvld $xr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI8_0) +; CHECK-NEXT: xvld $xr2, $a1, %pc_lo12(.LCPI8_0) +; CHECK-NEXT: xvbitsel.v $xr0, $xr1, $xr0, $xr2 +; CHECK-NEXT: xvst $xr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <8 x i32>, ptr %a0 + %v1 = load <8 x i32>, ptr %a1 + %sel = select <8 x i1> <i1 true, i1 true, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false>, <8 x i32> %v0, <8 x i32> %v1 + store <8 x i32> %sel, ptr %res + ret void +} + +define void @select_v8f32(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v8f32: +; CHECK: # %bb.0: +; CHECK-NEXT: xvld $xr0, $a1, 0 +; CHECK-NEXT: xvld $xr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI9_0) +; CHECK-NEXT: xvld $xr2, $a1, %pc_lo12(.LCPI9_0) +; CHECK-NEXT: xvbitsel.v $xr0, $xr1, $xr0, $xr2 +; CHECK-NEXT: xvst $xr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <8 x float>, ptr %a0 + %v1 = load <8 x float>, ptr %a1 + %sel = select <8 x i1> <i1 false, i1 false, i1 true, i1 false, i1 false, i1 false, i1 true, i1 false>, <8 x float> %v0, <8 x float> %v1 + store <8 x float> %sel, ptr %res + ret void +} + define void @select_v4i64(ptr %res, ptr %a0, ptr %a1) nounwind { ; CHECK-LABEL: select_v4i64: ; CHECK: # %bb.0: ; CHECK-NEXT: xvld $xr0, $a1, 0 ; CHECK-NEXT: xvld $xr1, $a2, 0 -; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI4_0) -; CHECK-NEXT: xvld $xr2, $a1, %pc_lo12(.LCPI4_0) +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI10_0) +; CHECK-NEXT: xvld $xr2, $a1, %pc_lo12(.LCPI10_0) ; CHECK-NEXT: xvbitsel.v $xr0, $xr1, $xr0, $xr2 ; CHECK-NEXT: xvst $xr0, $a0, 0 ; CHECK-NEXT: ret %v0 = load <4 x i64>, ptr %a0 %v1 = load <4 x i64>, ptr %a1 - %sel = select <4 x i1> <i1 true, i1 false, i1 true, i1 false>, <4 x i64> %v0, <4 x i64> %v1 + %sel = select <4 x i1> <i1 true, i1 false, i1 false, i1 false>, <4 x i64> %v0, <4 x i64> %v1 store <4 x i64> %sel, ptr %res ret void } + +define void @select_v4f64(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v4f64: +; CHECK: # %bb.0: +; CHECK-NEXT: xvld $xr0, $a1, 0 +; CHECK-NEXT: xvld $xr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI11_0) +; CHECK-NEXT: xvld $xr2, $a1, %pc_lo12(.LCPI11_0) +; CHECK-NEXT: xvbitsel.v $xr0, $xr1, $xr0, $xr2 +; CHECK-NEXT: xvst $xr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <4 x double>, ptr %a0 + %v1 = load <4 x double>, ptr %a1 + %sel = select <4 x i1> <i1 true, i1 false, i1 true, i1 false>, <4 x double> %v0, <4 x double> %v1 + store <4 x double> %sel, ptr %res + ret void +} diff --git a/llvm/test/CodeGen/LoongArch/lsx/vselect.ll b/llvm/test/CodeGen/LoongArch/lsx/vselect.ll index 8f25a6b..25c4f09 100644 --- a/llvm/test/CodeGen/LoongArch/lsx/vselect.ll +++ b/llvm/test/CodeGen/LoongArch/lsx/vselect.ll @@ -16,6 +16,20 @@ define void @select_v16i8_imm(ptr %res, ptr %a0) nounwind { ret void } +define void @select_v16i8_imm_1(ptr %res, ptr %a0) nounwind { +; CHECK-LABEL: select_v16i8_imm_1: +; CHECK: # %bb.0: +; CHECK-NEXT: vld $vr0, $a1, 0 +; CHECK-NEXT: vrepli.h $vr1, -256 +; CHECK-NEXT: vbitseli.b $vr1, $vr0, 1 +; CHECK-NEXT: vst $vr1, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <16 x i8>, ptr %a0 + %sel = select <16 x i1> <i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true, i1 false, i1 true>, <16 x i8> <i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>, <16 x i8> %v0 + store <16 x i8> %sel, ptr %res + ret void +} + define void @select_v16i8(ptr %res, ptr %a0, ptr %a1) nounwind { ; CHECK-LABEL: select_v16i8: ; CHECK: # %bb.0: @@ -32,6 +46,40 @@ define void @select_v16i8(ptr %res, ptr %a0, ptr %a1) nounwind { ret void } +define void @select_v16i8_1(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v16i8_1: +; CHECK: # %bb.0: +; CHECK-NEXT: vld $vr0, $a1, 0 +; CHECK-NEXT: vld $vr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI3_0) +; CHECK-NEXT: vld $vr2, $a1, %pc_lo12(.LCPI3_0) +; CHECK-NEXT: vbitsel.v $vr0, $vr1, $vr0, $vr2 +; CHECK-NEXT: vst $vr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <16 x i8>, ptr %a0 + %v1 = load <16 x i8>, ptr %a1 + %sel = select <16 x i1> <i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>, <16 x i8> %v0, <16 x i8> %v1 + store <16 x i8> %sel, ptr %res + ret void +} + +define void @select_v16i8_2(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v16i8_2: +; CHECK: # %bb.0: +; CHECK-NEXT: vld $vr0, $a1, 0 +; CHECK-NEXT: vld $vr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI4_0) +; CHECK-NEXT: vld $vr2, $a1, %pc_lo12(.LCPI4_0) +; CHECK-NEXT: vbitsel.v $vr0, $vr1, $vr0, $vr2 +; CHECK-NEXT: vst $vr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <16 x i8>, ptr %a0 + %v1 = load <16 x i8>, ptr %a1 + %sel = select <16 x i1> <i1 true, i1 true, i1 false, i1 false, i1 false, i1 false, i1 false, i1 true, i1 false, i1 true, i1 true, i1 true, i1 true, i1 true, i1 false, i1 false>, <16 x i8> %v0, <16 x i8> %v1 + store <16 x i8> %sel, ptr %res + ret void +} + define void @select_v8i16(ptr %res, ptr %a0, ptr %a1) nounwind { ; CHECK-LABEL: select_v8i16: ; CHECK: # %bb.0: @@ -49,6 +97,40 @@ define void @select_v8i16(ptr %res, ptr %a0, ptr %a1) nounwind { ret void } +define void @select_v8i16_1(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v8i16_1: +; CHECK: # %bb.0: +; CHECK-NEXT: vld $vr0, $a1, 0 +; CHECK-NEXT: vld $vr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI6_0) +; CHECK-NEXT: vld $vr2, $a1, %pc_lo12(.LCPI6_0) +; CHECK-NEXT: vbitsel.v $vr0, $vr1, $vr0, $vr2 +; CHECK-NEXT: vst $vr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <8 x i16>, ptr %a0 + %v1 = load <8 x i16>, ptr %a1 + %sel = select <8 x i1> <i1 true, i1 true, i1 true, i1 true, i1 false, i1 false, i1 false, i1 false>, <8 x i16> %v0, <8 x i16> %v1 + store <8 x i16> %sel, ptr %res + ret void +} + +define void @select_v8i16_2(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v8i16_2: +; CHECK: # %bb.0: +; CHECK-NEXT: vld $vr0, $a1, 0 +; CHECK-NEXT: vld $vr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI7_0) +; CHECK-NEXT: vld $vr2, $a1, %pc_lo12(.LCPI7_0) +; CHECK-NEXT: vbitsel.v $vr0, $vr1, $vr0, $vr2 +; CHECK-NEXT: vst $vr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <8 x i16>, ptr %a0 + %v1 = load <8 x i16>, ptr %a1 + %sel = select <8 x i1> <i1 false, i1 false, i1 true, i1 true, i1 false, i1 false, i1 false, i1 false>, <8 x i16> %v0, <8 x i16> %v1 + store <8 x i16> %sel, ptr %res + ret void +} + define void @select_v4i32(ptr %res, ptr %a0, ptr %a1) nounwind { ; CHECK-LABEL: select_v4i32: ; CHECK: # %bb.0: @@ -65,13 +147,47 @@ define void @select_v4i32(ptr %res, ptr %a0, ptr %a1) nounwind { ret void } +define void @select_v4i32_1(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v4i32_1: +; CHECK: # %bb.0: +; CHECK-NEXT: vld $vr0, $a1, 0 +; CHECK-NEXT: vld $vr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI9_0) +; CHECK-NEXT: vld $vr2, $a1, %pc_lo12(.LCPI9_0) +; CHECK-NEXT: vbitsel.v $vr0, $vr1, $vr0, $vr2 +; CHECK-NEXT: vst $vr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <4 x i32>, ptr %a0 + %v1 = load <4 x i32>, ptr %a1 + %sel = select <4 x i1> <i1 true, i1 true, i1 false, i1 false>, <4 x i32> %v0, <4 x i32> %v1 + store <4 x i32> %sel, ptr %res + ret void +} + +define void @select_v4f32(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v4f32: +; CHECK: # %bb.0: +; CHECK-NEXT: vld $vr0, $a1, 0 +; CHECK-NEXT: vld $vr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI10_0) +; CHECK-NEXT: vld $vr2, $a1, %pc_lo12(.LCPI10_0) +; CHECK-NEXT: vbitsel.v $vr0, $vr1, $vr0, $vr2 +; CHECK-NEXT: vst $vr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <4 x float>, ptr %a0 + %v1 = load <4 x float>, ptr %a1 + %sel = select <4 x i1> <i1 false, i1 true, i1 true, i1 true>, <4 x float> %v0, <4 x float> %v1 + store <4 x float> %sel, ptr %res + ret void +} + define void @select_v2i64(ptr %res, ptr %a0, ptr %a1) nounwind { ; CHECK-LABEL: select_v2i64: ; CHECK: # %bb.0: ; CHECK-NEXT: vld $vr0, $a1, 0 ; CHECK-NEXT: vld $vr1, $a2, 0 -; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI4_0) -; CHECK-NEXT: vld $vr2, $a1, %pc_lo12(.LCPI4_0) +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI11_0) +; CHECK-NEXT: vld $vr2, $a1, %pc_lo12(.LCPI11_0) ; CHECK-NEXT: vbitsel.v $vr0, $vr1, $vr0, $vr2 ; CHECK-NEXT: vst $vr0, $a0, 0 ; CHECK-NEXT: ret @@ -81,3 +197,20 @@ define void @select_v2i64(ptr %res, ptr %a0, ptr %a1) nounwind { store <2 x i64> %sel, ptr %res ret void } + +define void @select_v2f64(ptr %res, ptr %a0, ptr %a1) nounwind { +; CHECK-LABEL: select_v2f64: +; CHECK: # %bb.0: +; CHECK-NEXT: vld $vr0, $a1, 0 +; CHECK-NEXT: vld $vr1, $a2, 0 +; CHECK-NEXT: pcalau12i $a1, %pc_hi20(.LCPI12_0) +; CHECK-NEXT: vld $vr2, $a1, %pc_lo12(.LCPI12_0) +; CHECK-NEXT: vbitsel.v $vr0, $vr1, $vr0, $vr2 +; CHECK-NEXT: vst $vr0, $a0, 0 +; CHECK-NEXT: ret + %v0 = load <2 x double>, ptr %a0 + %v1 = load <2 x double>, ptr %a1 + %sel = select <2 x i1> <i1 false, i1 true>, <2 x double> %v0, <2 x double> %v1 + store <2 x double> %sel, ptr %res + ret void +} diff --git a/llvm/test/CodeGen/PowerPC/fmf-propagation.ll b/llvm/test/CodeGen/PowerPC/fmf-propagation.ll index e71f59c..cad684e 100644 --- a/llvm/test/CodeGen/PowerPC/fmf-propagation.ll +++ b/llvm/test/CodeGen/PowerPC/fmf-propagation.ll @@ -325,24 +325,21 @@ define float @sqrt_afn_ieee(float %x) #0 { ; ; GLOBAL-LABEL: sqrt_afn_ieee: ; GLOBAL: # %bb.0: -; GLOBAL-NEXT: addis 3, 2, .LCPI11_1@toc@ha -; GLOBAL-NEXT: xsabsdp 0, 1 -; GLOBAL-NEXT: lfs 2, .LCPI11_1@toc@l(3) -; GLOBAL-NEXT: fcmpu 0, 0, 2 -; GLOBAL-NEXT: xxlxor 0, 0, 0 -; GLOBAL-NEXT: blt 0, .LBB11_2 -; GLOBAL-NEXT: # %bb.1: ; GLOBAL-NEXT: xsrsqrtesp 0, 1 ; GLOBAL-NEXT: vspltisw 2, -3 ; GLOBAL-NEXT: addis 3, 2, .LCPI11_0@toc@ha -; GLOBAL-NEXT: xvcvsxwdp 2, 34 -; GLOBAL-NEXT: xsmulsp 1, 1, 0 -; GLOBAL-NEXT: xsmaddasp 2, 1, 0 +; GLOBAL-NEXT: xvcvsxwdp 3, 34 +; GLOBAL-NEXT: xsmulsp 2, 1, 0 +; GLOBAL-NEXT: xsabsdp 1, 1 +; GLOBAL-NEXT: xsmaddasp 3, 2, 0 ; GLOBAL-NEXT: lfs 0, .LCPI11_0@toc@l(3) -; GLOBAL-NEXT: xsmulsp 0, 1, 0 -; GLOBAL-NEXT: xsmulsp 0, 0, 2 -; GLOBAL-NEXT: .LBB11_2: -; GLOBAL-NEXT: fmr 1, 0 +; GLOBAL-NEXT: addis 3, 2, .LCPI11_1@toc@ha +; GLOBAL-NEXT: xsmulsp 0, 2, 0 +; GLOBAL-NEXT: lfs 2, .LCPI11_1@toc@l(3) +; GLOBAL-NEXT: xssubsp 1, 1, 2 +; GLOBAL-NEXT: xxlxor 2, 2, 2 +; GLOBAL-NEXT: xsmulsp 0, 0, 3 +; GLOBAL-NEXT: fsel 1, 1, 0, 2 ; GLOBAL-NEXT: blr %rt = call afn ninf float @llvm.sqrt.f32(float %x) ret float %rt @@ -393,21 +390,19 @@ define float @sqrt_afn_preserve_sign(float %x) #1 { ; ; GLOBAL-LABEL: sqrt_afn_preserve_sign: ; GLOBAL: # %bb.0: -; GLOBAL-NEXT: xxlxor 0, 0, 0 -; GLOBAL-NEXT: fcmpu 0, 1, 0 -; GLOBAL-NEXT: beq 0, .LBB13_2 -; GLOBAL-NEXT: # %bb.1: ; GLOBAL-NEXT: xsrsqrtesp 0, 1 ; GLOBAL-NEXT: vspltisw 2, -3 ; GLOBAL-NEXT: addis 3, 2, .LCPI13_0@toc@ha -; GLOBAL-NEXT: xvcvsxwdp 2, 34 -; GLOBAL-NEXT: xsmulsp 1, 1, 0 -; GLOBAL-NEXT: xsmaddasp 2, 1, 0 +; GLOBAL-NEXT: xvcvsxwdp 3, 34 +; GLOBAL-NEXT: xsmulsp 2, 1, 0 +; GLOBAL-NEXT: xsmaddasp 3, 2, 0 ; GLOBAL-NEXT: lfs 0, .LCPI13_0@toc@l(3) -; GLOBAL-NEXT: xsmulsp 0, 1, 0 -; GLOBAL-NEXT: xsmulsp 0, 0, 2 -; GLOBAL-NEXT: .LBB13_2: -; GLOBAL-NEXT: fmr 1, 0 +; GLOBAL-NEXT: xsmulsp 0, 2, 0 +; GLOBAL-NEXT: xxlxor 2, 2, 2 +; GLOBAL-NEXT: xsmulsp 0, 0, 3 +; GLOBAL-NEXT: fsel 2, 1, 2, 0 +; GLOBAL-NEXT: xsnegdp 1, 1 +; GLOBAL-NEXT: fsel 1, 1, 2, 0 ; GLOBAL-NEXT: blr %rt = call afn ninf float @llvm.sqrt.f32(float %x) ret float %rt @@ -462,24 +457,21 @@ define float @sqrt_fast_ieee(float %x) #0 { ; ; GLOBAL-LABEL: sqrt_fast_ieee: ; GLOBAL: # %bb.0: -; GLOBAL-NEXT: addis 3, 2, .LCPI15_1@toc@ha -; GLOBAL-NEXT: xsabsdp 0, 1 -; GLOBAL-NEXT: lfs 2, .LCPI15_1@toc@l(3) -; GLOBAL-NEXT: fcmpu 0, 0, 2 -; GLOBAL-NEXT: xxlxor 0, 0, 0 -; GLOBAL-NEXT: blt 0, .LBB15_2 -; GLOBAL-NEXT: # %bb.1: ; GLOBAL-NEXT: xsrsqrtesp 0, 1 ; GLOBAL-NEXT: vspltisw 2, -3 ; GLOBAL-NEXT: addis 3, 2, .LCPI15_0@toc@ha -; GLOBAL-NEXT: xvcvsxwdp 2, 34 -; GLOBAL-NEXT: xsmulsp 1, 1, 0 -; GLOBAL-NEXT: xsmaddasp 2, 1, 0 +; GLOBAL-NEXT: xvcvsxwdp 3, 34 +; GLOBAL-NEXT: xsmulsp 2, 1, 0 +; GLOBAL-NEXT: xsabsdp 1, 1 +; GLOBAL-NEXT: xsmaddasp 3, 2, 0 ; GLOBAL-NEXT: lfs 0, .LCPI15_0@toc@l(3) -; GLOBAL-NEXT: xsmulsp 0, 1, 0 -; GLOBAL-NEXT: xsmulsp 0, 0, 2 -; GLOBAL-NEXT: .LBB15_2: -; GLOBAL-NEXT: fmr 1, 0 +; GLOBAL-NEXT: addis 3, 2, .LCPI15_1@toc@ha +; GLOBAL-NEXT: xsmulsp 0, 2, 0 +; GLOBAL-NEXT: lfs 2, .LCPI15_1@toc@l(3) +; GLOBAL-NEXT: xssubsp 1, 1, 2 +; GLOBAL-NEXT: xxlxor 2, 2, 2 +; GLOBAL-NEXT: xsmulsp 0, 0, 3 +; GLOBAL-NEXT: fsel 1, 1, 0, 2 ; GLOBAL-NEXT: blr %rt = call contract reassoc afn ninf float @llvm.sqrt.f32(float %x) ret float %rt @@ -517,21 +509,19 @@ define float @sqrt_fast_preserve_sign(float %x) #1 { ; ; GLOBAL-LABEL: sqrt_fast_preserve_sign: ; GLOBAL: # %bb.0: -; GLOBAL-NEXT: xxlxor 0, 0, 0 -; GLOBAL-NEXT: fcmpu 0, 1, 0 -; GLOBAL-NEXT: beq 0, .LBB16_2 -; GLOBAL-NEXT: # %bb.1: ; GLOBAL-NEXT: xsrsqrtesp 0, 1 ; GLOBAL-NEXT: vspltisw 2, -3 ; GLOBAL-NEXT: addis 3, 2, .LCPI16_0@toc@ha -; GLOBAL-NEXT: xvcvsxwdp 2, 34 -; GLOBAL-NEXT: xsmulsp 1, 1, 0 -; GLOBAL-NEXT: xsmaddasp 2, 1, 0 +; GLOBAL-NEXT: xvcvsxwdp 3, 34 +; GLOBAL-NEXT: xsmulsp 2, 1, 0 +; GLOBAL-NEXT: xsmaddasp 3, 2, 0 ; GLOBAL-NEXT: lfs 0, .LCPI16_0@toc@l(3) -; GLOBAL-NEXT: xsmulsp 0, 1, 0 -; GLOBAL-NEXT: xsmulsp 0, 0, 2 -; GLOBAL-NEXT: .LBB16_2: -; GLOBAL-NEXT: fmr 1, 0 +; GLOBAL-NEXT: xsmulsp 0, 2, 0 +; GLOBAL-NEXT: xxlxor 2, 2, 2 +; GLOBAL-NEXT: xsmulsp 0, 0, 3 +; GLOBAL-NEXT: fsel 2, 1, 2, 0 +; GLOBAL-NEXT: xsnegdp 1, 1 +; GLOBAL-NEXT: fsel 1, 1, 2, 0 ; GLOBAL-NEXT: blr %rt = call contract reassoc ninf afn float @llvm.sqrt.f32(float %x) ret float %rt diff --git a/llvm/test/CodeGen/PowerPC/lxvkq-vec-constant.ll b/llvm/test/CodeGen/PowerPC/lxvkq-vec-constant.ll new file mode 100644 index 0000000..0ee4524 --- /dev/null +++ b/llvm/test/CodeGen/PowerPC/lxvkq-vec-constant.ll @@ -0,0 +1,307 @@ +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5 + +; RUN: llc -verify-machineinstrs -mcpu=pwr10 -mtriple=powerpc64le-unknown-unknown \ +; RUN: -ppc-asm-full-reg-names --ppc-vsr-nums-as-vr < %s | FileCheck %s --check-prefix=POWERPC64-LE-10 + +; RUN: llc -verify-machineinstrs -mcpu=pwr10 -mtriple=powerpc64-unknown-unknown \ +; RUN: -ppc-asm-full-reg-names --ppc-vsr-nums-as-vr < %s | FileCheck %s --check-prefix=POWERPC64-BE-10 + +; Test LXVKQ instruction generation for special vector constants matching 128 bit patterns: +; 0x8000_0000_0000_0000_0000_0000_0000_0000 (MSB set pattern) +; 0x0000_0000_0000_0000_0000_0000_0000_0001 (LSB set pattern) + +; ============================================================================= +; v2i64 tests - MSB set pattern (0x8000_0000_0000_0000_0000_0000_0000_0000) +; ============================================================================= + +; Big-Endian: 0x8000_0000_0000_0000_0000_0000_0000_0000 represents <-9223372036854775808, 0> +define dso_local noundef <2 x i64> @test_v2i64_msb_set_bigendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v2i64_msb_set_bigendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: plxv v2, .LCPI0_0@PCREL(0), 1 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v2i64_msb_set_bigendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: lxvkq v2, 16 +; POWERPC64-BE-10-NEXT: blr +entry: + ret <2 x i64> <i64 -9223372036854775808, i64 0> +} + +; Little-Endian: 0x8000_0000_0000_0000_0000_0000_0000_0000 represents <0, -9223372036854775808> +define dso_local noundef <2 x i64> @test_v2i64_msb_set_littleendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v2i64_msb_set_littleendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: lxvkq v2, 16 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v2i64_msb_set_littleendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: addis r3, r2, .LCPI1_0@toc@ha +; POWERPC64-BE-10-NEXT: addi r3, r3, .LCPI1_0@toc@l +; POWERPC64-BE-10-NEXT: lxv v2, 0(r3) +; POWERPC64-BE-10-NEXT: blr +entry: + ret <2 x i64> <i64 0, i64 -9223372036854775808> +} + +; ============================================================================= +; v4i32 tests - MSB set pattern (0x8000_0000_0000_0000_0000_0000_0000_0000) +; ============================================================================= + +; Big-Endian: 0x8000_0000_0000_0000_0000_0000_0000_0000 represents <-2147483648, 0, 0, 0> +define dso_local noundef <4 x i32> @test_v4i32_msb_set_bigendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v4i32_msb_set_bigendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: plxv v2, .LCPI2_0@PCREL(0), 1 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v4i32_msb_set_bigendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: lxvkq v2, 16 +; POWERPC64-BE-10-NEXT: blr +entry: + ret <4 x i32> <i32 -2147483648, i32 0, i32 0, i32 0> +} + +; Little-Endian: 0x8000_0000_0000_0000_0000_0000_0000_0000 represents <0, 0, 0, -2147483648> +define dso_local noundef <4 x i32> @test_v4i32_msb_set_littleendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v4i32_msb_set_littleendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: lxvkq v2, 16 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v4i32_msb_set_littleendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: addis r3, r2, .LCPI3_0@toc@ha +; POWERPC64-BE-10-NEXT: addi r3, r3, .LCPI3_0@toc@l +; POWERPC64-BE-10-NEXT: lxv v2, 0(r3) +; POWERPC64-BE-10-NEXT: blr +entry: + ret <4 x i32> <i32 0, i32 0, i32 0, i32 -2147483648> +} + +; ============================================================================= +; v8i16 tests - MSB set pattern (0x8000_0000_0000_0000_0000_0000_0000_0000) +; ============================================================================= + +; Big-Endian: 0x8000_0000_0000_0000_0000_0000_0000_0000 represents <-32768, 0, 0, 0, 0, 0, 0, 0> +define dso_local noundef <8 x i16> @test_v8i16_msb_set_bigendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v8i16_msb_set_bigendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: plxv v2, .LCPI4_0@PCREL(0), 1 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v8i16_msb_set_bigendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: lxvkq v2, 16 +; POWERPC64-BE-10-NEXT: blr +entry: + ret <8 x i16> <i16 -32768, i16 0, i16 0, i16 0, i16 0, i16 0, i16 0, i16 0> +} + +; Little-Endian: 0x8000_0000_0000_0000_0000_0000_0000_0000 represents <0, 0, 0, 0, 0, 0, 0, -32768> +define dso_local noundef <8 x i16> @test_v8i16_msb_set_littleendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v8i16_msb_set_littleendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: lxvkq v2, 16 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v8i16_msb_set_littleendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: addis r3, r2, .LCPI5_0@toc@ha +; POWERPC64-BE-10-NEXT: addi r3, r3, .LCPI5_0@toc@l +; POWERPC64-BE-10-NEXT: lxv v2, 0(r3) +; POWERPC64-BE-10-NEXT: blr +entry: + ret <8 x i16> <i16 0, i16 0, i16 0, i16 0, i16 0, i16 0, i16 0, i16 -32768> +} + +; ============================================================================= +; v16i8 tests - MSB set pattern (0x8000_0000_0000_0000_0000_0000_0000_0000) +; ============================================================================= + +; Big-Endian: 0x8000_0000_0000_0000_0000_0000_0000_0000 represents <-128, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0> +define dso_local noundef <16 x i8> @test_v16i8_msb_set_bigendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v16i8_msb_set_bigendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: plxv v2, .LCPI6_0@PCREL(0), 1 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v16i8_msb_set_bigendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: lxvkq v2, 16 +; POWERPC64-BE-10-NEXT: blr +entry: + ret <16 x i8> <i8 -128, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0> +} + +; Little-Endian: 0x8000_0000_0000_0000_0000_0000_0000_0000 represents <0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, -128> +define dso_local noundef <16 x i8> @test_v16i8_msb_set_littleendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v16i8_msb_set_littleendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: lxvkq v2, 16 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v16i8_msb_set_littleendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: addis r3, r2, .LCPI7_0@toc@ha +; POWERPC64-BE-10-NEXT: addi r3, r3, .LCPI7_0@toc@l +; POWERPC64-BE-10-NEXT: lxv v2, 0(r3) +; POWERPC64-BE-10-NEXT: blr +entry: + ret <16 x i8> <i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 -128> +} + +; ============================================================================= +; v2i64 tests - LSB set pattern (0x0000_0000_0000_0000_0000_0000_0000_0001) +; ============================================================================= + +; Big-Endian: 0x0000_0000_0000_0000_0000_0000_0000_0001 represents <0, 1> +define dso_local noundef <2 x i64> @test_v2i64_lsb_set_bigendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v2i64_lsb_set_bigendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: plxv v2, .LCPI8_0@PCREL(0), 1 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v2i64_lsb_set_bigendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: xxspltib v2, 255 +; POWERPC64-BE-10-NEXT: vsrq v2, v2, v2 +; POWERPC64-BE-10-NEXT: blr +entry: + ret <2 x i64> <i64 0, i64 1> +} + +; Little-Endian: 0x0000_0000_0000_0000_0000_0000_0000_0001 represents <1, 0> +define dso_local noundef <2 x i64> @test_v2i64_lsb_set_littleendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v2i64_lsb_set_littleendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: xxspltib v2, 255 +; POWERPC64-LE-10-NEXT: vsrq v2, v2, v2 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v2i64_lsb_set_littleendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: addis r3, r2, .LCPI9_0@toc@ha +; POWERPC64-BE-10-NEXT: addi r3, r3, .LCPI9_0@toc@l +; POWERPC64-BE-10-NEXT: lxv v2, 0(r3) +; POWERPC64-BE-10-NEXT: blr +entry: + ret <2 x i64> <i64 1, i64 0> +} + +; ============================================================================= +; v4i32 tests - LSB set pattern (0x0000_0000_0000_0000_0000_0000_0000_0001) +; ============================================================================= + +; Big-Endian: 0x0000_0000_0000_0000_0000_0000_0000_0001 represents <0, 0, 0, 1> +define dso_local noundef <4 x i32> @test_v4i32_lsb_set_bigendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v4i32_lsb_set_bigendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: plxv v2, .LCPI10_0@PCREL(0), 1 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v4i32_lsb_set_bigendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: xxspltib v2, 255 +; POWERPC64-BE-10-NEXT: vsrq v2, v2, v2 +; POWERPC64-BE-10-NEXT: blr +entry: + ret <4 x i32> <i32 0, i32 0, i32 0, i32 1> +} + +; Little-Endian: 0x0000_0000_0000_0000_0000_0000_0000_0001 represents <1, 0, 0, 0> +define dso_local noundef <4 x i32> @test_v4i32_lsb_set_littleendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v4i32_lsb_set_littleendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: xxspltib v2, 255 +; POWERPC64-LE-10-NEXT: vsrq v2, v2, v2 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v4i32_lsb_set_littleendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: addis r3, r2, .LCPI11_0@toc@ha +; POWERPC64-BE-10-NEXT: addi r3, r3, .LCPI11_0@toc@l +; POWERPC64-BE-10-NEXT: lxv v2, 0(r3) +; POWERPC64-BE-10-NEXT: blr +entry: + ret <4 x i32> <i32 1, i32 0, i32 0, i32 0> +} + +; ============================================================================= +; v8i16 tests - LSB set pattern (0x0000_0000_0000_0000_0000_0000_0000_0001) +; ============================================================================= + +; Big-Endian: 0x0000_0000_0000_0000_0000_0000_0000_0001 represents <0, 0, 0, 0, 0, 0, 0, 1> +define dso_local noundef <8 x i16> @test_v8i16_lsb_set_bigendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v8i16_lsb_set_bigendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: plxv v2, .LCPI12_0@PCREL(0), 1 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v8i16_lsb_set_bigendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: xxspltib v2, 255 +; POWERPC64-BE-10-NEXT: vsrq v2, v2, v2 +; POWERPC64-BE-10-NEXT: blr +entry: + ret <8 x i16> <i16 0, i16 0, i16 0, i16 0, i16 0, i16 0, i16 0, i16 1> +} + +; Little-Endian: 0x0000_0000_0000_0000_0000_0000_0000_0001 represents <1, 0, 0, 0, 0, 0, 0, 0> +define dso_local noundef <8 x i16> @test_v8i16_lsb_set_littleendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v8i16_lsb_set_littleendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: xxspltib v2, 255 +; POWERPC64-LE-10-NEXT: vsrq v2, v2, v2 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v8i16_lsb_set_littleendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: addis r3, r2, .LCPI13_0@toc@ha +; POWERPC64-BE-10-NEXT: addi r3, r3, .LCPI13_0@toc@l +; POWERPC64-BE-10-NEXT: lxv v2, 0(r3) +; POWERPC64-BE-10-NEXT: blr +entry: + ret <8 x i16> <i16 1, i16 0, i16 0, i16 0, i16 0, i16 0, i16 0, i16 0> +} + +; ============================================================================= +; v16i8 tests - LSB set pattern (0x0000_0000_0000_0000_0000_0000_0000_0001) +; ============================================================================= + +; Big-Endian: 0x0000_0000_0000_0000_0000_0000_0000_0001 represents <0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1> +define dso_local noundef <16 x i8> @test_v16i8_lsb_set_bigendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v16i8_lsb_set_bigendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: plxv v2, .LCPI14_0@PCREL(0), 1 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v16i8_lsb_set_bigendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: xxspltib v2, 255 +; POWERPC64-BE-10-NEXT: vsrq v2, v2, v2 +; POWERPC64-BE-10-NEXT: blr +entry: + ret <16 x i8> <i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 1> +} + +; Little-Endian: 0x0000_0000_0000_0000_0000_0000_0000_0001 represents <1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0> +define dso_local noundef <16 x i8> @test_v16i8_lsb_set_littleendian() local_unnamed_addr { +; POWERPC64-LE-10-LABEL: test_v16i8_lsb_set_littleendian: +; POWERPC64-LE-10: # %bb.0: # %entry +; POWERPC64-LE-10-NEXT: xxspltib v2, 255 +; POWERPC64-LE-10-NEXT: vsrq v2, v2, v2 +; POWERPC64-LE-10-NEXT: blr +; +; POWERPC64-BE-10-LABEL: test_v16i8_lsb_set_littleendian: +; POWERPC64-BE-10: # %bb.0: # %entry +; POWERPC64-BE-10-NEXT: addis r3, r2, .LCPI15_0@toc@ha +; POWERPC64-BE-10-NEXT: addi r3, r3, .LCPI15_0@toc@l +; POWERPC64-BE-10-NEXT: lxv v2, 0(r3) +; POWERPC64-BE-10-NEXT: blr +entry: + ret <16 x i8> <i8 1, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0, i8 0> +}
\ No newline at end of file diff --git a/llvm/test/CodeGen/PowerPC/vector-reduce-add.ll b/llvm/test/CodeGen/PowerPC/vector-reduce-add.ll index 0892210..d506d20 100644 --- a/llvm/test/CodeGen/PowerPC/vector-reduce-add.ll +++ b/llvm/test/CodeGen/PowerPC/vector-reduce-add.ll @@ -1566,12 +1566,16 @@ define dso_local i64 @v16i8tov16i64_sign(<16 x i8> %a) local_unnamed_addr #0 { ; PWR10BE-LABEL: v16i8tov16i64_sign: ; PWR10BE: # %bb.0: # %entry ; PWR10BE-NEXT: addis r3, r2, .LCPI23_0@toc@ha +; PWR10BE-NEXT: xxspltib v1, 255 ; PWR10BE-NEXT: addi r3, r3, .LCPI23_0@toc@l +; PWR10BE-NEXT: vsrq v1, v1, v1 ; PWR10BE-NEXT: lxv v3, 0(r3) ; PWR10BE-NEXT: addis r3, r2, .LCPI23_1@toc@ha ; PWR10BE-NEXT: addi r3, r3, .LCPI23_1@toc@l +; PWR10BE-NEXT: vperm v1, v2, v2, v1 ; PWR10BE-NEXT: lxv v4, 0(r3) ; PWR10BE-NEXT: addis r3, r2, .LCPI23_2@toc@ha +; PWR10BE-NEXT: vextsb2d v1, v1 ; PWR10BE-NEXT: vperm v3, v2, v2, v3 ; PWR10BE-NEXT: addi r3, r3, .LCPI23_2@toc@l ; PWR10BE-NEXT: vextsb2d v3, v3 @@ -1585,23 +1589,18 @@ define dso_local i64 @v16i8tov16i64_sign(<16 x i8> %a) local_unnamed_addr #0 { ; PWR10BE-NEXT: vperm v5, v2, v2, v5 ; PWR10BE-NEXT: addi r3, r3, .LCPI23_4@toc@l ; PWR10BE-NEXT: vextsb2d v5, v5 -; PWR10BE-NEXT: lxv v1, 0(r3) +; PWR10BE-NEXT: lxv v6, 0(r3) ; PWR10BE-NEXT: addis r3, r2, .LCPI23_5@toc@ha ; PWR10BE-NEXT: vperm v0, v2, v2, v0 ; PWR10BE-NEXT: addi r3, r3, .LCPI23_5@toc@l ; PWR10BE-NEXT: vextsb2d v0, v0 -; PWR10BE-NEXT: lxv v6, 0(r3) +; PWR10BE-NEXT: lxv v7, 0(r3) ; PWR10BE-NEXT: addis r3, r2, .LCPI23_6@toc@ha -; PWR10BE-NEXT: vperm v1, v2, v2, v1 +; PWR10BE-NEXT: vperm v6, v2, v2, v6 ; PWR10BE-NEXT: vaddudm v5, v0, v5 ; PWR10BE-NEXT: vaddudm v3, v4, v3 ; PWR10BE-NEXT: vaddudm v3, v3, v5 ; PWR10BE-NEXT: addi r3, r3, .LCPI23_6@toc@l -; PWR10BE-NEXT: vextsb2d v1, v1 -; PWR10BE-NEXT: lxv v7, 0(r3) -; PWR10BE-NEXT: addis r3, r2, .LCPI23_7@toc@ha -; PWR10BE-NEXT: vperm v6, v2, v2, v6 -; PWR10BE-NEXT: addi r3, r3, .LCPI23_7@toc@l ; PWR10BE-NEXT: vextsb2d v6, v6 ; PWR10BE-NEXT: lxv v8, 0(r3) ; PWR10BE-NEXT: vperm v7, v2, v2, v7 @@ -1609,7 +1608,7 @@ define dso_local i64 @v16i8tov16i64_sign(<16 x i8> %a) local_unnamed_addr #0 { ; PWR10BE-NEXT: vperm v2, v2, v2, v8 ; PWR10BE-NEXT: vextsb2d v2, v2 ; PWR10BE-NEXT: vaddudm v2, v2, v7 -; PWR10BE-NEXT: vaddudm v4, v6, v1 +; PWR10BE-NEXT: vaddudm v4, v1, v6 ; PWR10BE-NEXT: vaddudm v2, v4, v2 ; PWR10BE-NEXT: vaddudm v2, v2, v3 ; PWR10BE-NEXT: xxswapd v3, v2 diff --git a/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-eqv.ll b/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-eqv.ll index 24a1724..ba7680b 100644 --- a/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-eqv.ll +++ b/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-eqv.ll @@ -15,11 +15,9 @@ define <4 x i32> @ternary_A_or_BC_eqv_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i3 ; CHECK-LABEL: ternary_A_or_BC_eqv_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxleqv vs1, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 151 ; CHECK-NEXT: blr entry: %or = or <4 x i32> %B, %C @@ -34,12 +32,10 @@ define <2 x i64> @ternary_A_or_BC_eqv_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i6 ; CHECK-LABEL: ternary_A_or_BC_eqv_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxleqv vs1, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 151 ; CHECK-NEXT: blr entry: %or = or <2 x i64> %B, %C @@ -54,11 +50,9 @@ define <16 x i8> @ternary_A_or_BC_eqv_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_or_BC_eqv_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxleqv vs1, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 151 ; CHECK-NEXT: blr entry: %or = or <16 x i8> %B, %C @@ -73,11 +67,9 @@ define <8 x i16> @ternary_A_or_BC_eqv_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i1 ; CHECK-LABEL: ternary_A_or_BC_eqv_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxleqv vs1, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 151 ; CHECK-NEXT: blr entry: %or = or <8 x i16> %B, %C @@ -92,11 +84,9 @@ define <4 x i32> @ternary_A_nor_BC_eqv_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i ; CHECK-LABEL: ternary_A_nor_BC_eqv_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v3, v4 -; CHECK-NEXT: xxleqv vs1, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 152 ; CHECK-NEXT: blr entry: %or = or <4 x i32> %B, %C @@ -112,12 +102,10 @@ define <2 x i64> @ternary_A_nor_BC_eqv_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i ; CHECK-LABEL: ternary_A_nor_BC_eqv_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v3, v4 -; CHECK-NEXT: xxleqv vs1, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 152 ; CHECK-NEXT: blr entry: %or = or <2 x i64> %B, %C @@ -133,11 +121,9 @@ define <16 x i8> @ternary_A_nor_BC_eqv_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_nor_BC_eqv_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnor vs0, v3, v4 -; CHECK-NEXT: xxleqv vs1, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 152 ; CHECK-NEXT: blr entry: %or = or <16 x i8> %B, %C @@ -153,11 +139,9 @@ define <8 x i16> @ternary_A_nor_BC_eqv_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i ; CHECK-LABEL: ternary_A_nor_BC_eqv_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnor vs0, v3, v4 -; CHECK-NEXT: xxleqv vs1, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 152 ; CHECK-NEXT: blr entry: %or = or <8 x i16> %B, %C @@ -173,10 +157,9 @@ define <4 x i32> @ternary_A_not_C_eqv_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i3 ; CHECK-LABEL: ternary_A_not_C_eqv_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v4, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxeval v2, v2, vs0, v3, 99 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 154 ; CHECK-NEXT: blr entry: %not = xor <4 x i32> %C, <i32 -1, i32 -1, i32 -1, i32 -1> ; Vector not operation @@ -191,12 +174,10 @@ define <2 x i64> @ternary_A_not_C_eqv_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i6 ; CHECK-LABEL: ternary_A_not_C_eqv_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v4, v4 -; CHECK-NEXT: xxleqv vs1, v4, v3 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 154 ; CHECK-NEXT: blr entry: %not = xor <2 x i64> %C, <i64 -1, i64 -1> ; Vector not operation @@ -211,11 +192,9 @@ define <16 x i8> @ternary_A_not_C_eqv_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_not_C_eqv_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnor vs0, v4, v4 -; CHECK-NEXT: xxleqv vs1, v4, v3 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 154 ; CHECK-NEXT: blr entry: %not = xor <16 x i8> %C, <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1> ; Vector not operation @@ -230,11 +209,9 @@ define <8 x i16> @ternary_A_not_C_eqv_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i1 ; CHECK-LABEL: ternary_A_not_C_eqv_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnor vs0, v4, v4 -; CHECK-NEXT: xxleqv vs1, v4, v3 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 154 ; CHECK-NEXT: blr entry: %not = xor <8 x i16> %C, <i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1> ; Vector not operation @@ -249,11 +226,9 @@ define <4 x i32> @ternary_A_nand_BC_eqv_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x ; CHECK-LABEL: ternary_A_nand_BC_eqv_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxleqv vs1, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 158 ; CHECK-NEXT: blr entry: %and = and <4 x i32> %B, %C @@ -269,12 +244,10 @@ define <2 x i64> @ternary_A_nand_BC_eqv_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x ; CHECK-LABEL: ternary_A_nand_BC_eqv_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxleqv vs1, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 158 ; CHECK-NEXT: blr entry: %and = and <2 x i64> %B, %C @@ -290,11 +263,9 @@ define <16 x i8> @ternary_A_nand_BC_eqv_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 ; CHECK-LABEL: ternary_A_nand_BC_eqv_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxleqv vs1, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 158 ; CHECK-NEXT: blr entry: %and = and <16 x i8> %B, %C @@ -310,11 +281,9 @@ define <8 x i16> @ternary_A_nand_BC_eqv_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x ; CHECK-LABEL: ternary_A_nand_BC_eqv_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxleqv vs1, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 158 ; CHECK-NEXT: blr entry: %and = and <8 x i16> %B, %C diff --git a/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-nand.ll b/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-nand.ll index 7a6733d3..067b089 100644 --- a/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-nand.ll +++ b/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-nand.ll @@ -15,10 +15,9 @@ define <4 x i32> @ternary_A_B_nand_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i32> ; CHECK-LABEL: ternary_A_B_nand_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnand vs0, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v3, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 227 ; CHECK-NEXT: blr entry: %and = and <4 x i32> %B, %C @@ -32,11 +31,10 @@ define <2 x i64> @ternary_A_B_nand_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i64> ; CHECK-LABEL: ternary_A_B_nand_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnand vs0, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v3, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 227 ; CHECK-NEXT: blr entry: %and = and <2 x i64> %B, %C @@ -50,10 +48,9 @@ define <16 x i8> @ternary_A_B_nand_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 x i8> ; CHECK-LABEL: ternary_A_B_nand_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnand vs0, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v3, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 227 ; CHECK-NEXT: blr entry: %and = and <16 x i8> %B, %C @@ -67,10 +64,9 @@ define <8 x i16> @ternary_A_B_nand_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i16> ; CHECK-LABEL: ternary_A_B_nand_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnand vs0, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v3, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 227 ; CHECK-NEXT: blr entry: %and = and <8 x i16> %B, %C @@ -84,10 +80,9 @@ define <4 x i32> @ternary_A_C_nand_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i32> ; CHECK-LABEL: ternary_A_C_nand_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnand vs0, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v4, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 229 ; CHECK-NEXT: blr entry: %and = and <4 x i32> %B, %C @@ -101,11 +96,10 @@ define <2 x i64> @ternary_A_C_nand_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i64> ; CHECK-LABEL: ternary_A_C_nand_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnand vs0, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v4, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 229 ; CHECK-NEXT: blr entry: %and = and <2 x i64> %B, %C @@ -119,10 +113,9 @@ define <16 x i8> @ternary_A_C_nand_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 x i8> ; CHECK-LABEL: ternary_A_C_nand_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnand vs0, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v4, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 229 ; CHECK-NEXT: blr entry: %and = and <16 x i8> %B, %C @@ -136,10 +129,9 @@ define <8 x i16> @ternary_A_C_nand_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i16> ; CHECK-LABEL: ternary_A_C_nand_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnand vs0, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v4, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 229 ; CHECK-NEXT: blr entry: %and = and <8 x i16> %B, %C @@ -153,11 +145,9 @@ define <4 x i32> @ternary_A_xor_BC_nand_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x ; CHECK-LABEL: ternary_A_xor_BC_nand_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnand vs1, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 230 ; CHECK-NEXT: blr entry: %xor = xor <4 x i32> %B, %C @@ -172,12 +162,10 @@ define <2 x i64> @ternary_A_xor_BC_nand_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x ; CHECK-LABEL: ternary_A_xor_BC_nand_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnand vs1, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 230 ; CHECK-NEXT: blr entry: %xor = xor <2 x i64> %B, %C @@ -192,11 +180,9 @@ define <16 x i8> @ternary_A_xor_BC_nand_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 ; CHECK-LABEL: ternary_A_xor_BC_nand_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnand vs1, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 230 ; CHECK-NEXT: blr entry: %xor = xor <16 x i8> %B, %C @@ -211,11 +197,9 @@ define <8 x i16> @ternary_A_xor_BC_nand_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x ; CHECK-LABEL: ternary_A_xor_BC_nand_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnand vs1, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 230 ; CHECK-NEXT: blr entry: %xor = xor <8 x i16> %B, %C @@ -230,11 +214,9 @@ define <4 x i32> @ternary_A_or_BC_nand_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i ; CHECK-LABEL: ternary_A_or_BC_nand_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxlnand vs1, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 231 ; CHECK-NEXT: blr entry: %or = or <4 x i32> %B, %C @@ -249,12 +231,10 @@ define <2 x i64> @ternary_A_or_BC_nand_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i ; CHECK-LABEL: ternary_A_or_BC_nand_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxlnand vs1, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 231 ; CHECK-NEXT: blr entry: %or = or <2 x i64> %B, %C @@ -269,11 +249,9 @@ define <16 x i8> @ternary_A_or_BC_nand_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_or_BC_nand_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxlnand vs1, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 231 ; CHECK-NEXT: blr entry: %or = or <16 x i8> %B, %C @@ -288,11 +266,9 @@ define <8 x i16> @ternary_A_or_BC_nand_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i ; CHECK-LABEL: ternary_A_or_BC_nand_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxlnand vs1, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 231 ; CHECK-NEXT: blr entry: %or = or <8 x i16> %B, %C @@ -307,11 +283,9 @@ define <4 x i32> @ternary_A_eqv_BC_nand_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x ; CHECK-LABEL: ternary_A_eqv_BC_nand_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxleqv vs0, v3, v4 -; CHECK-NEXT: xxlnand vs1, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 233 ; CHECK-NEXT: blr entry: %xor = xor <4 x i32> %B, %C @@ -327,12 +301,10 @@ define <2 x i64> @ternary_A_eqv_BC_nand_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x ; CHECK-LABEL: ternary_A_eqv_BC_nand_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxleqv vs0, v3, v4 -; CHECK-NEXT: xxlnand vs1, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 233 ; CHECK-NEXT: blr entry: %xor = xor <2 x i64> %B, %C @@ -348,11 +320,9 @@ define <16 x i8> @ternary_A_eqv_BC_nand_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 ; CHECK-LABEL: ternary_A_eqv_BC_nand_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxleqv vs0, v3, v4 -; CHECK-NEXT: xxlnand vs1, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 233 ; CHECK-NEXT: blr entry: %xor = xor <16 x i8> %B, %C @@ -368,11 +338,9 @@ define <8 x i16> @ternary_A_eqv_BC_nand_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x ; CHECK-LABEL: ternary_A_eqv_BC_nand_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxleqv vs0, v3, v4 -; CHECK-NEXT: xxlnand vs1, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 233 ; CHECK-NEXT: blr entry: %xor = xor <8 x i16> %B, %C diff --git a/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-nor.ll b/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-nor.ll index d635952..3695874 100644 --- a/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-nor.ll +++ b/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-nor.ll @@ -15,11 +15,9 @@ define <4 x i32> @ternary_A_and_BC_nor_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i ; CHECK-LABEL: ternary_A_and_BC_nor_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxland vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 129 ; CHECK-NEXT: blr entry: %and = and <4 x i32> %B, %C @@ -34,12 +32,10 @@ define <2 x i64> @ternary_A_and_BC_nor_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i ; CHECK-LABEL: ternary_A_and_BC_nor_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxland vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 129 ; CHECK-NEXT: blr entry: %and = and <2 x i64> %B, %C @@ -54,11 +50,9 @@ define <16 x i8> @ternary_A_and_BC_nor_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_and_BC_nor_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxland vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 129 ; CHECK-NEXT: blr entry: %and = and <16 x i8> %B, %C @@ -73,11 +67,9 @@ define <8 x i16> @ternary_A_and_BC_nor_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i ; CHECK-LABEL: ternary_A_and_BC_nor_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxland vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 129 ; CHECK-NEXT: blr entry: %and = and <8 x i16> %B, %C @@ -92,10 +84,9 @@ define <4 x i32> @ternary_A_B_nor_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i32> % ; CHECK-LABEL: ternary_A_B_nor_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v3, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 131 ; CHECK-NEXT: blr entry: %or = or <4 x i32> %B, %C @@ -109,11 +100,10 @@ define <2 x i64> @ternary_A_B_nor_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i64> % ; CHECK-LABEL: ternary_A_B_nor_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v3, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 131 ; CHECK-NEXT: blr entry: %or = or <2 x i64> %B, %C @@ -127,10 +117,9 @@ define <16 x i8> @ternary_A_B_nor_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 x i8> ; CHECK-LABEL: ternary_A_B_nor_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnor vs0, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v3, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 131 ; CHECK-NEXT: blr entry: %or = or <16 x i8> %B, %C @@ -144,10 +133,9 @@ define <8 x i16> @ternary_A_B_nor_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i16> % ; CHECK-LABEL: ternary_A_B_nor_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnor vs0, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v3, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 131 ; CHECK-NEXT: blr entry: %or = or <8 x i16> %B, %C @@ -161,10 +149,9 @@ define <4 x i32> @ternary_A_C_nor_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i32> % ; CHECK-LABEL: ternary_A_C_nor_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v4, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 133 ; CHECK-NEXT: blr entry: %or = or <4 x i32> %B, %C @@ -178,11 +165,10 @@ define <2 x i64> @ternary_A_C_nor_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i64> % ; CHECK-LABEL: ternary_A_C_nor_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v4, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 133 ; CHECK-NEXT: blr entry: %or = or <2 x i64> %B, %C @@ -196,10 +182,9 @@ define <16 x i8> @ternary_A_C_nor_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 x i8> ; CHECK-LABEL: ternary_A_C_nor_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnor vs0, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v4, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 133 ; CHECK-NEXT: blr entry: %or = or <16 x i8> %B, %C @@ -213,10 +198,9 @@ define <8 x i16> @ternary_A_C_nor_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i16> % ; CHECK-LABEL: ternary_A_C_nor_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnor vs0, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v4, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 133 ; CHECK-NEXT: blr entry: %or = or <8 x i16> %B, %C @@ -230,11 +214,9 @@ define <4 x i32> @ternary_A_xor_BC_nor_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i ; CHECK-LABEL: ternary_A_xor_BC_nor_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 134 ; CHECK-NEXT: blr entry: %xor = xor <4 x i32> %B, %C @@ -249,12 +231,10 @@ define <2 x i64> @ternary_A_xor_BC_nor_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i ; CHECK-LABEL: ternary_A_xor_BC_nor_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 134 ; CHECK-NEXT: blr entry: %xor = xor <2 x i64> %B, %C @@ -269,11 +249,9 @@ define <16 x i8> @ternary_A_xor_BC_nor_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_xor_BC_nor_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 134 ; CHECK-NEXT: blr entry: %xor = xor <16 x i8> %B, %C @@ -288,11 +266,9 @@ define <8 x i16> @ternary_A_xor_BC_nor_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i ; CHECK-LABEL: ternary_A_xor_BC_nor_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 134 ; CHECK-NEXT: blr entry: %xor = xor <8 x i16> %B, %C @@ -307,11 +283,9 @@ define <4 x i32> @ternary_A_not_C_nor_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i3 ; CHECK-LABEL: ternary_A_not_C_nor_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v4, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 138 ; CHECK-NEXT: blr entry: %not = xor <4 x i32> %C, <i32 -1, i32 -1, i32 -1, i32 -1> ; Vector not operation @@ -326,12 +300,10 @@ define <2 x i64> @ternary_A_not_C_nor_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i6 ; CHECK-LABEL: ternary_A_not_C_nor_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v4, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 138 ; CHECK-NEXT: blr entry: %not = xor <2 x i64> %C, <i64 -1, i64 -1> ; Vector not operation @@ -346,11 +318,9 @@ define <16 x i8> @ternary_A_not_C_nor_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_not_C_nor_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnor vs0, v4, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 138 ; CHECK-NEXT: blr entry: %not = xor <16 x i8> %C, <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1> ; Vector not operation @@ -365,11 +335,9 @@ define <8 x i16> @ternary_A_not_C_nor_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i1 ; CHECK-LABEL: ternary_A_not_C_nor_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnor vs0, v4, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 138 ; CHECK-NEXT: blr entry: %not = xor <8 x i16> %C, <i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1> ; Vector not operation @@ -384,11 +352,9 @@ define <4 x i32> @ternary_A_not_B_nor_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i3 ; CHECK-LABEL: ternary_A_not_B_nor_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v3, v3 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 140 ; CHECK-NEXT: blr entry: %not = xor <4 x i32> %B, <i32 -1, i32 -1, i32 -1, i32 -1> ; Vector not operation @@ -403,12 +369,10 @@ define <2 x i64> @ternary_A_not_B_nor_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i6 ; CHECK-LABEL: ternary_A_not_B_nor_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v3, v3 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 140 ; CHECK-NEXT: blr entry: %not = xor <2 x i64> %B, <i64 -1, i64 -1> ; Vector not operation @@ -423,11 +387,9 @@ define <16 x i8> @ternary_A_not_B_nor_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_not_B_nor_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnor vs0, v3, v3 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 140 ; CHECK-NEXT: blr entry: %not = xor <16 x i8> %B, <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1> ; Vector not operation @@ -442,11 +404,9 @@ define <8 x i16> @ternary_A_not_B_nor_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i1 ; CHECK-LABEL: ternary_A_not_B_nor_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnor vs0, v3, v3 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 140 ; CHECK-NEXT: blr entry: %not = xor <8 x i16> %B, <i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1> ; Vector not operation @@ -461,11 +421,9 @@ define <4 x i32> @ternary_A_nand_BC_nor_BC_4x32(<4 x i1> %A, <4 x i32> %B, <4 x ; CHECK-LABEL: ternary_A_nand_BC_nor_BC_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 142 ; CHECK-NEXT: blr entry: %and = and <4 x i32> %B, %C @@ -481,12 +439,10 @@ define <2 x i64> @ternary_A_nand_BC_nor_BC_2x64(<2 x i1> %A, <2 x i64> %B, <2 x ; CHECK-LABEL: ternary_A_nand_BC_nor_BC_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 142 ; CHECK-NEXT: blr entry: %and = and <2 x i64> %B, %C @@ -502,11 +458,9 @@ define <16 x i8> @ternary_A_nand_BC_nor_BC_16x8(<16 x i1> %A, <16 x i8> %B, <16 ; CHECK-LABEL: ternary_A_nand_BC_nor_BC_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 142 ; CHECK-NEXT: blr entry: %and = and <16 x i8> %B, %C @@ -522,11 +476,9 @@ define <8 x i16> @ternary_A_nand_BC_nor_BC_8x16(<8 x i1> %A, <8 x i16> %B, <8 x ; CHECK-LABEL: ternary_A_nand_BC_nor_BC_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 142 ; CHECK-NEXT: blr entry: %and = and <8 x i16> %B, %C diff --git a/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-not-b.ll b/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-not-b.ll index 6203a96..a67d9cf 100644 --- a/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-not-b.ll +++ b/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-not-b.ll @@ -15,11 +15,9 @@ define <4 x i32> @ternary_A_and_BC_not_B_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i3 ; CHECK-LABEL: ternary_A_and_BC_not_B_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxland vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 193 ; CHECK-NEXT: blr entry: %and = and <4 x i32> %B, %C @@ -33,12 +31,10 @@ define <2 x i64> @ternary_A_and_BC_not_B_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i6 ; CHECK-LABEL: ternary_A_and_BC_not_B_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxland vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 193 ; CHECK-NEXT: blr entry: %and = and <2 x i64> %B, %C @@ -52,11 +48,9 @@ define <16 x i8> @ternary_A_and_BC_not_B_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_and_BC_not_B_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxland vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 193 ; CHECK-NEXT: blr entry: %and = and <16 x i8> %B, %C @@ -70,11 +64,9 @@ define <8 x i16> @ternary_A_and_BC_not_B_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i1 ; CHECK-LABEL: ternary_A_and_BC_not_B_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxland vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 193 ; CHECK-NEXT: blr entry: %and = and <8 x i16> %B, %C @@ -88,11 +80,9 @@ define <4 x i32> @ternary_A_xor_BC_not_B_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i3 ; CHECK-LABEL: ternary_A_xor_BC_not_B_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 198 ; CHECK-NEXT: blr entry: %xor = xor <4 x i32> %B, %C @@ -106,12 +96,10 @@ define <2 x i64> @ternary_A_xor_BC_not_B_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i6 ; CHECK-LABEL: ternary_A_xor_BC_not_B_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 198 ; CHECK-NEXT: blr entry: %xor = xor <2 x i64> %B, %C @@ -125,11 +113,9 @@ define <16 x i8> @ternary_A_xor_BC_not_B_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_xor_BC_not_B_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 198 ; CHECK-NEXT: blr entry: %xor = xor <16 x i8> %B, %C @@ -143,11 +129,9 @@ define <8 x i16> @ternary_A_xor_BC_not_B_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i1 ; CHECK-LABEL: ternary_A_xor_BC_not_B_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 198 ; CHECK-NEXT: blr entry: %xor = xor <8 x i16> %B, %C @@ -161,11 +145,9 @@ define <4 x i32> @ternary_A_or_BC_not_B_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i32 ; CHECK-LABEL: ternary_A_or_BC_not_B_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 199 ; CHECK-NEXT: blr entry: %or = or <4 x i32> %B, %C @@ -179,12 +161,10 @@ define <2 x i64> @ternary_A_or_BC_not_B_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i64 ; CHECK-LABEL: ternary_A_or_BC_not_B_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 199 ; CHECK-NEXT: blr entry: %or = or <2 x i64> %B, %C @@ -198,11 +178,9 @@ define <16 x i8> @ternary_A_or_BC_not_B_16x8(<16 x i1> %A, <16 x i8> %B, <16 x i ; CHECK-LABEL: ternary_A_or_BC_not_B_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 199 ; CHECK-NEXT: blr entry: %or = or <16 x i8> %B, %C @@ -216,11 +194,9 @@ define <8 x i16> @ternary_A_or_BC_not_B_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i16 ; CHECK-LABEL: ternary_A_or_BC_not_B_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 199 ; CHECK-NEXT: blr entry: %or = or <8 x i16> %B, %C @@ -234,11 +210,9 @@ define <4 x i32> @ternary_A_nand_BC_not_B_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i ; CHECK-LABEL: ternary_A_nand_BC_not_B_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 206 ; CHECK-NEXT: blr entry: %and = and <4 x i32> %B, %C @@ -253,12 +227,10 @@ define <2 x i64> @ternary_A_nand_BC_not_B_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i ; CHECK-LABEL: ternary_A_nand_BC_not_B_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 206 ; CHECK-NEXT: blr entry: %and = and <2 x i64> %B, %C @@ -273,11 +245,9 @@ define <16 x i8> @ternary_A_nand_BC_not_B_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_nand_BC_not_B_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 206 ; CHECK-NEXT: blr entry: %and = and <16 x i8> %B, %C @@ -292,11 +262,9 @@ define <8 x i16> @ternary_A_nand_BC_not_B_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i ; CHECK-LABEL: ternary_A_nand_BC_not_B_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v3, v3 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 206 ; CHECK-NEXT: blr entry: %and = and <8 x i16> %B, %C diff --git a/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-not-c.ll b/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-not-c.ll index 3479d94..98c1f28 100644 --- a/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-not-c.ll +++ b/llvm/test/CodeGen/PowerPC/xxeval-vselect-x-not-c.ll @@ -15,11 +15,9 @@ define <4 x i32> @ternary_A_and_BC_not_C_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i3 ; CHECK-LABEL: ternary_A_and_BC_not_C_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxland vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 161 ; CHECK-NEXT: blr entry: %and = and <4 x i32> %B, %C @@ -33,12 +31,10 @@ define <2 x i64> @ternary_A_and_BC_not_C_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i6 ; CHECK-LABEL: ternary_A_and_BC_not_C_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxland vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 161 ; CHECK-NEXT: blr entry: %and = and <2 x i64> %B, %C @@ -52,11 +48,9 @@ define <16 x i8> @ternary_A_and_BC_not_C_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_and_BC_not_C_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxland vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 161 ; CHECK-NEXT: blr entry: %and = and <16 x i8> %B, %C @@ -70,11 +64,9 @@ define <8 x i16> @ternary_A_and_BC_not_C_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i1 ; CHECK-LABEL: ternary_A_and_BC_not_C_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxland vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 161 ; CHECK-NEXT: blr entry: %and = and <8 x i16> %B, %C @@ -88,10 +80,9 @@ define <4 x i32> @ternary_A_B_not_C_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i32> %C ; CHECK-LABEL: ternary_A_B_not_C_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v4, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v3, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 163 ; CHECK-NEXT: blr entry: %not = xor <4 x i32> %C, <i32 -1, i32 -1, i32 -1, i32 -1> ; Vector not operation @@ -104,11 +95,10 @@ define <2 x i64> @ternary_A_B_not_C_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i64> %C ; CHECK-LABEL: ternary_A_B_not_C_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v4, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v3, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 163 ; CHECK-NEXT: blr entry: %not = xor <2 x i64> %C, <i64 -1, i64 -1> ; Vector not operation @@ -121,10 +111,9 @@ define <16 x i8> @ternary_A_B_not_C_16x8(<16 x i1> %A, <16 x i8> %B, <16 x i8> % ; CHECK-LABEL: ternary_A_B_not_C_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnor vs0, v4, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v3, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 163 ; CHECK-NEXT: blr entry: %not = xor <16 x i8> %C, <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1> ; Vector not operation @@ -137,10 +126,9 @@ define <8 x i16> @ternary_A_B_not_C_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i16> %C ; CHECK-LABEL: ternary_A_B_not_C_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnor vs0, v4, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs0, v3, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 163 ; CHECK-NEXT: blr entry: %not = xor <8 x i16> %C, <i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1> ; Vector not operation @@ -153,11 +141,9 @@ define <4 x i32> @ternary_A_xor_BC_not_C_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i3 ; CHECK-LABEL: ternary_A_xor_BC_not_C_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 166 ; CHECK-NEXT: blr entry: %xor = xor <4 x i32> %B, %C @@ -171,12 +157,10 @@ define <2 x i64> @ternary_A_xor_BC_not_C_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i6 ; CHECK-LABEL: ternary_A_xor_BC_not_C_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 166 ; CHECK-NEXT: blr entry: %xor = xor <2 x i64> %B, %C @@ -190,11 +174,9 @@ define <16 x i8> @ternary_A_xor_BC_not_C_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_xor_BC_not_C_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 166 ; CHECK-NEXT: blr entry: %xor = xor <16 x i8> %B, %C @@ -208,11 +190,9 @@ define <8 x i16> @ternary_A_xor_BC_not_C_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i1 ; CHECK-LABEL: ternary_A_xor_BC_not_C_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlxor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 166 ; CHECK-NEXT: blr entry: %xor = xor <8 x i16> %B, %C @@ -226,11 +206,9 @@ define <4 x i32> @ternary_A_or_BC_not_C_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i32 ; CHECK-LABEL: ternary_A_or_BC_not_C_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 167 ; CHECK-NEXT: blr entry: %or = or <4 x i32> %B, %C @@ -244,12 +222,10 @@ define <2 x i64> @ternary_A_or_BC_not_C_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i64 ; CHECK-LABEL: ternary_A_or_BC_not_C_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 167 ; CHECK-NEXT: blr entry: %or = or <2 x i64> %B, %C @@ -263,11 +239,9 @@ define <16 x i8> @ternary_A_or_BC_not_C_16x8(<16 x i1> %A, <16 x i8> %B, <16 x i ; CHECK-LABEL: ternary_A_or_BC_not_C_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 167 ; CHECK-NEXT: blr entry: %or = or <16 x i8> %B, %C @@ -281,11 +255,9 @@ define <8 x i16> @ternary_A_or_BC_not_C_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i16 ; CHECK-LABEL: ternary_A_or_BC_not_C_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlor vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 167 ; CHECK-NEXT: blr entry: %or = or <8 x i16> %B, %C @@ -299,11 +271,9 @@ define <4 x i32> @ternary_A_not_B_not_C_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i32 ; CHECK-LABEL: ternary_A_not_B_not_C_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v3, v3 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 172 ; CHECK-NEXT: blr entry: %not_b = xor <4 x i32> %B, <i32 -1, i32 -1, i32 -1, i32 -1> ; Vector not operation @@ -317,12 +287,10 @@ define <2 x i64> @ternary_A_not_B_not_C_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i64 ; CHECK-LABEL: ternary_A_not_B_not_C_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnor vs0, v3, v3 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 172 ; CHECK-NEXT: blr entry: %not_b = xor <2 x i64> %B, <i64 -1, i64 -1> ; Vector not operation @@ -336,11 +304,9 @@ define <16 x i8> @ternary_A_not_B_not_C_16x8(<16 x i1> %A, <16 x i8> %B, <16 x i ; CHECK-LABEL: ternary_A_not_B_not_C_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnor vs0, v3, v3 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 172 ; CHECK-NEXT: blr entry: %not_b = xor <16 x i8> %B, <i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1, i8 -1> ; Vector not operation @@ -354,11 +320,9 @@ define <8 x i16> @ternary_A_not_B_not_C_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i16 ; CHECK-LABEL: ternary_A_not_B_not_C_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnor vs0, v3, v3 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 172 ; CHECK-NEXT: blr entry: %not_b = xor <8 x i16> %B, <i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1, i16 -1> ; Vector not operation @@ -372,11 +336,9 @@ define <4 x i32> @ternary_A_nand_BC_not_C_4x32(<4 x i1> %A, <4 x i32> %B, <4 x i ; CHECK-LABEL: ternary_A_nand_BC_not_C_4x32: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxleqv v5, v5, v5 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslw v2, v2, v5 ; CHECK-NEXT: vsraw v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 174 ; CHECK-NEXT: blr entry: %and = and <4 x i32> %B, %C @@ -391,12 +353,10 @@ define <2 x i64> @ternary_A_nand_BC_not_C_2x64(<2 x i1> %A, <2 x i64> %B, <2 x i ; CHECK-LABEL: ternary_A_nand_BC_not_C_2x64: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxlxor v5, v5, v5 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: xxsplti32dx v5, 1, 63 ; CHECK-NEXT: vsld v2, v2, v5 ; CHECK-NEXT: vsrad v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 174 ; CHECK-NEXT: blr entry: %and = and <2 x i64> %B, %C @@ -411,11 +371,9 @@ define <16 x i8> @ternary_A_nand_BC_not_C_16x8(<16 x i1> %A, <16 x i8> %B, <16 x ; CHECK-LABEL: ternary_A_nand_BC_not_C_16x8: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltib v5, 7 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslb v2, v2, v5 ; CHECK-NEXT: vsrab v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 174 ; CHECK-NEXT: blr entry: %and = and <16 x i8> %B, %C @@ -430,11 +388,9 @@ define <8 x i16> @ternary_A_nand_BC_not_C_8x16(<8 x i1> %A, <8 x i16> %B, <8 x i ; CHECK-LABEL: ternary_A_nand_BC_not_C_8x16: ; CHECK: # %bb.0: # %entry ; CHECK-NEXT: xxspltiw v5, 983055 -; CHECK-NEXT: xxlnand vs0, v3, v4 -; CHECK-NEXT: xxlnor vs1, v4, v4 ; CHECK-NEXT: vslh v2, v2, v5 ; CHECK-NEXT: vsrah v2, v2, v5 -; CHECK-NEXT: xxsel v2, vs1, vs0, v2 +; CHECK-NEXT: xxeval v2, v2, v3, v4, 174 ; CHECK-NEXT: blr entry: %and = and <8 x i16> %B, %C diff --git a/llvm/test/CodeGen/RISCV/branch-rel.mir b/llvm/test/CodeGen/RISCV/branch-rel.mir new file mode 100644 index 0000000..1ed5f57 --- /dev/null +++ b/llvm/test/CodeGen/RISCV/branch-rel.mir @@ -0,0 +1,39 @@ +# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 5 +# RUN: llc %s -mtriple=riscv64 -run-pass=branch-relaxation -o - -verify-machineinstrs | FileCheck %s + +--- | + define void @foo() { + ret void + } +... +--- +name: foo +tracksRegLiveness: true +body: | + ; CHECK-LABEL: name: foo + ; CHECK: bb.0: + ; CHECK-NEXT: successors: %bb.2(0x80000000) + ; CHECK-NEXT: liveins: $x1 + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: PseudoBR %bb.2 + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: bb.1: + ; CHECK-NEXT: successors: %bb.2(0x80000000) + ; CHECK-NEXT: liveins: $x1 + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: INLINEASM &".space 4096", 1 /* sideeffect attdialect */ + ; CHECK-NEXT: BGE $x1, $x0, %bb.2 + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: bb.2: + ; CHECK-NEXT: PseudoRET + bb.0: + liveins: $x1 + BNE $x1, $x0, %bb.3 + PseudoBR %bb.3 + bb.1: + liveins: $x1 + INLINEASM &".space 4096", 1 + BGE $x1, $x0, %bb.3 + bb.3: + PseudoRET +## NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line: diff --git a/llvm/test/CodeGen/RISCV/rvv/regcoal-liveinterval-pruning-crash.ll b/llvm/test/CodeGen/RISCV/rvv/regcoal-liveinterval-pruning-crash.ll new file mode 100644 index 0000000..c19e93d --- /dev/null +++ b/llvm/test/CodeGen/RISCV/rvv/regcoal-liveinterval-pruning-crash.ll @@ -0,0 +1,76 @@ +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5 +; RUN: llc -O1 -mtriple=riscv64 -mattr=+v < %s | FileCheck %s + +define i32 @pr134424(i64 %input_value, i32 %base_value, i1 %cond_flag1, i1 %cond_flag2, i1 %cond_flag3) { +; CHECK-LABEL: pr134424: +; CHECK: # %bb.0: # %for.body.us.preheader.i +; CHECK-NEXT: andi a3, a3, 1 +; CHECK-NEXT: andi a5, a2, 1 +; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, ma +; CHECK-NEXT: vmv.v.x v8, a0 +; CHECK-NEXT: vsetvli zero, zero, e64, m2, tu, ma +; CHECK-NEXT: vmv.s.x v8, zero +; CHECK-NEXT: vsetivli zero, 1, e8, mf8, ta, ma +; CHECK-NEXT: vmv.v.i v0, 14 +; CHECK-NEXT: mv a2, a1 +; CHECK-NEXT: bnez a5, .LBB0_2 +; CHECK-NEXT: # %bb.1: # %for.body.us.preheader.i +; CHECK-NEXT: li a2, 1 +; CHECK-NEXT: .LBB0_2: # %for.body.us.preheader.i +; CHECK-NEXT: vsetivli zero, 4, e64, m2, ta, ma +; CHECK-NEXT: vmerge.vxm v8, v8, a0, v0 +; CHECK-NEXT: andi a4, a4, 1 +; CHECK-NEXT: mv a0, a1 +; CHECK-NEXT: bnez a3, .LBB0_4 +; CHECK-NEXT: # %bb.3: # %for.body.us.preheader.i +; CHECK-NEXT: li a0, 1 +; CHECK-NEXT: .LBB0_4: # %for.body.us.preheader.i +; CHECK-NEXT: vmsle.vi v0, v8, 0 +; CHECK-NEXT: sext.w a2, a2 +; CHECK-NEXT: bnez a4, .LBB0_6 +; CHECK-NEXT: # %bb.5: # %for.body.us.preheader.i +; CHECK-NEXT: li a1, 1 +; CHECK-NEXT: .LBB0_6: # %for.body.us.preheader.i +; CHECK-NEXT: sext.w a0, a0 +; CHECK-NEXT: vsetvli zero, zero, e32, m1, ta, ma +; CHECK-NEXT: vmv.v.i v8, 0 +; CHECK-NEXT: vmerge.vim v8, v8, 1, v0 +; CHECK-NEXT: vsetivli zero, 8, e32, m2, ta, ma +; CHECK-NEXT: vredmin.vs v8, v8, v8 +; CHECK-NEXT: vmv.x.s a3, v8 +; CHECK-NEXT: sext.w a1, a1 +; CHECK-NEXT: bge a3, a2, .LBB0_11 +; CHECK-NEXT: # %bb.7: # %for.body.us.preheader.i +; CHECK-NEXT: bge a0, a1, .LBB0_12 +; CHECK-NEXT: .LBB0_8: # %for.body.us.preheader.i +; CHECK-NEXT: blt a3, a0, .LBB0_10 +; CHECK-NEXT: .LBB0_9: # %for.body.us.preheader.i +; CHECK-NEXT: mv a3, a0 +; CHECK-NEXT: .LBB0_10: # %for.body.us.preheader.i +; CHECK-NEXT: sw a3, 0(zero) +; CHECK-NEXT: li a0, 0 +; CHECK-NEXT: ret +; CHECK-NEXT: .LBB0_11: # %for.body.us.preheader.i +; CHECK-NEXT: mv a3, a2 +; CHECK-NEXT: blt a0, a1, .LBB0_8 +; CHECK-NEXT: .LBB0_12: # %for.body.us.preheader.i +; CHECK-NEXT: mv a0, a1 +; CHECK-NEXT: bge a3, a0, .LBB0_9 +; CHECK-NEXT: j .LBB0_10 +for.body.us.preheader.i: + %partial_vector = insertelement <4 x i64> zeroinitializer, i64 %input_value, i64 1 + %comparison_vector = shufflevector <4 x i64> %partial_vector, <4 x i64> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 1, i32 1> + %comparison_result = icmp sle <4 x i64> %comparison_vector, zeroinitializer + %selected_value1 = select i1 %cond_flag1, i32 %base_value, i32 1 + %selected_value2 = select i1 %cond_flag2, i32 %base_value, i32 1 + %selected_value3 = select i1 %cond_flag3, i32 %base_value, i32 1 + %bool_to_int = zext <4 x i1> %comparison_result to <4 x i32> + %extended_vector = shufflevector <4 x i32> %bool_to_int, <4 x i32> zeroinitializer, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison> + %vector_min = call i32 @llvm.vector.reduce.smin.v8i32(<8 x i32> %extended_vector) + %min1 = call i32 @llvm.smin.i32(i32 %vector_min, i32 %selected_value1) + %min2 = call i32 @llvm.smin.i32(i32 %selected_value2, i32 %selected_value3) + %final_min = call i32 @llvm.smin.i32(i32 %min1, i32 %min2) + store i32 %final_min, ptr null, align 4 + ret i32 0 +} + diff --git a/llvm/test/CodeGen/RISCV/rvv/regcoal-liveinterval-pruning-crash.mir b/llvm/test/CodeGen/RISCV/rvv/regcoal-liveinterval-pruning-crash.mir new file mode 100644 index 0000000..aeab8f6 --- /dev/null +++ b/llvm/test/CodeGen/RISCV/rvv/regcoal-liveinterval-pruning-crash.mir @@ -0,0 +1,57 @@ +# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py UTC_ARGS: --version 5 +# RUN: llc -mtriple=riscv64 -mattr=+v -run-pass=register-coalescer -o - %s | FileCheck %s + +--- +name: pr71023 +tracksRegLiveness: true +body: | + ; CHECK-LABEL: name: pr71023 + ; CHECK: bb.0: + ; CHECK-NEXT: successors: %bb.3(0x40000000), %bb.1(0x40000000) + ; CHECK-NEXT: liveins: $x10, $v8, $v10 + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: dead [[DEF:%[0-9]+]]:gpr = IMPLICIT_DEF + ; CHECK-NEXT: undef [[PseudoVMV_V_I_M1_:%[0-9]+]].sub_vrm1_2:vrn8m1 = PseudoVMV_V_I_M1 undef [[PseudoVMV_V_I_M1_]].sub_vrm1_2, 0, -1, 3 /* e8 */, 0 /* tu, mu */, implicit $vl, implicit $vtype + ; CHECK-NEXT: [[PseudoVMV_V_I_M1_:%[0-9]+]].sub_vrm1_6:vrn8m1 = COPY undef [[PseudoVMV_V_I_M1_]].sub_vrm1_2 + ; CHECK-NEXT: BNE undef [[DEF]], $x0, %bb.3 + ; CHECK-NEXT: PseudoBR %bb.1 + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: bb.1: + ; CHECK-NEXT: successors: %bb.3(0x40000000), %bb.2(0x40000000) + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: BNE undef [[DEF]], $x0, %bb.3 + ; CHECK-NEXT: PseudoBR %bb.2 + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: bb.2: + ; CHECK-NEXT: successors: %bb.3(0x80000000) + ; CHECK-NEXT: {{ $}} + ; CHECK-NEXT: bb.3: + ; CHECK-NEXT: dead [[DEF1:%[0-9]+]]:vr = IMPLICIT_DEF + ; CHECK-NEXT: early-clobber [[PseudoVMV_V_I_M1_]].sub_vrm1_0:vrn8m1 = PseudoVRGATHER_VI_M1 undef [[PseudoVMV_V_I_M1_]].sub_vrm1_0, [[PseudoVMV_V_I_M1_]].sub_vrm1_2, 0, 0, 3 /* e8 */, 0 /* tu, mu */, implicit $vl, implicit $vtype + ; CHECK-NEXT: PseudoVSSEG6E8_V_M1_MASK [[PseudoVMV_V_I_M1_]].sub_vrm1_0_sub_vrm1_1_sub_vrm1_2_sub_vrm1_3_sub_vrm1_4_sub_vrm1_5, undef [[DEF]], killed undef $v0, 0, 3 /* e8 */, implicit $vl, implicit $vtype :: (store unknown-size, align 1) + ; CHECK-NEXT: PseudoRET + bb.0: + successors: %bb.3(0x40000000), %bb.1(0x40000000) + liveins: $x10, $v8, $v10 + %0:gpr = IMPLICIT_DEF + %1:vrnov0 = PseudoVMV_V_I_M1 undef %1, 0, -1, 3 /* e8 */, 0 /* tu, mu */, implicit $vl, implicit $vtype + %2:vrnov0 = IMPLICIT_DEF + undef %3.sub_vrm1_0:vrn6m1nov0 = COPY undef %1 + %3.sub_vrm1_3:vrn6m1nov0 = COPY %2 + %3.sub_vrm1_4:vrn6m1nov0 = COPY undef %1 + BNE undef %0, $x0, %bb.3 + PseudoBR %bb.1 + bb.1: + successors: %bb.3(0x40000000), %bb.2(0x40000000) + BNE killed undef %0, $x0, %bb.3 + PseudoBR %bb.2 + bb.2: + successors: %bb.3(0x80000000) + bb.3: + %4:vr = IMPLICIT_DEF + early-clobber %4:vr = PseudoVRGATHER_VI_M1 undef %4, killed %1, 0, 0, 3 /* e8 */, 0 /* tu, mu */, implicit $vl, implicit $vtype + undef %5.sub_vrm1_0:vrn6m1 = COPY killed %4 + %5.sub_vrm1_5:vrn6m1 = COPY killed %2 + PseudoVSSEG6E8_V_M1_MASK killed %5, undef %0, killed undef $v0, 0, 3 /* e8 */, implicit $vl, implicit $vtype :: (store unknown-size, align 1) + PseudoRET +... diff --git a/llvm/test/CodeGen/SPIRV/FCmpFalse.ll b/llvm/test/CodeGen/SPIRV/FCmpFalse.ll new file mode 100644 index 0000000..55d64196 --- /dev/null +++ b/llvm/test/CodeGen/SPIRV/FCmpFalse.ll @@ -0,0 +1,10 @@ +; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv64-unknown-unknown %s -o - | FileCheck %s +; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv32-unknown-unknown %s -o - -filetype=obj | spirv-val %} + +; CHECK: %[[#FalseVal:]] = OpConstantFalse %[[#]] +; CHECK: OpReturnValue %[[#FalseVal:]] + +define spir_func i1 @f(float %0) { + %2 = fcmp false float %0, %0 + ret i1 %2 +} diff --git a/llvm/test/CodeGen/SPIRV/FCmpFalse_Vec.ll b/llvm/test/CodeGen/SPIRV/FCmpFalse_Vec.ll new file mode 100644 index 0000000..c410b64 --- /dev/null +++ b/llvm/test/CodeGen/SPIRV/FCmpFalse_Vec.ll @@ -0,0 +1,13 @@ +; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv64-unknown-unknown %s -o - | FileCheck %s +; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv64-unknown-unknown %s -o - -filetype=obj | spirv-val %} + +; CHECK: %[[#BoolTy:]] = OpTypeBool +; CHECK: %[[#VecTy:]] = OpTypeVector %[[#BoolTy]] 4 +; CHECK: %[[#False:]] = OpConstantFalse %[[#BoolTy]] +; CHECK: %[[#Composite:]] = OpConstantComposite %[[#VecTy]] %[[#False]] %[[#False]] %[[#False]] %[[#False]] +; CHECK: OpReturnValue %[[#Composite]] + +define spir_func <4 x i1> @test(<4 x float> %a) { + %compare = fcmp false <4 x float> %a, %a + ret <4 x i1> %compare +} diff --git a/llvm/test/CodeGen/SPIRV/builtin_duplicate.ll b/llvm/test/CodeGen/SPIRV/builtin_duplicate.ll new file mode 100644 index 0000000..8786554 --- /dev/null +++ b/llvm/test/CodeGen/SPIRV/builtin_duplicate.ll @@ -0,0 +1,20 @@ +;; This test checks if we generate a single builtin variable for the following +;; LLVM IR. +;; @__spirv_BuiltInLocalInvocationId - A global variable +;; %3 = tail call i64 @_Z12get_local_idj(i32 0) - A function call + +; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv64-unknown-unknown %s -o - | FileCheck %s +; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv64-unknown-unknown %s -o - -filetype=obj | spirv-val %} + +; CHECK: OpName %[[#]] "__spirv_BuiltInLocalInvocationId" +; CHECK-NOT: OpName %[[#]] "__spirv_BuiltInLocalInvocationId.1" + +@__spirv_BuiltInLocalInvocationId = external dso_local local_unnamed_addr addrspace(1) constant <3 x i64>, align 32 + +declare spir_func i64 @_Z12get_local_idj(i32) local_unnamed_addr + +define spir_kernel void @test(i32 %a) { +entry: + %builtin_call = tail call i64 @_Z12get_local_idj(i32 0) + ret void +} diff --git a/llvm/test/CodeGen/SPIRV/complex-constexpr.ll b/llvm/test/CodeGen/SPIRV/complex-constexpr.ll new file mode 100644 index 0000000..e2c1d00 --- /dev/null +++ b/llvm/test/CodeGen/SPIRV/complex-constexpr.ll @@ -0,0 +1,21 @@ +; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv64-unknown-unknown %s -o - | FileCheck %s +; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv64-unknown-unknown %s -o - -filetype=obj | spirv-val %} + +@.str.1 = private unnamed_addr addrspace(1) constant [1 x i8] zeroinitializer, align 1 + +define linkonce_odr hidden spir_func void @test() { +entry: +; CHECK: %[[#MinusOne:]] = OpConstant %[[#]] 18446744073709551615 +; CHECK: %[[#Ptr:]] = OpConvertUToPtr %[[#]] %[[#MinusOne]] +; CHECK: %[[#PtrCast:]] = OpPtrCastToGeneric %[[#]] %[[#]] +; CHECK: %[[#]] = OpFunctionCall %[[#]] %[[#]] %[[#PtrCast]] %[[#Ptr]] + + %cast = bitcast ptr addrspace(4) inttoptr (i64 -1 to ptr addrspace(4)) to ptr addrspace(4) + call spir_func void @bar(ptr addrspace(4) addrspacecast (ptr addrspace(1) @.str.1 to ptr addrspace(4)), ptr addrspace(4) %cast) + ret void +} + +define linkonce_odr hidden spir_func void @bar(ptr addrspace(4) %begin, ptr addrspace(4) %end) { +entry: + ret void +} diff --git a/llvm/test/CodeGen/SPIRV/dominator-order.ll b/llvm/test/CodeGen/SPIRV/dominator-order.ll new file mode 100644 index 0000000..2ecdddc --- /dev/null +++ b/llvm/test/CodeGen/SPIRV/dominator-order.ll @@ -0,0 +1,25 @@ +; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv64-unknown-unknown %s -o - | FileCheck %s +; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv64-unknown-unknown %s -o - -filetype=obj | spirv-val %} + +; This test checks that basic blocks are reordered in SPIR-V so that dominators +; are emitted ahead of their dominated blocks as required by the SPIR-V +; specification. + +; CHECK-DAG: OpName %[[#ENTRY:]] "entry" +; CHECK-DAG: OpName %[[#FOR_BODY137_LR_PH:]] "for.body137.lr.ph" +; CHECK-DAG: OpName %[[#FOR_BODY:]] "for.body" + +; CHECK: %[[#ENTRY]] = OpLabel +; CHECK: %[[#FOR_BODY]] = OpLabel +; CHECK: %[[#FOR_BODY137_LR_PH]] = OpLabel + +define spir_kernel void @test(ptr addrspace(1) %arg, i1 %cond) { +entry: + br label %for.body + +for.body137.lr.ph: ; preds = %for.body + ret void + +for.body: ; preds = %for.body, %entry + br i1 %cond, label %for.body, label %for.body137.lr.ph +} diff --git a/llvm/test/CodeGen/SPIRV/llvm-intrinsics/fake_use.ll b/llvm/test/CodeGen/SPIRV/llvm-intrinsics/fake_use.ll new file mode 100644 index 0000000..5370b51 --- /dev/null +++ b/llvm/test/CodeGen/SPIRV/llvm-intrinsics/fake_use.ll @@ -0,0 +1,13 @@ +; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv64-unknown-unknown %s -o - | FileCheck %s +; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv64-unknown-unknown %s -o - -filetype=obj | spirv-val %} + +; CHECK-DAG: OpCapability Addresses +; CHECK-DAG: OpName %[[#]] "foo" + +declare void @llvm.fake.use(...) + +define spir_kernel void @foo(ptr addrspace(1) %a) { +entry: + call void (...) @llvm.fake.use(ptr addrspace(1) %a) + ret void +} diff --git a/llvm/test/CodeGen/SPIRV/transcoding/AtomicCompareExchange_cl20.ll b/llvm/test/CodeGen/SPIRV/transcoding/AtomicCompareExchange_cl20.ll new file mode 100644 index 0000000..8357373 --- /dev/null +++ b/llvm/test/CodeGen/SPIRV/transcoding/AtomicCompareExchange_cl20.ll @@ -0,0 +1,84 @@ +; RUN: llc -verify-machineinstrs -O0 -mtriple=spirv64-unknown-unknown %s -o - | FileCheck %s +; RUN: %if spirv-tools %{ llc -O0 -mtriple=spirv64v1.2-unknown-unknown %s -o - -filetype=obj | spirv-val %} + +; CHECK-NOT: OpCapability Int64Atomics + +; CHECK-DAG: %[[#int:]] = OpTypeInt 32 0 +; CHECK-DAG: %[[#int8:]] = OpTypeInt 8 0 +; CHECK-DAG: %[[#DeviceScope:]] = OpConstant %[[#int]] 1 +; CHECK-DAG: %[[#SequentiallyConsistent_MS:]] = OpConstant %[[#int]] 16 +; CHECK-DAG: %[[#int_ptr:]] = OpTypePointer Generic %[[#int]] +; CHECK-DAG: %[[#int_ptr8:]] = OpTypePointer Generic %[[#int8]] +; CHECK-DAG: %[[#bool:]] = OpTypeBool + +define spir_func void @test(ptr addrspace(4) %object, ptr addrspace(4) %expected, i32 %desired) { + +; CHECK: %[[#object:]] = OpFunctionParameter %[[#int_ptr8]] +; CHECK: %[[#expected:]] = OpFunctionParameter %[[#int_ptr8]] +; CHECK: %[[#desired:]] = OpFunctionParameter %[[#int]] + +entry: + %object.addr = alloca ptr addrspace(4), align 4 + %expected.addr = alloca ptr addrspace(4), align 4 + %desired.addr = alloca i32, align 4 + %strong_res = alloca i8, align 1 + %res = alloca i8, align 1 + %weak_res = alloca i8, align 1 + store ptr addrspace(4) %object, ptr %object.addr, align 4 + store ptr addrspace(4) %expected, ptr %expected.addr, align 4 + store i32 %desired, ptr %desired.addr, align 4 + %0 = load ptr addrspace(4), ptr %object.addr, align 4 + %1 = load ptr addrspace(4), ptr %expected.addr, align 4 + %2 = load i32, ptr %desired.addr, align 4 + +; CHECK-DAG: OpStore %[[#object_addr:]] %[[#object]] +; CHECK-DAG: OpStore %[[#expected_addr:]] %[[#expected]] +; CHECK-DAG: OpStore %[[#desired_addr:]] %[[#desired]] + +; CHECK: %[[#Pointer:]] = OpLoad %[[#int_ptr]] %[[#]] +; CHECK: %[[#exp:]] = OpLoad %[[#int_ptr]] %[[#]] +; CHECK: %[[#Value:]] = OpLoad %[[#int]] %[[#desired_addr]] +; CHECK: %[[#Comparator:]] = OpLoad %[[#int]] %[[#exp]] + +; CHECK: %[[#Result:]] = OpAtomicCompareExchange %[[#int]] %[[#]] %[[#DeviceScope]] %[[#SequentiallyConsistent_MS]] %[[#SequentiallyConsistent_MS]] %[[#Value]] %[[#Comparator]] + %call = call spir_func zeroext i1 @_Z30atomic_compare_exchange_strongPVU3AS4U7_AtomiciPU3AS4ii(ptr addrspace(4) %0, ptr addrspace(4) %1, i32 %2) + +; CHECK-NEXT: OpStore %[[#exp]] %[[#Result]] +; CHECK-NEXT: %[[#CallRes:]] = OpIEqual %[[#bool]] %[[#Result]] %[[#Comparator]] +; CHECK-NOT: %[[#Result]] + + %frombool = zext i1 %call to i8 + store i8 %frombool, ptr %strong_res, align 1 + %3 = load i8, ptr %strong_res, align 1 + %tobool = trunc i8 %3 to i1 + %lnot = xor i1 %tobool, true + %frombool1 = zext i1 %lnot to i8 + store i8 %frombool1, ptr %res, align 1 + %4 = load ptr addrspace(4), ptr %object.addr, align 4 + %5 = load ptr addrspace(4), ptr %expected.addr, align 4 + %6 = load i32, ptr %desired.addr, align 4 + +; CHECK: %[[#Pointer:]] = OpLoad %[[#int_ptr]] %[[#]] +; CHECK: %[[#exp:]] = OpLoad %[[#int_ptr]] %[[#]] +; CHECK: %[[#Value:]] = OpLoad %[[#int]] %[[#desired_addr]] +; CHECK: %[[#ComparatorWeak:]] = OpLoad %[[#int]] %[[#exp]] + +; CHECK: %[[#Result:]] = OpAtomicCompareExchangeWeak %[[#int]] %[[#]] %[[#DeviceScope]] %[[#SequentiallyConsistent_MS]] %[[#SequentiallyConsistent_MS]] %[[#Value]] %[[#ComparatorWeak]] + %call2 = call spir_func zeroext i1 @_Z28atomic_compare_exchange_weakPVU3AS4U7_AtomiciPU3AS4ii(ptr addrspace(4) %4, ptr addrspace(4) %5, i32 %6) + +; CHECK-NEXT: OpStore %[[#exp]] %[[#Result]] +; CHECK-NEXT: %[[#CallRes:]] = OpIEqual %[[#bool]] %[[#Result]] %[[#ComparatorWeak]] +; CHECK-NOT: %[[#Result]] + + %frombool3 = zext i1 %call2 to i8 + store i8 %frombool3, ptr %weak_res, align 1 + %7 = load i8, ptr %weak_res, align 1 + %tobool4 = trunc i8 %7 to i1 + %lnot5 = xor i1 %tobool4, true + %frombool6 = zext i1 %lnot5 to i8 + store i8 %frombool6, ptr %res, align 1 + ret void +} + +declare spir_func zeroext i1 @_Z30atomic_compare_exchange_strongPVU3AS4U7_AtomiciPU3AS4ii(ptr addrspace(4), ptr addrspace(4), i32) #1 +declare spir_func zeroext i1 @_Z28atomic_compare_exchange_weakPVU3AS4U7_AtomiciPU3AS4ii(ptr addrspace(4), ptr addrspace(4), i32) #1 diff --git a/llvm/test/CodeGen/Thumb2/LowOverheadLoops/tail-pred-basic.ll b/llvm/test/CodeGen/Thumb2/LowOverheadLoops/tail-pred-basic.ll index c2b4494..11e7e5c 100644 --- a/llvm/test/CodeGen/Thumb2/LowOverheadLoops/tail-pred-basic.ll +++ b/llvm/test/CodeGen/Thumb2/LowOverheadLoops/tail-pred-basic.ll @@ -1,16 +1,41 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 6 ; RUN: opt -mtriple=thumbv8.1m.main -mve-tail-predication -tail-predication=enabled -mattr=+mve,+lob %s -S -o - | FileCheck %s -; CHECK-LABEL: mul_v16i8 -; CHECK-NOT: %num.elements = add i32 %trip.count.minus.1, 1 -; CHECK: vector.body: -; CHECK: %index = phi i32 -; CHECK: [[ELEMS:%[^ ]+]] = phi i32 [ %N, %vector.ph ], [ [[REMAINING:%[^ ]+]], %vector.body ] -; CHECK: [[VCTP:%[^ ]+]] = call <16 x i1> @llvm.arm.mve.vctp8(i32 [[ELEMS]]) -; CHECK: [[REMAINING]] = sub i32 [[ELEMS]], 16 -; CHECK: [[LD0:%[^ ]+]] = tail call <16 x i8> @llvm.masked.load.v16i8.p0(ptr {{.*}}, i32 4, <16 x i1> [[VCTP]], <16 x i8> undef) -; CHECK: [[LD1:%[^ ]+]] = tail call <16 x i8> @llvm.masked.load.v16i8.p0(ptr {{.*}}, i32 4, <16 x i1> [[VCTP]], <16 x i8> undef) -; CHECK: tail call void @llvm.masked.store.v16i8.p0(<16 x i8> {{.*}}, ptr {{.*}}, i32 4, <16 x i1> [[VCTP]]) define dso_local arm_aapcs_vfpcc void @mul_v16i8(ptr noalias nocapture readonly %a, ptr noalias nocapture readonly %b, ptr noalias nocapture %c, i32 %N) { +; CHECK-LABEL: define dso_local arm_aapcs_vfpcc void @mul_v16i8( +; CHECK-SAME: ptr noalias readonly captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], ptr noalias captures(none) [[C:%.*]], i32 [[N:%.*]]) #[[ATTR0:[0-9]+]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp eq i32 [[N]], 0 +; CHECK-NEXT: [[TMP8:%.*]] = add i32 [[N]], 15 +; CHECK-NEXT: [[TMP9:%.*]] = lshr i32 [[TMP8]], 4 +; CHECK-NEXT: [[TMP10:%.*]] = shl nuw i32 [[TMP9]], 4 +; CHECK-NEXT: [[TMP11:%.*]] = add i32 [[TMP10]], -16 +; CHECK-NEXT: [[TMP12:%.*]] = lshr i32 [[TMP11]], 4 +; CHECK-NEXT: [[TMP13:%.*]] = add nuw nsw i32 [[TMP12]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[FOR_COND_CLEANUP:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP13]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP14:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP15:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP0:%.*]] = phi i32 [ [[N]], %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP:%.*]] = getelementptr inbounds i8, ptr [[A]], i32 [[INDEX]] +; CHECK-NEXT: [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.vctp8(i32 [[TMP0]]) +; CHECK-NEXT: [[TMP2]] = sub i32 [[TMP0]], 16 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = tail call <16 x i8> @llvm.masked.load.v16i8.p0(ptr [[TMP]], i32 4, <16 x i1> [[TMP1]], <16 x i8> undef) +; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i8, ptr [[B]], i32 [[INDEX]] +; CHECK-NEXT: [[WIDE_MASKED_LOAD2:%.*]] = tail call <16 x i8> @llvm.masked.load.v16i8.p0(ptr [[TMP3]], i32 4, <16 x i1> [[TMP1]], <16 x i8> undef) +; CHECK-NEXT: [[MUL:%.*]] = mul nsw <16 x i8> [[WIDE_MASKED_LOAD2]], [[WIDE_MASKED_LOAD]] +; CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds i8, ptr [[C]], i32 [[INDEX]] +; CHECK-NEXT: tail call void @llvm.masked.store.v16i8.p0(<16 x i8> [[MUL]], ptr [[TMP6]], i32 4, <16 x i1> [[TMP1]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 16 +; CHECK-NEXT: [[TMP15]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP14]], i32 1) +; CHECK-NEXT: [[TMP16:%.*]] = icmp ne i32 [[TMP15]], 0 +; CHECK-NEXT: br i1 [[TMP16]], label %[[VECTOR_BODY]], label %[[FOR_COND_CLEANUP]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp eq i32 %N, 0 %tmp8 = add i32 %N, 15 @@ -45,17 +70,41 @@ for.cond.cleanup: ; preds = %vector.body, %entry ret void } -; CHECK-LABEL: mul_v8i16 -; CHECK-NOT: %num.elements = add i32 %trip.count.minus.1, 1 -; CHECK: vector.body: -; CHECK: %index = phi i32 -; CHECK: [[ELEMS:%[^ ]+]] = phi i32 [ %N, %vector.ph ], [ [[REMAINING:%[^ ]+]], %vector.body ] -; CHECK: [[VCTP:%[^ ]+]] = call <8 x i1> @llvm.arm.mve.vctp16(i32 [[ELEMS]]) -; CHECK: [[REMAINING]] = sub i32 [[ELEMS]], 8 -; CHECK: [[LD0:%[^ ]+]] = tail call <8 x i16> @llvm.masked.load.v8i16.p0(ptr {{.*}}, i32 4, <8 x i1> [[VCTP]], <8 x i16> undef) -; CHECK: [[LD1:%[^ ]+]] = tail call <8 x i16> @llvm.masked.load.v8i16.p0(ptr {{.*}}, i32 4, <8 x i1> [[VCTP]], <8 x i16> undef) -; CHECK: tail call void @llvm.masked.store.v8i16.p0(<8 x i16> {{.*}}, ptr {{.*}}, i32 4, <8 x i1> [[VCTP]]) define dso_local arm_aapcs_vfpcc void @mul_v8i16(ptr noalias nocapture readonly %a, ptr noalias nocapture readonly %b, ptr noalias nocapture %c, i32 %N) { +; CHECK-LABEL: define dso_local arm_aapcs_vfpcc void @mul_v8i16( +; CHECK-SAME: ptr noalias readonly captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], ptr noalias captures(none) [[C:%.*]], i32 [[N:%.*]]) #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp eq i32 [[N]], 0 +; CHECK-NEXT: [[TMP8:%.*]] = add i32 [[N]], 7 +; CHECK-NEXT: [[TMP9:%.*]] = lshr i32 [[TMP8]], 3 +; CHECK-NEXT: [[TMP10:%.*]] = shl nuw i32 [[TMP9]], 3 +; CHECK-NEXT: [[TMP11:%.*]] = add i32 [[TMP10]], -8 +; CHECK-NEXT: [[TMP12:%.*]] = lshr i32 [[TMP11]], 3 +; CHECK-NEXT: [[TMP13:%.*]] = add nuw nsw i32 [[TMP12]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[FOR_COND_CLEANUP:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP13]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP14:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP15:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP0:%.*]] = phi i32 [ [[N]], %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP:%.*]] = getelementptr inbounds i16, ptr [[A]], i32 [[INDEX]] +; CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.vctp16(i32 [[TMP0]]) +; CHECK-NEXT: [[TMP2]] = sub i32 [[TMP0]], 8 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = tail call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[TMP]], i32 4, <8 x i1> [[TMP1]], <8 x i16> undef) +; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i16, ptr [[B]], i32 [[INDEX]] +; CHECK-NEXT: [[WIDE_MASKED_LOAD2:%.*]] = tail call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[TMP3]], i32 4, <8 x i1> [[TMP1]], <8 x i16> undef) +; CHECK-NEXT: [[MUL:%.*]] = mul nsw <8 x i16> [[WIDE_MASKED_LOAD2]], [[WIDE_MASKED_LOAD]] +; CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds i16, ptr [[C]], i32 [[INDEX]] +; CHECK-NEXT: tail call void @llvm.masked.store.v8i16.p0(<8 x i16> [[MUL]], ptr [[TMP6]], i32 4, <8 x i1> [[TMP1]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 8 +; CHECK-NEXT: [[TMP15]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP14]], i32 1) +; CHECK-NEXT: [[TMP16:%.*]] = icmp ne i32 [[TMP15]], 0 +; CHECK-NEXT: br i1 [[TMP16]], label %[[VECTOR_BODY]], label %[[FOR_COND_CLEANUP]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp eq i32 %N, 0 %tmp8 = add i32 %N, 7 @@ -90,16 +139,41 @@ for.cond.cleanup: ; preds = %vector.body, %entry ret void } -; CHECK-LABEL: mul_v4i32 -; CHECK-NOT: %num.elements = add i32 %trip.count.minus.1, 1 -; CHECK: vector.body: -; CHECK: [[ELEMS:%[^ ]+]] = phi i32 [ %N, %vector.ph ], [ [[REMAINING:%[^ ]+]], %vector.body ] -; CHECK: [[VCTP:%[^ ]+]] = call <4 x i1> @llvm.arm.mve.vctp32(i32 [[ELEMS]]) -; CHECK: [[REMAINING]] = sub i32 [[ELEMS]], 4 -; CHECK: [[LD0:%[^ ]+]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr {{.*}}, i32 4, <4 x i1> [[VCTP]], <4 x i32> undef) -; CHECK: [[LD1:%[^ ]+]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr {{.*}}, i32 4, <4 x i1> [[VCTP]], <4 x i32> undef) -; CHECK: tail call void @llvm.masked.store.v4i32.p0(<4 x i32> {{.*}}, ptr {{.*}}, i32 4, <4 x i1> [[VCTP]]) define dso_local arm_aapcs_vfpcc void @mul_v4i32(ptr noalias nocapture readonly %a, ptr noalias nocapture readonly %b, ptr noalias nocapture %c, i32 %N) { +; CHECK-LABEL: define dso_local arm_aapcs_vfpcc void @mul_v4i32( +; CHECK-SAME: ptr noalias readonly captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], ptr noalias captures(none) [[C:%.*]], i32 [[N:%.*]]) #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp eq i32 [[N]], 0 +; CHECK-NEXT: [[TMP8:%.*]] = add i32 [[N]], 3 +; CHECK-NEXT: [[TMP9:%.*]] = lshr i32 [[TMP8]], 2 +; CHECK-NEXT: [[TMP10:%.*]] = shl nuw i32 [[TMP9]], 2 +; CHECK-NEXT: [[TMP11:%.*]] = add i32 [[TMP10]], -4 +; CHECK-NEXT: [[TMP12:%.*]] = lshr i32 [[TMP11]], 2 +; CHECK-NEXT: [[TMP13:%.*]] = add nuw nsw i32 [[TMP12]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[FOR_COND_CLEANUP:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP13]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP14:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP15:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP0:%.*]] = phi i32 [ [[N]], %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[INDEX]] +; CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.vctp32(i32 [[TMP0]]) +; CHECK-NEXT: [[TMP2]] = sub i32 [[TMP0]], 4 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[TMP]], i32 4, <4 x i1> [[TMP1]], <4 x i32> undef) +; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i32, ptr [[B]], i32 [[INDEX]] +; CHECK-NEXT: [[WIDE_MASKED_LOAD2:%.*]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[TMP3]], i32 4, <4 x i1> [[TMP1]], <4 x i32> undef) +; CHECK-NEXT: [[MUL:%.*]] = mul nsw <4 x i32> [[WIDE_MASKED_LOAD2]], [[WIDE_MASKED_LOAD]] +; CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[C]], i32 [[INDEX]] +; CHECK-NEXT: tail call void @llvm.masked.store.v4i32.p0(<4 x i32> [[MUL]], ptr [[TMP6]], i32 4, <4 x i1> [[TMP1]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 4 +; CHECK-NEXT: [[TMP15]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP14]], i32 1) +; CHECK-NEXT: [[TMP16:%.*]] = icmp ne i32 [[TMP15]], 0 +; CHECK-NEXT: br i1 [[TMP16]], label %[[VECTOR_BODY]], label %[[FOR_COND_CLEANUP]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp eq i32 %N, 0 %tmp8 = add i32 %N, 3 @@ -134,17 +208,47 @@ for.cond.cleanup: ; preds = %vector.body, %entry ret void } -; CHECK-LABEL: split_vector -; CHECK-NOT: %num.elements = add i32 %trip.count.minus.1, 1 -; CHECK: vector.body: -; CHECK: %index = phi i32 -; CHECK: [[ELEMS:%[^ ]+]] = phi i32 [ %N, %vector.ph ], [ [[REMAINING:%[^ ]+]], %vector.body ] -; CHECK: [[VCTP:%[^ ]+]] = call <4 x i1> @llvm.arm.mve.vctp32(i32 [[ELEMS]]) -; CHECK: [[REMAINING]] = sub i32 [[ELEMS]], 4 -; CHECK: [[LD0:%[^ ]+]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr {{.*}}, i32 4, <4 x i1> [[VCTP]], <4 x i32> undef) -; CHECK: [[LD1:%[^ ]+]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr {{.*}}, i32 4, <4 x i1> [[VCTP]], <4 x i32> undef) -; CHECK: tail call void @llvm.masked.store.v4i32.p0(<4 x i32> {{.*}}, ptr {{.*}}, i32 4, <4 x i1> [[VCTP]]) define dso_local arm_aapcs_vfpcc void @split_vector(ptr noalias nocapture readonly %a, ptr noalias nocapture readonly %b, ptr noalias nocapture %c, i32 %N) { +; CHECK-LABEL: define dso_local arm_aapcs_vfpcc void @split_vector( +; CHECK-SAME: ptr noalias readonly captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], ptr noalias captures(none) [[C:%.*]], i32 [[N:%.*]]) #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp eq i32 [[N]], 0 +; CHECK-NEXT: [[TMP8:%.*]] = add i32 [[N]], 3 +; CHECK-NEXT: [[TMP9:%.*]] = lshr i32 [[TMP8]], 2 +; CHECK-NEXT: [[TMP10:%.*]] = shl nuw i32 [[TMP9]], 2 +; CHECK-NEXT: [[TMP11:%.*]] = add i32 [[TMP10]], -4 +; CHECK-NEXT: [[TMP12:%.*]] = lshr i32 [[TMP11]], 2 +; CHECK-NEXT: [[TMP13:%.*]] = add nuw nsw i32 [[TMP12]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[FOR_COND_CLEANUP:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP13]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP14:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP15:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP0:%.*]] = phi i32 [ [[N]], %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[INDEX]] +; CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.vctp32(i32 [[TMP0]]) +; CHECK-NEXT: [[TMP2]] = sub i32 [[TMP0]], 4 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[TMP]], i32 4, <4 x i1> [[TMP1]], <4 x i32> undef) +; CHECK-NEXT: [[EXTRACT_1_LOW:%.*]] = shufflevector <4 x i32> [[WIDE_MASKED_LOAD]], <4 x i32> undef, <2 x i32> <i32 0, i32 2> +; CHECK-NEXT: [[EXTRACT_1_HIGH:%.*]] = shufflevector <4 x i32> [[WIDE_MASKED_LOAD]], <4 x i32> undef, <2 x i32> <i32 1, i32 3> +; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i32, ptr [[B]], i32 [[INDEX]] +; CHECK-NEXT: [[WIDE_MASKED_LOAD2:%.*]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[TMP3]], i32 4, <4 x i1> [[TMP1]], <4 x i32> undef) +; CHECK-NEXT: [[EXTRACT_2_LOW:%.*]] = shufflevector <4 x i32> [[WIDE_MASKED_LOAD2]], <4 x i32> undef, <2 x i32> <i32 0, i32 2> +; CHECK-NEXT: [[EXTRACT_2_HIGH:%.*]] = shufflevector <4 x i32> [[WIDE_MASKED_LOAD2]], <4 x i32> undef, <2 x i32> <i32 1, i32 3> +; CHECK-NEXT: [[MUL:%.*]] = mul nsw <2 x i32> [[EXTRACT_1_LOW]], [[EXTRACT_2_LOW]] +; CHECK-NEXT: [[SUB:%.*]] = sub nsw <2 x i32> [[EXTRACT_1_HIGH]], [[EXTRACT_2_HIGH]] +; CHECK-NEXT: [[COMBINE:%.*]] = shufflevector <2 x i32> [[MUL]], <2 x i32> [[SUB]], <4 x i32> <i32 0, i32 1, i32 2, i32 3> +; CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[C]], i32 [[INDEX]] +; CHECK-NEXT: tail call void @llvm.masked.store.v4i32.p0(<4 x i32> [[COMBINE]], ptr [[TMP6]], i32 4, <4 x i1> [[TMP1]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 4 +; CHECK-NEXT: [[TMP15]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP14]], i32 1) +; CHECK-NEXT: [[TMP16:%.*]] = icmp ne i32 [[TMP15]], 0 +; CHECK-NEXT: br i1 [[TMP16]], label %[[VECTOR_BODY]], label %[[FOR_COND_CLEANUP]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp eq i32 %N, 0 %tmp8 = add i32 %N, 3 @@ -186,14 +290,48 @@ for.cond.cleanup: ; preds = %vector.body, %entry } ; One of the loads now uses ult predicate. -; CHECK-LABEL: mismatch_load_pred -; CHECK: [[ELEMS:%[^ ]+]] = phi i32 [ %N, %vector.ph ], [ [[REMAINING:%[^ ]+]], %vector.body ] -; CHECK: [[VCTP:%[^ ]+]] = call <4 x i1> @llvm.arm.mve.vctp32(i32 [[ELEMS]]) -; CHECK: [[REMAINING]] = sub i32 [[ELEMS]], 4 -; CHECK: [[LD0:%[^ ]+]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr {{.*}}, i32 4, <4 x i1> [[VCTP]], <4 x i32> undef) -; CHECK: [[LD1:%[^ ]+]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr {{.*}}, i32 4, <4 x i1> %wrong, <4 x i32> undef) -; CHECK: tail call void @llvm.masked.store.v4i32.p0(<4 x i32> {{.*}}, ptr {{.*}}, i32 4, <4 x i1> [[VCTP]]) define dso_local arm_aapcs_vfpcc void @mismatch_load_pred(ptr noalias nocapture readonly %a, ptr noalias nocapture readonly %b, ptr noalias nocapture %c, i32 %N) { +; CHECK-LABEL: define dso_local arm_aapcs_vfpcc void @mismatch_load_pred( +; CHECK-SAME: ptr noalias readonly captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], ptr noalias captures(none) [[C:%.*]], i32 [[N:%.*]]) #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp eq i32 [[N]], 0 +; CHECK-NEXT: [[TMP8:%.*]] = add i32 [[N]], 3 +; CHECK-NEXT: [[TMP9:%.*]] = lshr i32 [[TMP8]], 2 +; CHECK-NEXT: [[TMP10:%.*]] = shl nuw i32 [[TMP9]], 2 +; CHECK-NEXT: [[TMP11:%.*]] = add i32 [[TMP10]], -4 +; CHECK-NEXT: [[TMP12:%.*]] = lshr i32 [[TMP11]], 2 +; CHECK-NEXT: [[TMP13:%.*]] = add nuw nsw i32 [[TMP12]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[FOR_COND_CLEANUP:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[TRIP_COUNT_MINUS_1:%.*]] = add i32 [[N]], -1 +; CHECK-NEXT: [[BROADCAST_SPLATINSERT10:%.*]] = insertelement <4 x i32> undef, i32 [[TRIP_COUNT_MINUS_1]], i32 0 +; CHECK-NEXT: [[BROADCAST_SPLAT11:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT10]], <4 x i32> undef, <4 x i32> zeroinitializer +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP13]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP14:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP15:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP0:%.*]] = phi i32 [ [[N]], %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> undef, i32 [[INDEX]], i32 0 +; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> undef, <4 x i32> zeroinitializer +; CHECK-NEXT: [[INDUCTION:%.*]] = add <4 x i32> [[BROADCAST_SPLAT]], <i32 0, i32 1, i32 2, i32 3> +; CHECK-NEXT: [[TMP:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[INDEX]] +; CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.vctp32(i32 [[TMP0]]) +; CHECK-NEXT: [[TMP2]] = sub i32 [[TMP0]], 4 +; CHECK-NEXT: [[WRONG:%.*]] = icmp ult <4 x i32> [[INDUCTION]], [[BROADCAST_SPLAT11]] +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[TMP]], i32 4, <4 x i1> [[TMP1]], <4 x i32> undef) +; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i32, ptr [[B]], i32 [[INDEX]] +; CHECK-NEXT: [[WIDE_MASKED_LOAD12:%.*]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[TMP3]], i32 4, <4 x i1> [[WRONG]], <4 x i32> undef) +; CHECK-NEXT: [[TMP5:%.*]] = mul nsw <4 x i32> [[WIDE_MASKED_LOAD12]], [[WIDE_MASKED_LOAD]] +; CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[C]], i32 [[INDEX]] +; CHECK-NEXT: tail call void @llvm.masked.store.v4i32.p0(<4 x i32> [[TMP5]], ptr [[TMP6]], i32 4, <4 x i1> [[TMP1]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 4 +; CHECK-NEXT: [[TMP15]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP14]], i32 1) +; CHECK-NEXT: [[TMP16:%.*]] = icmp ne i32 [[TMP15]], 0 +; CHECK-NEXT: br i1 [[TMP16]], label %[[VECTOR_BODY]], label %[[FOR_COND_CLEANUP]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp eq i32 %N, 0 %tmp8 = add i32 %N, 3 @@ -236,17 +374,48 @@ for.cond.cleanup: ; preds = %vector.body, %entry } ; The store now uses ult predicate. -; CHECK-LABEL: mismatch_store_pred -; CHECK-NOT: %num.elements = add i32 %trip.count.minus.1, 1 -; CHECK: vector.body: -; CHECK: %index = phi i32 -; CHECK: [[ELEMS:%[^ ]+]] = phi i32 [ %N, %vector.ph ], [ [[REMAINING:%[^ ]+]], %vector.body ] -; CHECK: [[VCTP:%[^ ]+]] = call <4 x i1> @llvm.arm.mve.vctp32(i32 [[ELEMS]]) -; CHECK: [[REMAINING]] = sub i32 [[ELEMS]], 4 -; CHECK: [[LD0:%[^ ]+]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr {{.*}}, i32 4, <4 x i1> [[VCTP]], <4 x i32> undef) -; CHECK: [[LD1:%[^ ]+]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr {{.*}}, i32 4, <4 x i1> [[VCTP]], <4 x i32> undef) -; CHECK: tail call void @llvm.masked.store.v4i32.p0(<4 x i32> {{.*}}, ptr {{.*}}, i32 4, <4 x i1> %wrong) define dso_local arm_aapcs_vfpcc void @mismatch_store_pred(ptr noalias nocapture readonly %a, ptr noalias nocapture readonly %b, ptr noalias nocapture %c, i32 %N) { +; CHECK-LABEL: define dso_local arm_aapcs_vfpcc void @mismatch_store_pred( +; CHECK-SAME: ptr noalias readonly captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], ptr noalias captures(none) [[C:%.*]], i32 [[N:%.*]]) #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp eq i32 [[N]], 0 +; CHECK-NEXT: [[TMP8:%.*]] = add i32 [[N]], 3 +; CHECK-NEXT: [[TMP9:%.*]] = lshr i32 [[TMP8]], 2 +; CHECK-NEXT: [[TMP10:%.*]] = shl nuw i32 [[TMP9]], 2 +; CHECK-NEXT: [[TMP11:%.*]] = add i32 [[TMP10]], -4 +; CHECK-NEXT: [[TMP12:%.*]] = lshr i32 [[TMP11]], 2 +; CHECK-NEXT: [[TMP13:%.*]] = add nuw nsw i32 [[TMP12]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[FOR_COND_CLEANUP:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[TRIP_COUNT_MINUS_1:%.*]] = add i32 [[N]], -1 +; CHECK-NEXT: [[BROADCAST_SPLATINSERT10:%.*]] = insertelement <4 x i32> undef, i32 [[TRIP_COUNT_MINUS_1]], i32 0 +; CHECK-NEXT: [[BROADCAST_SPLAT11:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT10]], <4 x i32> undef, <4 x i32> zeroinitializer +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP13]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP14:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP15:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP0:%.*]] = phi i32 [ [[N]], %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> undef, i32 [[INDEX]], i32 0 +; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> undef, <4 x i32> zeroinitializer +; CHECK-NEXT: [[INDUCTION:%.*]] = add <4 x i32> [[BROADCAST_SPLAT]], <i32 0, i32 1, i32 2, i32 3> +; CHECK-NEXT: [[TMP:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[INDEX]] +; CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.vctp32(i32 [[TMP0]]) +; CHECK-NEXT: [[TMP2]] = sub i32 [[TMP0]], 4 +; CHECK-NEXT: [[WRONG:%.*]] = icmp ult <4 x i32> [[INDUCTION]], [[BROADCAST_SPLAT11]] +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[TMP]], i32 4, <4 x i1> [[TMP1]], <4 x i32> undef) +; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i32, ptr [[B]], i32 [[INDEX]] +; CHECK-NEXT: [[WIDE_MASKED_LOAD12:%.*]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[TMP3]], i32 4, <4 x i1> [[TMP1]], <4 x i32> undef) +; CHECK-NEXT: [[TMP5:%.*]] = mul nsw <4 x i32> [[WIDE_MASKED_LOAD12]], [[WIDE_MASKED_LOAD]] +; CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[C]], i32 [[INDEX]] +; CHECK-NEXT: tail call void @llvm.masked.store.v4i32.p0(<4 x i32> [[TMP5]], ptr [[TMP6]], i32 4, <4 x i1> [[WRONG]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 4 +; CHECK-NEXT: [[TMP15]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP14]], i32 1) +; CHECK-NEXT: [[TMP16:%.*]] = icmp ne i32 [[TMP15]], 0 +; CHECK-NEXT: br i1 [[TMP16]], label %[[VECTOR_BODY]], label %[[FOR_COND_CLEANUP]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp eq i32 %N, 0 %tmp8 = add i32 %N, 3 @@ -294,14 +463,72 @@ for.cond.cleanup: ; preds = %vector.body, %entry ; ; Step value 16 doesn't match vector width 4 ; -; CHECK-LABEL: interleave4 -; CHECK: vector.body: -; CHECK: %active.lane.mask = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 %index, i32 %N) -; CHECK: %active.lane.mask{{.*}} = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 %v7, i32 %N) -; CHECK: %active.lane.mask{{.*}} = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 %v8, i32 %N) -; CHECK: %active.lane.mask{{.*}} = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 %v9, i32 %N) -; define dso_local void @interleave4(ptr noalias nocapture %A, ptr noalias nocapture readonly %B, ptr noalias nocapture readonly %C, i32 %N) local_unnamed_addr #0 { +; CHECK-LABEL: define dso_local void @interleave4( +; CHECK-SAME: ptr noalias captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], ptr noalias readonly captures(none) [[C:%.*]], i32 [[N:%.*]]) local_unnamed_addr #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp sgt i32 [[N]], 0 +; CHECK-NEXT: [[V0:%.*]] = add i32 [[N]], 15 +; CHECK-NEXT: [[V1:%.*]] = lshr i32 [[V0]], 4 +; CHECK-NEXT: [[V2:%.*]] = shl nuw i32 [[V1]], 4 +; CHECK-NEXT: [[V3:%.*]] = add i32 [[V2]], -16 +; CHECK-NEXT: [[V4:%.*]] = lshr i32 [[V3]], 4 +; CHECK-NEXT: [[V5:%.*]] = add nuw nsw i32 [[V4]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[VECTOR_PH:.*]], label %[[FOR_COND_CLEANUP:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[SCEVGEP:%.*]] = getelementptr i32, ptr [[A]], i32 8 +; CHECK-NEXT: [[SCEVGEP30:%.*]] = getelementptr i32, ptr [[C]], i32 8 +; CHECK-NEXT: [[SCEVGEP37:%.*]] = getelementptr i32, ptr [[B]], i32 8 +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[V5]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[LSR_IV38:%.*]] = phi ptr [ [[SCEVGEP39:%.*]], %[[VECTOR_BODY]] ], [ [[SCEVGEP37]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[LSR_IV31:%.*]] = phi ptr [ [[SCEVGEP32:%.*]], %[[VECTOR_BODY]] ], [ [[SCEVGEP30]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[LSR_IV:%.*]] = phi ptr [ [[SCEVGEP25:%.*]], %[[VECTOR_BODY]] ], [ [[SCEVGEP]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[V14:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[V6:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[V15:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 [[INDEX]], i32 [[N]]) +; CHECK-NEXT: [[V7:%.*]] = add i32 [[INDEX]], 4 +; CHECK-NEXT: [[ACTIVE_LANE_MASK15:%.*]] = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 [[V7]], i32 [[N]]) +; CHECK-NEXT: [[V8:%.*]] = add i32 [[V7]], 4 +; CHECK-NEXT: [[ACTIVE_LANE_MASK16:%.*]] = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 [[V8]], i32 [[N]]) +; CHECK-NEXT: [[V9:%.*]] = add i32 [[V8]], 4 +; CHECK-NEXT: [[ACTIVE_LANE_MASK17:%.*]] = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 [[V9]], i32 [[N]]) +; CHECK-NEXT: [[SCEVGEP42:%.*]] = getelementptr <4 x i32>, ptr [[LSR_IV38]], i32 -2 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[SCEVGEP42]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK]], <4 x i32> undef) +; CHECK-NEXT: [[SCEVGEP43:%.*]] = getelementptr <4 x i32>, ptr [[LSR_IV38]], i32 -1 +; CHECK-NEXT: [[WIDE_MASKED_LOAD18:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr nonnull [[SCEVGEP43]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK15]], <4 x i32> undef) +; CHECK-NEXT: [[WIDE_MASKED_LOAD19:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr nonnull [[LSR_IV38]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK16]], <4 x i32> undef) +; CHECK-NEXT: [[SCEVGEP41:%.*]] = getelementptr <4 x i32>, ptr [[LSR_IV38]], i32 1 +; CHECK-NEXT: [[WIDE_MASKED_LOAD20:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr nonnull [[SCEVGEP41]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK17]], <4 x i32> undef) +; CHECK-NEXT: [[SCEVGEP34:%.*]] = getelementptr <4 x i32>, ptr [[LSR_IV31]], i32 -2 +; CHECK-NEXT: [[WIDE_MASKED_LOAD21:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[SCEVGEP34]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK]], <4 x i32> undef) +; CHECK-NEXT: [[SCEVGEP35:%.*]] = getelementptr <4 x i32>, ptr [[LSR_IV31]], i32 -1 +; CHECK-NEXT: [[WIDE_MASKED_LOAD22:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr nonnull [[SCEVGEP35]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK15]], <4 x i32> undef) +; CHECK-NEXT: [[WIDE_MASKED_LOAD23:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr nonnull [[LSR_IV31]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK16]], <4 x i32> undef) +; CHECK-NEXT: [[SCEVGEP36:%.*]] = getelementptr <4 x i32>, ptr [[LSR_IV31]], i32 1 +; CHECK-NEXT: [[WIDE_MASKED_LOAD24:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr nonnull [[SCEVGEP36]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK17]], <4 x i32> undef) +; CHECK-NEXT: [[V10:%.*]] = add nsw <4 x i32> [[WIDE_MASKED_LOAD21]], [[WIDE_MASKED_LOAD]] +; CHECK-NEXT: [[V11:%.*]] = add nsw <4 x i32> [[WIDE_MASKED_LOAD22]], [[WIDE_MASKED_LOAD18]] +; CHECK-NEXT: [[V12:%.*]] = add nsw <4 x i32> [[WIDE_MASKED_LOAD23]], [[WIDE_MASKED_LOAD19]] +; CHECK-NEXT: [[V13:%.*]] = add nsw <4 x i32> [[WIDE_MASKED_LOAD24]], [[WIDE_MASKED_LOAD20]] +; CHECK-NEXT: [[SCEVGEP27:%.*]] = getelementptr <4 x i32>, ptr [[LSR_IV]], i32 -2 +; CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[V10]], ptr [[SCEVGEP27]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK]]) +; CHECK-NEXT: [[SCEVGEP28:%.*]] = getelementptr <4 x i32>, ptr [[LSR_IV]], i32 -1 +; CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[V11]], ptr [[SCEVGEP28]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK15]]) +; CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[V12]], ptr [[LSR_IV]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK16]]) +; CHECK-NEXT: [[SCEVGEP29:%.*]] = getelementptr <4 x i32>, ptr [[LSR_IV]], i32 1 +; CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[V13]], ptr [[SCEVGEP29]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK17]]) +; CHECK-NEXT: [[SCEVGEP25]] = getelementptr i32, ptr [[LSR_IV]], i32 16 +; CHECK-NEXT: [[SCEVGEP32]] = getelementptr i32, ptr [[LSR_IV31]], i32 16 +; CHECK-NEXT: [[SCEVGEP39]] = getelementptr i32, ptr [[LSR_IV38]], i32 16 +; CHECK-NEXT: [[V14]] = add i32 [[V9]], 4 +; CHECK-NEXT: [[V15]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[V6]], i32 1) +; CHECK-NEXT: [[V16:%.*]] = icmp ne i32 [[V15]], 0 +; CHECK-NEXT: br i1 [[V16]], label %[[VECTOR_BODY]], label %[[FOR_COND_CLEANUP]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp sgt i32 %N, 0 %v0 = add i32 %N, 15 @@ -370,12 +597,42 @@ for.cond.cleanup: ret void } -; CHECK-LABEL: const_expected_in_set_loop -; CHECK: call <4 x i1> @llvm.get.active.lane.mask -; CHECK-NOT: vctp -; CHECK: ret void -; define dso_local void @const_expected_in_set_loop(ptr noalias nocapture %A, ptr noalias nocapture readonly %B, ptr noalias nocapture readonly %C, i32 %N) local_unnamed_addr #0 { +; CHECK-LABEL: define dso_local void @const_expected_in_set_loop( +; CHECK-SAME: ptr noalias captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], ptr noalias readonly captures(none) [[C:%.*]], i32 [[N:%.*]]) local_unnamed_addr #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp sgt i32 [[N]], 0 +; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N]], 3 +; CHECK-NEXT: [[TMP1:%.*]] = lshr i32 [[TMP0]], 2 +; CHECK-NEXT: [[TMP2:%.*]] = shl nuw i32 [[TMP1]], 2 +; CHECK-NEXT: [[TMP3:%.*]] = add i32 [[TMP2]], -4 +; CHECK-NEXT: [[TMP4:%.*]] = lshr i32 [[TMP3]], 2 +; CHECK-NEXT: [[TMP5:%.*]] = add nuw nsw i32 [[TMP4]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[VECTOR_PH:.*]], label %[[FOR_COND_CLEANUP:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP5]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[LSR_IV17:%.*]] = phi ptr [ [[SCEVGEP18:%.*]], %[[VECTOR_BODY]] ], [ [[A]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[LSR_IV14:%.*]] = phi ptr [ [[SCEVGEP15:%.*]], %[[VECTOR_BODY]] ], [ [[C]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[LSR_IV:%.*]] = phi ptr [ [[SCEVGEP:%.*]], %[[VECTOR_BODY]] ], [ [[B]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP6:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP8:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 [[INDEX]], i32 42) +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[LSR_IV]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK]], <4 x i32> undef) +; CHECK-NEXT: [[WIDE_MASKED_LOAD12:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[LSR_IV14]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK]], <4 x i32> undef) +; CHECK-NEXT: [[TMP7:%.*]] = add nsw <4 x i32> [[WIDE_MASKED_LOAD12]], [[WIDE_MASKED_LOAD]] +; CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[TMP7]], ptr [[LSR_IV17]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 4 +; CHECK-NEXT: [[SCEVGEP]] = getelementptr i32, ptr [[LSR_IV]], i32 4 +; CHECK-NEXT: [[SCEVGEP15]] = getelementptr i32, ptr [[LSR_IV14]], i32 4 +; CHECK-NEXT: [[SCEVGEP18]] = getelementptr i32, ptr [[LSR_IV17]], i32 4 +; CHECK-NEXT: [[TMP8]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP6]], i32 1) +; CHECK-NEXT: [[TMP9:%.*]] = icmp ne i32 [[TMP8]], 0 +; CHECK-NEXT: br i1 [[TMP9]], label %[[VECTOR_BODY]], label %[[FOR_COND_CLEANUP]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp sgt i32 %N, 0 %0 = add i32 %N, 3 @@ -413,12 +670,42 @@ for.cond.cleanup: ; preds = %vector.body, %entry ret void } -; CHECK-LABEL: tripcount_arg_not_invariant -; CHECK: call <4 x i1> @llvm.get.active.lane.mask -; CHECK-NOT: vctp -; CHECK: ret void -; define dso_local void @tripcount_arg_not_invariant(ptr noalias nocapture %A, ptr noalias nocapture readonly %B, ptr noalias nocapture readonly %C, i32 %N) local_unnamed_addr #0 { +; CHECK-LABEL: define dso_local void @tripcount_arg_not_invariant( +; CHECK-SAME: ptr noalias captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], ptr noalias readonly captures(none) [[C:%.*]], i32 [[N:%.*]]) local_unnamed_addr #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp sgt i32 [[N]], 0 +; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N]], 3 +; CHECK-NEXT: [[TMP1:%.*]] = lshr i32 [[TMP0]], 2 +; CHECK-NEXT: [[TMP2:%.*]] = shl nuw i32 [[TMP1]], 2 +; CHECK-NEXT: [[TMP3:%.*]] = add i32 [[TMP2]], -4 +; CHECK-NEXT: [[TMP4:%.*]] = lshr i32 [[TMP3]], 2 +; CHECK-NEXT: [[TMP5:%.*]] = add nuw nsw i32 [[TMP4]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[VECTOR_PH:.*]], label %[[FOR_COND_CLEANUP:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP5]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[LSR_IV17:%.*]] = phi ptr [ [[SCEVGEP18:%.*]], %[[VECTOR_BODY]] ], [ [[A]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[LSR_IV14:%.*]] = phi ptr [ [[SCEVGEP15:%.*]], %[[VECTOR_BODY]] ], [ [[C]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[LSR_IV:%.*]] = phi ptr [ [[SCEVGEP:%.*]], %[[VECTOR_BODY]] ], [ [[B]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP6:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP8:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 [[INDEX]], i32 [[INDEX]]) +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[LSR_IV]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK]], <4 x i32> undef) +; CHECK-NEXT: [[WIDE_MASKED_LOAD12:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[LSR_IV14]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK]], <4 x i32> undef) +; CHECK-NEXT: [[TMP7:%.*]] = add nsw <4 x i32> [[WIDE_MASKED_LOAD12]], [[WIDE_MASKED_LOAD]] +; CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[TMP7]], ptr [[LSR_IV17]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 4 +; CHECK-NEXT: [[SCEVGEP]] = getelementptr i32, ptr [[LSR_IV]], i32 4 +; CHECK-NEXT: [[SCEVGEP15]] = getelementptr i32, ptr [[LSR_IV14]], i32 4 +; CHECK-NEXT: [[SCEVGEP18]] = getelementptr i32, ptr [[LSR_IV17]], i32 4 +; CHECK-NEXT: [[TMP8]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP6]], i32 1) +; CHECK-NEXT: [[TMP9:%.*]] = icmp ne i32 [[TMP8]], 0 +; CHECK-NEXT: br i1 [[TMP9]], label %[[VECTOR_BODY]], label %[[VECTOR_PH]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp sgt i32 %N, 0 %0 = add i32 %N, 3 @@ -458,12 +745,42 @@ for.cond.cleanup: ; preds = %vector.body, %entry ret void } -; CHECK-LABEL: addrec_base_not_zero -; CHECK: call <4 x i1> @llvm.get.active.lane.mask -; CHECK-NOT: vctp -; CHECK: ret void -; define dso_local void @addrec_base_not_zero(ptr noalias nocapture %A, ptr noalias nocapture readonly %B, ptr noalias nocapture readonly %C, i32 %N) local_unnamed_addr #0 { +; CHECK-LABEL: define dso_local void @addrec_base_not_zero( +; CHECK-SAME: ptr noalias captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], ptr noalias readonly captures(none) [[C:%.*]], i32 [[N:%.*]]) local_unnamed_addr #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp sgt i32 [[N]], 0 +; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N]], 3 +; CHECK-NEXT: [[TMP1:%.*]] = lshr i32 [[TMP0]], 2 +; CHECK-NEXT: [[TMP2:%.*]] = shl nuw i32 [[TMP1]], 2 +; CHECK-NEXT: [[TMP3:%.*]] = add i32 [[TMP2]], -4 +; CHECK-NEXT: [[TMP4:%.*]] = lshr i32 [[TMP3]], 2 +; CHECK-NEXT: [[TMP5:%.*]] = add nuw nsw i32 [[TMP4]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[VECTOR_PH:.*]], label %[[FOR_COND_CLEANUP:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP5]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[LSR_IV17:%.*]] = phi ptr [ [[SCEVGEP18:%.*]], %[[VECTOR_BODY]] ], [ [[A]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[LSR_IV14:%.*]] = phi ptr [ [[SCEVGEP15:%.*]], %[[VECTOR_BODY]] ], [ [[C]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[LSR_IV:%.*]] = phi ptr [ [[SCEVGEP:%.*]], %[[VECTOR_BODY]] ], [ [[B]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 1, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP6:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP8:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[ACTIVE_LANE_MASK:%.*]] = call <4 x i1> @llvm.get.active.lane.mask.v4i1.i32(i32 [[INDEX]], i32 [[N]]) +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[LSR_IV]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK]], <4 x i32> undef) +; CHECK-NEXT: [[WIDE_MASKED_LOAD12:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[LSR_IV14]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK]], <4 x i32> undef) +; CHECK-NEXT: [[TMP7:%.*]] = add nsw <4 x i32> [[WIDE_MASKED_LOAD12]], [[WIDE_MASKED_LOAD]] +; CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[TMP7]], ptr [[LSR_IV17]], i32 4, <4 x i1> [[ACTIVE_LANE_MASK]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 4 +; CHECK-NEXT: [[SCEVGEP]] = getelementptr i32, ptr [[LSR_IV]], i32 4 +; CHECK-NEXT: [[SCEVGEP15]] = getelementptr i32, ptr [[LSR_IV14]], i32 4 +; CHECK-NEXT: [[SCEVGEP18]] = getelementptr i32, ptr [[LSR_IV17]], i32 4 +; CHECK-NEXT: [[TMP8]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP6]], i32 1) +; CHECK-NEXT: [[TMP9:%.*]] = icmp ne i32 [[TMP8]], 0 +; CHECK-NEXT: br i1 [[TMP9]], label %[[VECTOR_BODY]], label %[[VECTOR_PH]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp sgt i32 %N, 0 %0 = add i32 %N, 3 diff --git a/llvm/test/CodeGen/Thumb2/LowOverheadLoops/tail-pred-reduce.ll b/llvm/test/CodeGen/Thumb2/LowOverheadLoops/tail-pred-reduce.ll index fa6a66b..9775cf9 100644 --- a/llvm/test/CodeGen/Thumb2/LowOverheadLoops/tail-pred-reduce.ll +++ b/llvm/test/CodeGen/Thumb2/LowOverheadLoops/tail-pred-reduce.ll @@ -1,15 +1,55 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --prefix-filecheck-ir-name INST --version 6 ; RUN: opt -mtriple=thumbv8.1m.main -mve-tail-predication -tail-predication=enabled -mattr=+mve %s -S -o - | FileCheck %s -; CHECK-LABEL: reduction_i32 -; CHECK: phi i32 [ 0, %vector.ph ] -; CHECK: phi <8 x i16> [ zeroinitializer, %vector.ph ] -; CHECK: phi i32 -; CHECK: [[PHI:%[^ ]+]] = phi i32 [ %N, %vector.ph ], [ [[ELEMS:%[^ ]+]], %vector.body ] -; CHECK: [[VCTP:%[^ ]+]] = call <8 x i1> @llvm.arm.mve.vctp16(i32 [[PHI]]) -; CHECK: [[ELEMS]] = sub i32 [[PHI]], 8 -; CHECK: call <8 x i16> @llvm.masked.load.v8i16.p0(ptr %tmp2, i32 4, <8 x i1> [[VCTP]], <8 x i16> undef) -; CHECK: call <8 x i16> @llvm.masked.load.v8i16.p0(ptr %tmp5, i32 4, <8 x i1> [[VCTP]], <8 x i16> undef) define i16 @reduction_i32(ptr nocapture readonly %A, ptr nocapture readonly %B, i32 %N) { +; CHECK-LABEL: define i16 @reduction_i32( +; CHECK-SAME: ptr readonly captures(none) [[A:%.*]], ptr readonly captures(none) [[B:%.*]], i32 [[N:%.*]]) #[[ATTR0:[0-9]+]] { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: [[CMP8:%.*]] = icmp eq i32 [[N]], 0 +; CHECK-NEXT: br i1 [[CMP8]], label %[[FOR_COND_CLEANUP:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[TMP:%.*]] = add i32 [[N]], -1 +; CHECK-NEXT: [[N_RND_UP:%.*]] = add i32 [[TMP]], 8 +; CHECK-NEXT: [[N_VEC:%.*]] = and i32 [[N_RND_UP]], -8 +; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N_VEC]], -8 +; CHECK-NEXT: [[TMP1:%.*]] = lshr i32 [[TMP0]], 3 +; CHECK-NEXT: [[TMP2:%.*]] = add i32 [[TMP1]], 1 +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP2]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <8 x i16> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP8:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP3:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP9:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP4:%.*]] = phi i32 [ [[N]], %[[VECTOR_PH]] ], [ [[TMP6:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[INSTTMP2:%.*]] = getelementptr inbounds i16, ptr [[A]], i32 [[INDEX]] +; CHECK-NEXT: [[TMP5:%.*]] = call <8 x i1> @llvm.arm.mve.vctp16(i32 [[TMP4]]) +; CHECK-NEXT: [[TMP6]] = sub i32 [[TMP4]], 8 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[INSTTMP2]], i32 4, <8 x i1> [[TMP5]], <8 x i16> undef) +; CHECK-NEXT: [[INSTTMP5:%.*]] = getelementptr inbounds i16, ptr [[B]], i32 [[INDEX]] +; CHECK-NEXT: [[WIDE_MASKED_LOAD3:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[INSTTMP5]], i32 4, <8 x i1> [[TMP5]], <8 x i16> undef) +; CHECK-NEXT: [[TMP7:%.*]] = add <8 x i16> [[WIDE_MASKED_LOAD]], [[VEC_PHI]] +; CHECK-NEXT: [[TMP8]] = add <8 x i16> [[TMP7]], [[WIDE_MASKED_LOAD3]] +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 8 +; CHECK-NEXT: [[TMP9]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP3]], i32 1) +; CHECK-NEXT: [[TMP12:%.*]] = icmp ne i32 [[TMP9]], 0 +; CHECK-NEXT: br i1 [[TMP12]], label %[[VECTOR_BODY]], label %[[MIDDLE_BLOCK:.*]] +; CHECK: [[MIDDLE_BLOCK]]: +; CHECK-NEXT: [[VEC_PHI_LCSSA:%.*]] = phi <8 x i16> [ [[VEC_PHI]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[DOTLCSSA3:%.*]] = phi <8 x i1> [ [[TMP5]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[DOTLCSSA:%.*]] = phi <8 x i16> [ [[TMP8]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP10:%.*]] = select <8 x i1> [[DOTLCSSA3]], <8 x i16> [[DOTLCSSA]], <8 x i16> [[VEC_PHI_LCSSA]] +; CHECK-NEXT: [[RDX_SHUF:%.*]] = shufflevector <8 x i16> [[TMP10]], <8 x i16> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison> +; CHECK-NEXT: [[BIN_RDX:%.*]] = add <8 x i16> [[RDX_SHUF]], [[TMP10]] +; CHECK-NEXT: [[RDX_SHUF4:%.*]] = shufflevector <8 x i16> [[BIN_RDX]], <8 x i16> undef, <8 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison> +; CHECK-NEXT: [[BIN_RDX5:%.*]] = add <8 x i16> [[RDX_SHUF4]], [[BIN_RDX]] +; CHECK-NEXT: [[RDX_SHUF6:%.*]] = shufflevector <8 x i16> [[BIN_RDX5]], <8 x i16> undef, <8 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison> +; CHECK-NEXT: [[BIN_RDX7:%.*]] = add <8 x i16> [[RDX_SHUF6]], [[BIN_RDX5]] +; CHECK-NEXT: [[TMP11:%.*]] = extractelement <8 x i16> [[BIN_RDX7]], i32 0 +; CHECK-NEXT: ret i16 [[TMP11]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: [[RES_0:%.*]] = phi i16 [ 0, %[[ENTRY]] ] +; CHECK-NEXT: ret i16 [[RES_0]] +; entry: %cmp8 = icmp eq i32 %N, 0 br i1 %cmp8, label %for.cond.cleanup, label %vector.ph @@ -59,16 +99,52 @@ for.cond.cleanup: ret i16 %res.0 } -; CHECK-LABEL: reduction_i32_with_scalar -; CHECK: vector.body: -; CHECK: %index = phi i32 [ 0, %vector.ph ], [ %index.next, %vector.body ] -; CHECK: %vec.phi = phi <8 x i16> [ zeroinitializer, %vector.ph ], [ %{{.*}}, %vector.body ] -; CHECK: %{{.*}} = phi i32 [ %{{.*}}, %vector.ph ], [ %{{.*}}, %vector.body ] -; CHECK: [[PHI:%[^ ]+]] = phi i32 [ %N, %vector.ph ], [ [[ELEMS:%[^ ]+]], %vector.body ] -; CHECK: [[VCTP:%[^ ]+]] = call <8 x i1> @llvm.arm.mve.vctp16(i32 [[PHI]]) -; CHECK: [[ELEMS]] = sub i32 [[PHI]], 8 -; CHECK: call <8 x i16> @llvm.masked.load.v8i16.p0(ptr %tmp2, i32 4, <8 x i1> [[VCTP]], <8 x i16> undef) define i16 @reduction_i32_with_scalar(ptr nocapture readonly %A, i16 %B, i32 %N) local_unnamed_addr { +; CHECK-LABEL: define i16 @reduction_i32_with_scalar( +; CHECK-SAME: ptr readonly captures(none) [[A:%.*]], i16 [[B:%.*]], i32 [[N:%.*]]) local_unnamed_addr #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: [[CMP8:%.*]] = icmp eq i32 [[N]], 0 +; CHECK-NEXT: br i1 [[CMP8]], label %[[FOR_COND_CLEANUP:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[TMP:%.*]] = add i32 [[N]], -1 +; CHECK-NEXT: [[N_RND_UP:%.*]] = add nuw nsw i32 [[TMP]], 8 +; CHECK-NEXT: [[N_VEC:%.*]] = and i32 [[N_RND_UP]], -8 +; CHECK-NEXT: [[BROADCAST_SPLATINSERT3:%.*]] = insertelement <8 x i16> undef, i16 [[B]], i32 0 +; CHECK-NEXT: [[BROADCAST_SPLAT4:%.*]] = shufflevector <8 x i16> [[BROADCAST_SPLATINSERT3]], <8 x i16> undef, <8 x i32> zeroinitializer +; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N_VEC]], -8 +; CHECK-NEXT: [[TMP1:%.*]] = lshr i32 [[TMP0]], 3 +; CHECK-NEXT: [[TMP2:%.*]] = add nuw nsw i32 [[TMP1]], 1 +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP2]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <8 x i16> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[INSTTMP6:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP3:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP7:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP4:%.*]] = phi i32 [ [[N]], %[[VECTOR_PH]] ], [ [[TMP6:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[INSTTMP2:%.*]] = getelementptr inbounds i16, ptr [[A]], i32 [[INDEX]] +; CHECK-NEXT: [[TMP5:%.*]] = call <8 x i1> @llvm.arm.mve.vctp16(i32 [[TMP4]]) +; CHECK-NEXT: [[TMP6]] = sub i32 [[TMP4]], 8 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[INSTTMP2]], i32 4, <8 x i1> [[TMP5]], <8 x i16> undef) +; CHECK-NEXT: [[INSTTMP5:%.*]] = add <8 x i16> [[VEC_PHI]], [[BROADCAST_SPLAT4]] +; CHECK-NEXT: [[INSTTMP6]] = add <8 x i16> [[INSTTMP5]], [[WIDE_MASKED_LOAD]] +; CHECK-NEXT: [[INDEX_NEXT]] = add nuw nsw i32 [[INDEX]], 8 +; CHECK-NEXT: [[TMP7]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP3]], i32 1) +; CHECK-NEXT: [[TMP8:%.*]] = icmp ne i32 [[TMP7]], 0 +; CHECK-NEXT: br i1 [[TMP8]], label %[[VECTOR_BODY]], label %[[MIDDLE_BLOCK:.*]] +; CHECK: [[MIDDLE_BLOCK]]: +; CHECK-NEXT: [[INSTTMP8:%.*]] = select <8 x i1> [[TMP5]], <8 x i16> [[INSTTMP6]], <8 x i16> [[VEC_PHI]] +; CHECK-NEXT: [[RDX_SHUF:%.*]] = shufflevector <8 x i16> [[INSTTMP8]], <8 x i16> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison> +; CHECK-NEXT: [[BIN_RDX:%.*]] = add <8 x i16> [[RDX_SHUF]], [[INSTTMP8]] +; CHECK-NEXT: [[RDX_SHUF5:%.*]] = shufflevector <8 x i16> [[BIN_RDX]], <8 x i16> undef, <8 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison> +; CHECK-NEXT: [[BIN_RDX6:%.*]] = add <8 x i16> [[RDX_SHUF5]], [[BIN_RDX]] +; CHECK-NEXT: [[RDX_SHUF7:%.*]] = shufflevector <8 x i16> [[BIN_RDX6]], <8 x i16> undef, <8 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison> +; CHECK-NEXT: [[BIN_RDX8:%.*]] = add <8 x i16> [[RDX_SHUF7]], [[BIN_RDX6]] +; CHECK-NEXT: [[TMP9:%.*]] = extractelement <8 x i16> [[BIN_RDX8]], i32 0 +; CHECK-NEXT: ret i16 [[TMP9]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: [[RES_0:%.*]] = phi i16 [ 0, %[[ENTRY]] ] +; CHECK-NEXT: ret i16 [[RES_0]] +; entry: %cmp8 = icmp eq i32 %N, 0 br i1 %cmp8, label %for.cond.cleanup, label %vector.ph @@ -119,15 +195,46 @@ for.cond.cleanup: ; despite this we can still calculate a precise enough range so that the ; the overflow checks for get.active.active.lane.mask don't reject ; tail-predication. -; -; CHECK-LABEL: @reduction_not_guarded -; -; CHECK: vector.body: -; CHECK: @llvm.arm.mve.vctp -; CHECK-NOT: @llvm.get.active.lane.mask.v8i1.i32 -; CHECK: ret -; define i16 @reduction_not_guarded(ptr nocapture readonly %A, i16 %B, i32 %N) local_unnamed_addr { +; CHECK-LABEL: define i16 @reduction_not_guarded( +; CHECK-SAME: ptr readonly captures(none) [[A:%.*]], i16 [[B:%.*]], i32 [[N:%.*]]) local_unnamed_addr #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: [[TMP:%.*]] = add i32 [[N]], -1 +; CHECK-NEXT: [[N_RND_UP:%.*]] = add nuw nsw i32 [[TMP]], 8 +; CHECK-NEXT: [[N_VEC:%.*]] = and i32 [[N_RND_UP]], -8 +; CHECK-NEXT: [[BROADCAST_SPLATINSERT3:%.*]] = insertelement <8 x i16> undef, i16 [[B]], i32 0 +; CHECK-NEXT: [[BROADCAST_SPLAT4:%.*]] = shufflevector <8 x i16> [[BROADCAST_SPLATINSERT3]], <8 x i16> undef, <8 x i32> zeroinitializer +; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N_VEC]], -8 +; CHECK-NEXT: [[TMP1:%.*]] = lshr i32 [[TMP0]], 3 +; CHECK-NEXT: [[TMP2:%.*]] = add nuw nsw i32 [[TMP1]], 1 +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP2]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <8 x i16> [ zeroinitializer, %[[ENTRY]] ], [ [[INSTTMP6:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP3:%.*]] = phi i32 [ [[START]], %[[ENTRY]] ], [ [[TMP7:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP4:%.*]] = phi i32 [ [[N]], %[[ENTRY]] ], [ [[TMP6:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[INSTTMP2:%.*]] = getelementptr inbounds i16, ptr [[A]], i32 [[INDEX]] +; CHECK-NEXT: [[TMP5:%.*]] = call <8 x i1> @llvm.arm.mve.vctp16(i32 [[TMP4]]) +; CHECK-NEXT: [[TMP6]] = sub i32 [[TMP4]], 8 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[INSTTMP2]], i32 4, <8 x i1> [[TMP5]], <8 x i16> undef) +; CHECK-NEXT: [[INSTTMP5:%.*]] = add <8 x i16> [[VEC_PHI]], [[BROADCAST_SPLAT4]] +; CHECK-NEXT: [[INSTTMP6]] = add <8 x i16> [[INSTTMP5]], [[WIDE_MASKED_LOAD]] +; CHECK-NEXT: [[INDEX_NEXT]] = add nuw nsw i32 [[INDEX]], 8 +; CHECK-NEXT: [[TMP7]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP3]], i32 1) +; CHECK-NEXT: [[TMP8:%.*]] = icmp ne i32 [[TMP7]], 0 +; CHECK-NEXT: br i1 [[TMP8]], label %[[VECTOR_BODY]], label %[[MIDDLE_BLOCK:.*]] +; CHECK: [[MIDDLE_BLOCK]]: +; CHECK-NEXT: [[INSTTMP8:%.*]] = select <8 x i1> [[TMP5]], <8 x i16> [[INSTTMP6]], <8 x i16> [[VEC_PHI]] +; CHECK-NEXT: [[RDX_SHUF:%.*]] = shufflevector <8 x i16> [[INSTTMP8]], <8 x i16> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison> +; CHECK-NEXT: [[BIN_RDX:%.*]] = add <8 x i16> [[RDX_SHUF]], [[INSTTMP8]] +; CHECK-NEXT: [[RDX_SHUF5:%.*]] = shufflevector <8 x i16> [[BIN_RDX]], <8 x i16> undef, <8 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison> +; CHECK-NEXT: [[BIN_RDX6:%.*]] = add <8 x i16> [[RDX_SHUF5]], [[BIN_RDX]] +; CHECK-NEXT: [[RDX_SHUF7:%.*]] = shufflevector <8 x i16> [[BIN_RDX6]], <8 x i16> undef, <8 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison> +; CHECK-NEXT: [[BIN_RDX8:%.*]] = add <8 x i16> [[RDX_SHUF7]], [[BIN_RDX6]] +; CHECK-NEXT: [[TMP9:%.*]] = extractelement <8 x i16> [[BIN_RDX8]], i32 0 +; CHECK-NEXT: ret i16 [[TMP9]] +; entry: %tmp = add i32 %N, -1 %n.rnd.up = add nuw nsw i32 %tmp, 8 @@ -166,12 +273,76 @@ middle.block: ; preds = %vector.body ret i16 %tmp9 } -; CHECK-LABEL: @Correlation -; CHECK: vector.body: -; CHECK: @llvm.arm.mve.vctp -; CHECK-NOT: %active.lane.mask = call <4 x i1> @llvm.get.active.lane.mask -; define dso_local void @Correlation(ptr nocapture readonly %Input, ptr nocapture %Output, i16 signext %Size, i16 signext %N, i16 signext %Scale) local_unnamed_addr #0 { +; CHECK-LABEL: define dso_local void @Correlation( +; CHECK-SAME: ptr readonly captures(none) [[INPUT:%.*]], ptr captures(none) [[OUTPUT:%.*]], i16 signext [[SIZE:%.*]], i16 signext [[N:%.*]], i16 signext [[SCALE:%.*]]) local_unnamed_addr #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CONV:%.*]] = sext i16 [[N]] to i32 +; CHECK-NEXT: [[CMP36:%.*]] = icmp sgt i16 [[N]], 0 +; CHECK-NEXT: br i1 [[CMP36]], label %[[FOR_BODY_LR_PH:.*]], label %[[FOR_END17:.*]] +; CHECK: [[FOR_BODY_LR_PH]]: +; CHECK-NEXT: [[CONV2:%.*]] = sext i16 [[SIZE]] to i32 +; CHECK-NEXT: [[CONV1032:%.*]] = zext i16 [[SCALE]] to i32 +; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[CONV2]], 3 +; CHECK-NEXT: br label %[[FOR_BODY:.*]] +; CHECK: [[FOR_BODY]]: +; CHECK-NEXT: [[LSR_IV51:%.*]] = phi i32 [ [[LSR_IV_NEXT:%.*]], %[[FOR_END:.*]] ], [ [[TMP0]], %[[FOR_BODY_LR_PH]] ] +; CHECK-NEXT: [[LSR_IV46:%.*]] = phi ptr [ [[SCEVGEP47:%.*]], %[[FOR_END]] ], [ [[INPUT]], %[[FOR_BODY_LR_PH]] ] +; CHECK-NEXT: [[I_037:%.*]] = phi i32 [ 0, %[[FOR_BODY_LR_PH]] ], [ [[INC16:%.*]], %[[FOR_END]] ] +; CHECK-NEXT: [[TMP1:%.*]] = mul nsw i32 [[I_037]], -1 +; CHECK-NEXT: [[TMP2:%.*]] = add i32 [[CONV2]], [[TMP1]] +; CHECK-NEXT: [[TMP3:%.*]] = mul nsw i32 [[I_037]], -1 +; CHECK-NEXT: [[TMP4:%.*]] = add i32 [[TMP0]], [[TMP3]] +; CHECK-NEXT: [[TMP5:%.*]] = lshr i32 [[TMP4]], 2 +; CHECK-NEXT: [[TMP6:%.*]] = shl nuw i32 [[TMP5]], 2 +; CHECK-NEXT: [[TMP7:%.*]] = add i32 [[TMP6]], -4 +; CHECK-NEXT: [[TMP8:%.*]] = lshr i32 [[TMP7]], 2 +; CHECK-NEXT: [[TMP9:%.*]] = add nuw nsw i32 [[TMP8]], 1 +; CHECK-NEXT: [[CMP433:%.*]] = icmp slt i32 [[I_037]], [[CONV2]] +; CHECK-NEXT: br i1 [[CMP433]], label %[[VECTOR_PH:.*]], label %[[FOR_END]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP9]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[LSR_IV48:%.*]] = phi ptr [ [[SCEVGEP49:%.*]], %[[VECTOR_BODY]] ], [ [[LSR_IV46]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[LSR_IV:%.*]] = phi ptr [ [[SCEVGEP:%.*]], %[[VECTOR_BODY]] ], [ [[INPUT]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP20:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP10:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP21:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP11:%.*]] = phi i32 [ [[TMP2]], %[[VECTOR_PH]] ], [ [[TMP13:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP12:%.*]] = call <4 x i1> @llvm.arm.mve.vctp32(i32 [[TMP11]]) +; CHECK-NEXT: [[TMP13]] = sub i32 [[TMP11]], 4 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <4 x i16> @llvm.masked.load.v4i16.p0(ptr [[LSR_IV]], i32 2, <4 x i1> [[TMP12]], <4 x i16> undef) +; CHECK-NEXT: [[TMP14:%.*]] = sext <4 x i16> [[WIDE_MASKED_LOAD]] to <4 x i32> +; CHECK-NEXT: [[WIDE_MASKED_LOAD42:%.*]] = call <4 x i16> @llvm.masked.load.v4i16.p0(ptr [[LSR_IV48]], i32 2, <4 x i1> [[TMP12]], <4 x i16> undef) +; CHECK-NEXT: [[TMP15:%.*]] = sext <4 x i16> [[WIDE_MASKED_LOAD42]] to <4 x i32> +; CHECK-NEXT: [[TMP16:%.*]] = mul nsw <4 x i32> [[TMP15]], [[TMP14]] +; CHECK-NEXT: [[TMP17:%.*]] = insertelement <4 x i32> undef, i32 [[CONV1032]], i32 0 +; CHECK-NEXT: [[TMP18:%.*]] = shufflevector <4 x i32> [[TMP17]], <4 x i32> undef, <4 x i32> zeroinitializer +; CHECK-NEXT: [[TMP19:%.*]] = ashr <4 x i32> [[TMP16]], [[TMP18]] +; CHECK-NEXT: [[TMP20]] = add <4 x i32> [[TMP19]], [[VEC_PHI]] +; CHECK-NEXT: [[SCEVGEP]] = getelementptr i16, ptr [[LSR_IV]], i32 4 +; CHECK-NEXT: [[SCEVGEP49]] = getelementptr i16, ptr [[LSR_IV48]], i32 4 +; CHECK-NEXT: [[TMP21]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP10]], i32 1) +; CHECK-NEXT: [[TMP22:%.*]] = icmp ne i32 [[TMP21]], 0 +; CHECK-NEXT: br i1 [[TMP22]], label %[[VECTOR_BODY]], label %[[MIDDLE_BLOCK:.*]] +; CHECK: [[MIDDLE_BLOCK]]: +; CHECK-NEXT: [[TMP23:%.*]] = select <4 x i1> [[TMP12]], <4 x i32> [[TMP20]], <4 x i32> [[VEC_PHI]] +; CHECK-NEXT: [[TMP24:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP23]]) +; CHECK-NEXT: br label %[[FOR_END]] +; CHECK: [[FOR_END]]: +; CHECK-NEXT: [[SUM_0_LCSSA:%.*]] = phi i32 [ 0, %[[FOR_BODY]] ], [ [[TMP24]], %[[MIDDLE_BLOCK]] ] +; CHECK-NEXT: [[TMP25:%.*]] = lshr i32 [[SUM_0_LCSSA]], 16 +; CHECK-NEXT: [[CONV13:%.*]] = trunc i32 [[TMP25]] to i16 +; CHECK-NEXT: [[ARRAYIDX14:%.*]] = getelementptr inbounds i16, ptr [[OUTPUT]], i32 [[I_037]] +; CHECK-NEXT: store i16 [[CONV13]], ptr [[ARRAYIDX14]], align 2 +; CHECK-NEXT: [[INC16]] = add nuw nsw i32 [[I_037]], 1 +; CHECK-NEXT: [[SCEVGEP47]] = getelementptr i16, ptr [[LSR_IV46]], i32 1 +; CHECK-NEXT: [[LSR_IV_NEXT]] = add i32 [[LSR_IV51]], -1 +; CHECK-NEXT: [[EXITCOND39:%.*]] = icmp eq i32 [[INC16]], [[CONV]] +; CHECK-NEXT: br i1 [[EXITCOND39]], label %[[FOR_END17]], label %[[FOR_BODY]] +; CHECK: [[FOR_END17]]: +; CHECK-NEXT: ret void +; entry: %conv = sext i16 %N to i32 %cmp36 = icmp sgt i16 %N, 0 diff --git a/llvm/test/CodeGen/Thumb2/LowOverheadLoops/tail-pred-widen.ll b/llvm/test/CodeGen/Thumb2/LowOverheadLoops/tail-pred-widen.ll index a8ad360..b54d526 100644 --- a/llvm/test/CodeGen/Thumb2/LowOverheadLoops/tail-pred-widen.ll +++ b/llvm/test/CodeGen/Thumb2/LowOverheadLoops/tail-pred-widen.ll @@ -1,8 +1,43 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 6 ; RUN: opt -mtriple=thumbv8.1m.main -mve-tail-predication -tail-predication=enabled -mattr=+mve,+lob %s -S -o - | FileCheck %s -; CHECK-LABEL: expand_v8i16_v8i32 -; CHECK-NOT: call i32 @llvm.arm.mve.vctp define void @expand_v8i16_v8i32(ptr noalias nocapture readonly %a, ptr noalias nocapture readonly %b, ptr noalias nocapture %c, i32 %N) { +; CHECK-LABEL: define void @expand_v8i16_v8i32( +; CHECK-SAME: ptr noalias readonly captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], ptr noalias captures(none) [[C:%.*]], i32 [[N:%.*]]) #[[ATTR0:[0-9]+]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp eq i32 [[N]], 0 +; CHECK-NEXT: [[TMP8:%.*]] = add i32 [[N]], 7 +; CHECK-NEXT: [[TMP9:%.*]] = lshr i32 [[TMP8]], 3 +; CHECK-NEXT: [[TMP10:%.*]] = shl nuw i32 [[TMP9]], 3 +; CHECK-NEXT: [[TMP11:%.*]] = add i32 [[TMP10]], -8 +; CHECK-NEXT: [[TMP12:%.*]] = lshr i32 [[TMP11]], 3 +; CHECK-NEXT: [[TMP13:%.*]] = add nuw nsw i32 [[TMP12]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[FOR_COND_CLEANUP:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP13]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP14:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP15:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP0:%.*]] = phi i32 [ [[N]], %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP:%.*]] = getelementptr inbounds i16, ptr [[A]], i32 [[INDEX]] +; CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.vctp16(i32 [[TMP0]]) +; CHECK-NEXT: [[TMP2]] = sub i32 [[TMP0]], 8 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = tail call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[TMP]], i32 4, <8 x i1> [[TMP1]], <8 x i16> undef) +; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i16, ptr [[B]], i32 [[INDEX]] +; CHECK-NEXT: [[WIDE_MASKED_LOAD2:%.*]] = tail call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[TMP3]], i32 4, <8 x i1> [[TMP1]], <8 x i16> undef) +; CHECK-NEXT: [[EXPAND_1:%.*]] = zext <8 x i16> [[WIDE_MASKED_LOAD]] to <8 x i32> +; CHECK-NEXT: [[EXPAND_2:%.*]] = zext <8 x i16> [[WIDE_MASKED_LOAD2]] to <8 x i32> +; CHECK-NEXT: [[MUL:%.*]] = mul nsw <8 x i32> [[EXPAND_2]], [[EXPAND_1]] +; CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[C]], i32 [[INDEX]] +; CHECK-NEXT: tail call void @llvm.masked.store.v8i32.p0(<8 x i32> [[MUL]], ptr [[TMP6]], i32 4, <8 x i1> [[TMP1]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 8 +; CHECK-NEXT: [[TMP15]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP14]], i32 1) +; CHECK-NEXT: [[TMP16:%.*]] = icmp ne i32 [[TMP15]], 0 +; CHECK-NEXT: br i1 [[TMP16]], label %[[VECTOR_BODY]], label %[[FOR_COND_CLEANUP]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp eq i32 %N, 0 %tmp8 = add i32 %N, 7 @@ -39,15 +74,57 @@ for.cond.cleanup: ; preds = %vector.body, %entry ret void } -; CHECK-LABEL: expand_v8i16_v4i32 -; CHECK: [[ELEMS:%[^ ]+]] = phi i32 [ %N, %vector.ph ], [ [[ELEMS_REM:%[^ ]+]], %vector.body ] -; CHECK: [[VCTP:%[^ ]+]] = call <8 x i1> @llvm.arm.mve.vctp16(i32 [[ELEMS]]) -; CHECK: [[ELEMS_REM]] = sub i32 [[ELEMS]], 8 -; CHECK: tail call <8 x i16> @llvm.masked.load.v8i16.p0(ptr {{.*}}, i32 4, <8 x i1> [[VCTP]], <8 x i16> undef) -; CHECK: %store.pred = icmp ule <4 x i32> %induction.store -; CHECK: tail call void @llvm.masked.store.v4i32.p0(<4 x i32> {{.*}}, ptr {{.*}}, i32 4, <4 x i1> %store.pred) -; CHECK: tail call void @llvm.masked.store.v4i32.p0(<4 x i32> {{.*}}, ptr {{.*}}, i32 4, <4 x i1> %store.pred) define void @expand_v8i16_v4i32(ptr readonly %a, ptr readonly %b, ptr %c, ptr %d, i32 %N) { +; CHECK-LABEL: define void @expand_v8i16_v4i32( +; CHECK-SAME: ptr readonly [[A:%.*]], ptr readonly [[B:%.*]], ptr [[C:%.*]], ptr [[D:%.*]], i32 [[N:%.*]]) #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp eq i32 [[N]], 0 +; CHECK-NEXT: [[TMP8:%.*]] = add i32 [[N]], 7 +; CHECK-NEXT: [[TMP9:%.*]] = lshr i32 [[TMP8]], 3 +; CHECK-NEXT: [[TMP10:%.*]] = shl nuw i32 [[TMP9]], 3 +; CHECK-NEXT: [[TMP11:%.*]] = add i32 [[TMP10]], -8 +; CHECK-NEXT: [[TMP12:%.*]] = lshr i32 [[TMP11]], 3 +; CHECK-NEXT: [[TMP13:%.*]] = add nuw nsw i32 [[TMP12]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[FOR_COND_CLEANUP:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[TRIP_COUNT_MINUS_1:%.*]] = add i32 [[N]], -1 +; CHECK-NEXT: [[BROADCAST_SPLATINSERT10_STORE:%.*]] = insertelement <4 x i32> undef, i32 [[TRIP_COUNT_MINUS_1]], i32 0 +; CHECK-NEXT: [[BROADCAST_SPLAT11_STORE:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT10_STORE]], <4 x i32> undef, <4 x i32> zeroinitializer +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP13]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[STORE_IDX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[STORE_IDX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP14:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP15:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP0:%.*]] = phi i32 [ [[N]], %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP:%.*]] = getelementptr inbounds i16, ptr [[A]], i32 [[INDEX]] +; CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.vctp16(i32 [[TMP0]]) +; CHECK-NEXT: [[TMP2]] = sub i32 [[TMP0]], 8 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = tail call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[TMP]], i32 4, <8 x i1> [[TMP1]], <8 x i16> undef) +; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i16, ptr [[B]], i32 [[INDEX]] +; CHECK-NEXT: [[WIDE_MASKED_LOAD2:%.*]] = tail call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[TMP3]], i32 4, <8 x i1> [[TMP1]], <8 x i16> undef) +; CHECK-NEXT: [[EXTRACT_2_LOW:%.*]] = shufflevector <8 x i16> [[WIDE_MASKED_LOAD2]], <8 x i16> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3> +; CHECK-NEXT: [[EXTRACT_2_HIGH:%.*]] = shufflevector <8 x i16> [[WIDE_MASKED_LOAD2]], <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7> +; CHECK-NEXT: [[EXPAND_1:%.*]] = zext <4 x i16> [[EXTRACT_2_LOW]] to <4 x i32> +; CHECK-NEXT: [[EXPAND_2:%.*]] = zext <4 x i16> [[EXTRACT_2_HIGH]] to <4 x i32> +; CHECK-NEXT: [[MUL:%.*]] = mul nsw <4 x i32> [[EXPAND_2]], [[EXPAND_1]] +; CHECK-NEXT: [[SUB:%.*]] = mul nsw <4 x i32> [[EXPAND_1]], [[EXPAND_2]] +; CHECK-NEXT: [[BROADCAST_SPLATINSERT_STORE:%.*]] = insertelement <4 x i32> undef, i32 [[STORE_IDX]], i32 0 +; CHECK-NEXT: [[BROADCAST_SPLAT_STORE:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT_STORE]], <4 x i32> undef, <4 x i32> zeroinitializer +; CHECK-NEXT: [[INDUCTION_STORE:%.*]] = add <4 x i32> [[BROADCAST_SPLAT_STORE]], <i32 0, i32 1, i32 2, i32 3> +; CHECK-NEXT: [[STORE_PRED:%.*]] = icmp ule <4 x i32> [[INDUCTION_STORE]], [[BROADCAST_SPLAT11_STORE]] +; CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds i32, ptr [[C]], i32 [[STORE_IDX]] +; CHECK-NEXT: tail call void @llvm.masked.store.v4i32.p0(<4 x i32> [[MUL]], ptr [[TMP6]], i32 4, <4 x i1> [[STORE_PRED]]) +; CHECK-NEXT: [[GEP:%.*]] = getelementptr inbounds i32, ptr [[D]], i32 [[STORE_IDX]] +; CHECK-NEXT: tail call void @llvm.masked.store.v4i32.p0(<4 x i32> [[SUB]], ptr [[GEP]], i32 4, <4 x i1> [[STORE_PRED]]) +; CHECK-NEXT: [[STORE_IDX_NEXT]] = add i32 [[STORE_IDX]], 4 +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 8 +; CHECK-NEXT: [[TMP15]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP14]], i32 1) +; CHECK-NEXT: [[TMP16:%.*]] = icmp ne i32 [[TMP15]], 0 +; CHECK-NEXT: br i1 [[TMP16]], label %[[VECTOR_BODY]], label %[[FOR_COND_CLEANUP]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp eq i32 %N, 0 %tmp8 = add i32 %N, 7 @@ -98,9 +175,43 @@ for.cond.cleanup: ; preds = %vector.body, %entry ret void } -; CHECK-LABEL: expand_v4i32_v4i64 -; CHECK-NOT: call i32 @llvm.arm.mve.vctp define void @expand_v4i32_v4i64(ptr noalias nocapture readonly %a, ptr noalias nocapture readonly %b, ptr noalias nocapture %c, i32 %N) { +; CHECK-LABEL: define void @expand_v4i32_v4i64( +; CHECK-SAME: ptr noalias readonly captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], ptr noalias captures(none) [[C:%.*]], i32 [[N:%.*]]) #[[ATTR0]] { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: [[CMP8:%.*]] = icmp eq i32 [[N]], 0 +; CHECK-NEXT: [[TMP8:%.*]] = add i32 [[N]], 3 +; CHECK-NEXT: [[TMP9:%.*]] = lshr i32 [[TMP8]], 2 +; CHECK-NEXT: [[TMP10:%.*]] = shl nuw i32 [[TMP9]], 2 +; CHECK-NEXT: [[TMP11:%.*]] = add i32 [[TMP10]], -4 +; CHECK-NEXT: [[TMP12:%.*]] = lshr i32 [[TMP11]], 2 +; CHECK-NEXT: [[TMP13:%.*]] = add nuw nsw i32 [[TMP12]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[FOR_COND_CLEANUP:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP13]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP14:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP15:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP0:%.*]] = phi i32 [ [[N]], %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP:%.*]] = getelementptr inbounds i32, ptr [[A]], i32 [[INDEX]] +; CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.vctp32(i32 [[TMP0]]) +; CHECK-NEXT: [[TMP2]] = sub i32 [[TMP0]], 4 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[TMP]], i32 4, <4 x i1> [[TMP1]], <4 x i32> undef) +; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds i32, ptr [[B]], i32 [[INDEX]] +; CHECK-NEXT: [[WIDE_MASKED_LOAD2:%.*]] = tail call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[TMP3]], i32 4, <4 x i1> [[TMP1]], <4 x i32> undef) +; CHECK-NEXT: [[EXPAND_1:%.*]] = zext <4 x i32> [[WIDE_MASKED_LOAD]] to <4 x i64> +; CHECK-NEXT: [[EXPAND_2:%.*]] = zext <4 x i32> [[WIDE_MASKED_LOAD2]] to <4 x i64> +; CHECK-NEXT: [[MUL:%.*]] = mul nsw <4 x i64> [[EXPAND_2]], [[EXPAND_1]] +; CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds i64, ptr [[C]], i32 [[INDEX]] +; CHECK-NEXT: tail call void @llvm.masked.store.v4i64.p0(<4 x i64> [[MUL]], ptr [[TMP6]], i32 4, <4 x i1> [[TMP1]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add i32 [[INDEX]], 4 +; CHECK-NEXT: [[TMP15]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP14]], i32 1) +; CHECK-NEXT: [[TMP16:%.*]] = icmp ne i32 [[TMP15]], 0 +; CHECK-NEXT: br i1 [[TMP16]], label %[[VECTOR_BODY]], label %[[FOR_COND_CLEANUP]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: ret void +; entry: %cmp8 = icmp eq i32 %N, 0 %tmp8 = add i32 %N, 3 diff --git a/llvm/test/CodeGen/Thumb2/LowOverheadLoops/vector-reduce-mve-tail.ll b/llvm/test/CodeGen/Thumb2/LowOverheadLoops/vector-reduce-mve-tail.ll index ec542df..fb1a4a4 100644 --- a/llvm/test/CodeGen/Thumb2/LowOverheadLoops/vector-reduce-mve-tail.ll +++ b/llvm/test/CodeGen/Thumb2/LowOverheadLoops/vector-reduce-mve-tail.ll @@ -1,24 +1,47 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 6 ; RUN: opt -mtriple=thumbv8.1m.main -mve-tail-predication -tail-predication=enabled -mattr=+mve %s -S -o - | FileCheck %s -; CHECK-LABEL: vec_mul_reduce_add - -; CHECK: vector.ph: -; CHECK: %start = call i32 @llvm.start.loop.iterations.i32 -; CHECK: br label %vector.body - -; CHECK: vector.body: -; CHECK: [[ELTS:%[^ ]+]] = phi i32 [ %N, %vector.ph ], [ [[SUB:%[^ ]+]], %vector.body ] -; CHECK: [[VCTP:%[^ ]+]] = call <4 x i1> @llvm.arm.mve.vctp32(i32 [[ELTS]]) -; CHECK: [[SUB]] = sub i32 [[ELTS]], 4 -; CHECK: call <4 x i32> @llvm.masked.load.v4i32.p0(ptr {{.*}}, i32 4, <4 x i1> [[VCTP]] -; CHECK: call <4 x i32> @llvm.masked.load.v4i32.p0(ptr {{.*}}, i32 4, <4 x i1> [[VCTP]], - -; CHECK: middle.block: -; CHECK: [[VPSEL:%[^ ]+]] = select <4 x i1> [[VCTP]], -; CHECK: call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[VPSEL]]) - define i32 @vec_mul_reduce_add(ptr noalias nocapture readonly %a, ptr noalias nocapture readonly %b, i32 %N) { +; CHECK-LABEL: define i32 @vec_mul_reduce_add( +; CHECK-SAME: ptr noalias readonly captures(none) [[A:%.*]], ptr noalias readonly captures(none) [[B:%.*]], i32 [[N:%.*]]) #[[ATTR0:[0-9]+]] { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: [[CMP8:%.*]] = icmp eq i32 [[N]], 0 +; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N]], 3 +; CHECK-NEXT: [[TMP1:%.*]] = lshr i32 [[TMP0]], 2 +; CHECK-NEXT: [[TMP2:%.*]] = shl nuw i32 [[TMP1]], 2 +; CHECK-NEXT: [[TMP3:%.*]] = add i32 [[TMP2]], -4 +; CHECK-NEXT: [[TMP4:%.*]] = lshr i32 [[TMP3]], 2 +; CHECK-NEXT: [[TMP5:%.*]] = add nuw nsw i32 [[TMP4]], 1 +; CHECK-NEXT: br i1 [[CMP8]], label %[[FOR_COND_CLEANUP:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[START:%.*]] = call i32 @llvm.start.loop.iterations.i32(i32 [[TMP5]]) +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[LSR_IV2:%.*]] = phi ptr [ [[SCEVGEP3:%.*]], %[[VECTOR_BODY]] ], [ [[A]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[LSR_IV:%.*]] = phi ptr [ [[SCEVGEP:%.*]], %[[VECTOR_BODY]] ], [ [[B]], %[[VECTOR_PH]] ] +; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP11:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP6:%.*]] = phi i32 [ [[START]], %[[VECTOR_PH]] ], [ [[TMP12:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP7:%.*]] = phi i32 [ [[N]], %[[VECTOR_PH]] ], [ [[TMP9:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP8:%.*]] = call <4 x i1> @llvm.arm.mve.vctp32(i32 [[TMP7]]) +; CHECK-NEXT: [[TMP9]] = sub i32 [[TMP7]], 4 +; CHECK-NEXT: [[WIDE_MASKED_LOAD:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[LSR_IV2]], i32 4, <4 x i1> [[TMP8]], <4 x i32> undef) +; CHECK-NEXT: [[WIDE_MASKED_LOAD13:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[LSR_IV]], i32 4, <4 x i1> [[TMP8]], <4 x i32> undef) +; CHECK-NEXT: [[TMP10:%.*]] = mul nsw <4 x i32> [[WIDE_MASKED_LOAD13]], [[WIDE_MASKED_LOAD]] +; CHECK-NEXT: [[TMP11]] = add nsw <4 x i32> [[TMP10]], [[VEC_PHI]] +; CHECK-NEXT: [[SCEVGEP]] = getelementptr i32, ptr [[LSR_IV]], i32 4 +; CHECK-NEXT: [[SCEVGEP3]] = getelementptr i32, ptr [[LSR_IV2]], i32 4 +; CHECK-NEXT: [[TMP12]] = call i32 @llvm.loop.decrement.reg.i32(i32 [[TMP6]], i32 1) +; CHECK-NEXT: [[TMP13:%.*]] = icmp ne i32 [[TMP12]], 0 +; CHECK-NEXT: br i1 [[TMP13]], label %[[VECTOR_BODY]], label %[[MIDDLE_BLOCK:.*]] +; CHECK: [[MIDDLE_BLOCK]]: +; CHECK-NEXT: [[TMP14:%.*]] = select <4 x i1> [[TMP8]], <4 x i32> [[TMP11]], <4 x i32> [[VEC_PHI]] +; CHECK-NEXT: [[TMP15:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[TMP14]]) +; CHECK-NEXT: br label %[[FOR_COND_CLEANUP]] +; CHECK: [[FOR_COND_CLEANUP]]: +; CHECK-NEXT: [[RES_0_LCSSA:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[TMP15]], %[[MIDDLE_BLOCK]] ] +; CHECK-NEXT: ret i32 [[RES_0_LCSSA]] +; entry: %cmp8 = icmp eq i32 %N, 0 %0 = add i32 %N, 3 diff --git a/llvm/test/CodeGen/WebAssembly/bulk-memory.ll b/llvm/test/CodeGen/WebAssembly/bulk-memory.ll index ae170d7..d949068 100644 --- a/llvm/test/CodeGen/WebAssembly/bulk-memory.ll +++ b/llvm/test/CodeGen/WebAssembly/bulk-memory.ll @@ -104,6 +104,31 @@ define void @memset_i32(ptr %dest, i8 %val, i32 %len) { ret void } +; CHECK-LABEL: memcpy_0: +; CHECK-NEXT: .functype memcpy_0 (i32, i32) -> () +; CHECK-NEXT: return +define void @memcpy_0(ptr %dest, ptr %src) { + call void @llvm.memcpy.p0.p0.i32(ptr %dest, ptr %src, i32 0, i1 0) + ret void +} + +; CHECK-LABEL: memmove_0: +; CHECK-NEXT: .functype memmove_0 (i32, i32) -> () +; CHECK-NEXT: return +define void @memmove_0(ptr %dest, ptr %src) { + call void @llvm.memmove.p0.p0.i32(ptr %dest, ptr %src, i32 0, i1 0) + ret void +} + +; CHECK-LABEL: memset_0: +; NO-BULK-MEM-NOT: memory.fill +; BULK-MEM-NEXT: .functype memset_0 (i32, i32) -> () +; BULK-MEM-NEXT: return +define void @memset_0(ptr %dest, i8 %val) { + call void @llvm.memset.p0.i32(ptr %dest, i8 %val, i32 0, i1 0) + ret void +} + ; CHECK-LABEL: memcpy_1: ; CHECK-NEXT: .functype memcpy_1 (i32, i32) -> () ; CHECK-NEXT: i32.load8_u $push[[L0:[0-9]+]]=, 0($1) @@ -137,14 +162,8 @@ define void @memset_1(ptr %dest, i8 %val) { ; CHECK-LABEL: memcpy_1024: ; NO-BULK-MEM-NOT: memory.copy ; BULK-MEM-NEXT: .functype memcpy_1024 (i32, i32) -> () -; BULK-MEM-NEXT: block ; BULK-MEM-NEXT: i32.const $push[[L0:[0-9]+]]=, 1024 -; BULK-MEM-NEXT: i32.eqz $push[[L1:[0-9]+]]=, $pop[[L0]] -; BULK-MEM-NEXT: br_if 0, $pop[[L1]] -; BULK-MEM-NEXT: i32.const $push[[L2:[0-9]+]]=, 1024 -; BULK-MEM-NEXT: memory.copy 0, 0, $0, $1, $pop[[L2]] -; BULK-MEM-NEXT: .LBB{{.*}}: -; BULK-MEM-NEXT: end_block +; BULK-MEM-NEXT: memory.copy 0, 0, $0, $1, $pop[[L0]] ; BULK-MEM-NEXT: return define void @memcpy_1024(ptr %dest, ptr %src) { call void @llvm.memcpy.p0.p0.i32(ptr %dest, ptr %src, i32 1024, i1 0) @@ -154,14 +173,8 @@ define void @memcpy_1024(ptr %dest, ptr %src) { ; CHECK-LABEL: memmove_1024: ; NO-BULK-MEM-NOT: memory.copy ; BULK-MEM-NEXT: .functype memmove_1024 (i32, i32) -> () -; BULK-MEM-NEXT: block ; BULK-MEM-NEXT: i32.const $push[[L0:[0-9]+]]=, 1024 -; BULK-MEM-NEXT: i32.eqz $push[[L1:[0-9]+]]=, $pop[[L0]] -; BULK-MEM-NEXT: br_if 0, $pop[[L1]] -; BULK-MEM-NEXT: i32.const $push[[L2:[0-9]+]]=, 1024 -; BULK-MEM-NEXT: memory.copy 0, 0, $0, $1, $pop[[L2]] -; BULK-MEM-NEXT: .LBB{{.*}}: -; BULK-MEM-NEXT: end_block +; BULK-MEM-NEXT: memory.copy 0, 0, $0, $1, $pop[[L0]] ; BULK-MEM-NEXT: return define void @memmove_1024(ptr %dest, ptr %src) { call void @llvm.memmove.p0.p0.i32(ptr %dest, ptr %src, i32 1024, i1 0) @@ -171,14 +184,8 @@ define void @memmove_1024(ptr %dest, ptr %src) { ; CHECK-LABEL: memset_1024: ; NO-BULK-MEM-NOT: memory.fill ; BULK-MEM-NEXT: .functype memset_1024 (i32, i32) -> () -; BULK-MEM-NEXT: block ; BULK-MEM-NEXT: i32.const $push[[L0:[0-9]+]]=, 1024 -; BULK-MEM-NEXT: i32.eqz $push[[L1:[0-9]+]]=, $pop[[L0]] -; BULK-MEM-NEXT: br_if 0, $pop[[L1]] -; BULK-MEM-NEXT: i32.const $push[[L2:[0-9]+]]=, 1024 -; BULK-MEM-NEXT: memory.fill 0, $0, $1, $pop[[L2]] -; BULK-MEM-NEXT: .LBB{{.*}}: -; BULK-MEM-NEXT: end_block +; BULK-MEM-NEXT: memory.fill 0, $0, $1, $pop[[L0]] ; BULK-MEM-NEXT: return define void @memset_1024(ptr %dest, i8 %val) { call void @llvm.memset.p0.i32(ptr %dest, i8 %val, i32 1024, i1 0) @@ -201,17 +208,11 @@ define void @memset_1024(ptr %dest, i8 %val) { ; BULK-MEM-NEXT: .functype memcpy_alloca_src (i32) -> () ; BULK-MEM-NEXT: global.get $push[[L0:[0-9]+]]=, __stack_pointer ; BULK-MEM-NEXT: i32.const $push[[L1:[0-9]+]]=, 112 -; BULK-MEM-NEXT: i32.sub $[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]] -; BULK-MEM-NEXT: block -; BULK-MEM-NEXT: i32.const $push[[L3:[0-9]+]]=, 100 -; BULK-MEM-NEXT: i32.eqz $push[[L4:[0-9]+]]=, $pop[[L3]] -; BULK-MEM-NEXT: br_if 0, $pop[[L4]] -; BULK-MEM-NEXT: i32.const $push[[L5:[0-9]+]]=, 12 -; BULK-MEM-NEXT: i32.add $push[[L6:[0-9]+]]=, $[[L2]], $pop[[L5]] -; BULK-MEM-NEXT: i32.const $push[[L7:[0-9]+]]=, 100 -; BULK-MEM-NEXT: memory.copy 0, 0, $0, $pop[[L6]], $pop[[L7]] -; BULK-MEM-NEXT: .LBB{{.*}}: -; BULK-MEM-NEXT: end_block +; BULK-MEM-NEXT: i32.sub $push[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]] +; BULK-MEM-NEXT: i32.const $push[[L3:[0-9]+]]=, 12 +; BULK-MEM-NEXT: i32.add $push[[L4:[0-9]+]]=, $pop[[L2]], $pop[[L3]] +; BULK-MEM-NEXT: i32.const $push[[L5:[0-9]+]]=, 100 +; BULK-MEM-NEXT: memory.copy 0, 0, $0, $pop[[L4]], $pop[[L5]] ; BULK-MEM-NEXT: return define void @memcpy_alloca_src(ptr %dst) { %a = alloca [100 x i8] @@ -224,17 +225,11 @@ define void @memcpy_alloca_src(ptr %dst) { ; BULK-MEM-NEXT: .functype memcpy_alloca_dst (i32) -> () ; BULK-MEM-NEXT: global.get $push[[L0:[0-9]+]]=, __stack_pointer ; BULK-MEM-NEXT: i32.const $push[[L1:[0-9]+]]=, 112 -; BULK-MEM-NEXT: i32.sub $[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]] -; BULK-MEM-NEXT: block -; BULK-MEM-NEXT: i32.const $push[[L3:[0-9]+]]=, 100 -; BULK-MEM-NEXT: i32.eqz $push[[L4:[0-9]+]]=, $pop[[L3]] -; BULK-MEM-NEXT: br_if 0, $pop[[L4]] -; BULK-MEM-NEXT: i32.const $push[[L5:[0-9]+]]=, 12 -; BULK-MEM-NEXT: i32.add $push[[L6:[0-9]+]]=, $[[L2]], $pop[[L5]] -; BULK-MEM-NEXT: i32.const $push[[L7:[0-9]+]]=, 100 -; BULK-MEM-NEXT: memory.copy 0, 0, $pop[[L6]], $0, $pop[[L7]] -; BULK-MEM-NEXT: .LBB{{.*}}: -; BULK-MEM-NEXT: end_block +; BULK-MEM-NEXT: i32.sub $push[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]] +; BULK-MEM-NEXT: i32.const $push[[L3:[0-9]+]]=, 12 +; BULK-MEM-NEXT: i32.add $push[[L4:[0-9]+]]=, $pop[[L2]], $pop[[L3]] +; BULK-MEM-NEXT: i32.const $push[[L5:[0-9]+]]=, 100 +; BULK-MEM-NEXT: memory.copy 0, 0, $pop[[L4]], $0, $pop[[L5]] ; BULK-MEM-NEXT: return define void @memcpy_alloca_dst(ptr %src) { %a = alloca [100 x i8] @@ -247,17 +242,11 @@ define void @memcpy_alloca_dst(ptr %src) { ; BULK-MEM-NEXT: .functype memset_alloca (i32) -> () ; BULK-MEM-NEXT: global.get $push[[L0:[0-9]+]]=, __stack_pointer ; BULK-MEM-NEXT: i32.const $push[[L1:[0-9]+]]=, 112 -; BULK-MEM-NEXT: i32.sub $1=, $pop[[L0]], $pop[[L1]] -; BULK-MEM-NEXT: block -; BULK-MEM-NEXT: i32.const $push[[L2:[0-9]+]]=, 100 -; BULK-MEM-NEXT: i32.eqz $push[[L3:[0-9]+]]=, $pop[[L2]] -; BULK-MEM-NEXT: br_if 0, $pop[[L3]] -; BULK-MEM-NEXT: i32.const $push[[L4:[0-9]+]]=, 12 -; BULK-MEM-NEXT: i32.add $push[[L5:[0-9]+]]=, $1, $pop[[L4]] -; BULK-MEM-NEXT: i32.const $push[[L6:[0-9]+]]=, 100 -; BULK-MEM-NEXT: memory.fill 0, $pop[[L5]], $0, $pop[[L6]] -; BULK-MEM-NEXT: .LBB{{.*}}: -; BULK-MEM-NEXT: end_block +; BULK-MEM-NEXT: i32.sub $push[[L2:[0-9]+]]=, $pop[[L0]], $pop[[L1]] +; BULK-MEM-NEXT: i32.const $push[[L3:[0-9]+]]=, 12 +; BULK-MEM-NEXT: i32.add $push[[L4:[0-9]+]]=, $pop[[L2]], $pop[[L3]] +; BULK-MEM-NEXT: i32.const $push[[L5:[0-9]+]]=, 100 +; BULK-MEM-NEXT: memory.fill 0, $pop[[L4]], $0, $pop[[L5]] ; BULK-MEM-NEXT: return define void @memset_alloca(i8 %val) { %a = alloca [100 x i8] diff --git a/llvm/test/CodeGen/WebAssembly/bulk-memory64.ll b/llvm/test/CodeGen/WebAssembly/bulk-memory64.ll index 0cf8493..d0206a3 100644 --- a/llvm/test/CodeGen/WebAssembly/bulk-memory64.ll +++ b/llvm/test/CodeGen/WebAssembly/bulk-memory64.ll @@ -110,6 +110,31 @@ define void @memset_i32(ptr %dest, i8 %val, i64 %len) { ret void } +; CHECK-LABEL: memcpy_0: +; CHECK-NEXT: .functype memcpy_0 (i64, i64) -> () +; CHECK-NEXT: return +define void @memcpy_0(ptr %dest, ptr %src) { + call void @llvm.memcpy.p0.p0.i64(ptr %dest, ptr %src, i64 0, i1 0) + ret void +} + +; CHECK-LABEL: memmove_0: +; CHECK-NEXT: .functype memmove_0 (i64, i64) -> () +; CHECK-NEXT: return +define void @memmove_0(ptr %dest, ptr %src) { + call void @llvm.memmove.p0.p0.i64(ptr %dest, ptr %src, i64 0, i1 0) + ret void +} + +; CHECK-LABEL: memset_0: +; NO-BULK-MEM-NOT: memory.fill +; BULK-MEM-NEXT: .functype memset_0 (i64, i32) -> () +; BULK-MEM-NEXT: return +define void @memset_0(ptr %dest, i8 %val) { + call void @llvm.memset.p0.i64(ptr %dest, i8 %val, i64 0, i1 0) + ret void +} + ; CHECK-LABEL: memcpy_1: ; CHECK-NEXT: .functype memcpy_1 (i64, i64) -> () ; CHECK-NEXT: i32.load8_u $push[[L0:[0-9]+]]=, 0($1) @@ -143,14 +168,8 @@ define void @memset_1(ptr %dest, i8 %val) { ; CHECK-LABEL: memcpy_1024: ; NO-BULK-MEM-NOT: memory.copy ; BULK-MEM-NEXT: .functype memcpy_1024 (i64, i64) -> () -; BULK-MEM-NEXT: block -; BULK-MEM-NEXT: i64.const $push[[L1:[0-9]+]]=, 1024 -; BULK-MEM-NEXT: i64.eqz $push0=, $pop[[L1]] -; BULK-MEM-NEXT: br_if 0, $pop0 ; BULK-MEM-NEXT: i64.const $push[[L0:[0-9]+]]=, 1024 ; BULK-MEM-NEXT: memory.copy 0, 0, $0, $1, $pop[[L0]] -; BULK-MEM-NEXT: .LBB{{.*}}: -; BULK-MEM-NEXT: end_block ; BULK-MEM-NEXT: return define void @memcpy_1024(ptr %dest, ptr %src) { call void @llvm.memcpy.p0.p0.i64(ptr %dest, ptr %src, i64 1024, i1 0) @@ -160,14 +179,8 @@ define void @memcpy_1024(ptr %dest, ptr %src) { ; CHECK-LABEL: memmove_1024: ; NO-BULK-MEM-NOT: memory.copy ; BULK-MEM-NEXT: .functype memmove_1024 (i64, i64) -> () -; BULK-MEM-NEXT: block -; BULK-MEM-NEXT: i64.const $push[[L1:[0-9]+]]=, 1024 -; BULK-MEM-NEXT: i64.eqz $push0=, $pop[[L1]] -; BULK-MEM-NEXT: br_if 0, $pop0 ; BULK-MEM-NEXT: i64.const $push[[L0:[0-9]+]]=, 1024 ; BULK-MEM-NEXT: memory.copy 0, 0, $0, $1, $pop[[L0]] -; BULK-MEM-NEXT: .LBB{{.*}}: -; BULK-MEM-NEXT: end_block ; BULK-MEM-NEXT: return define void @memmove_1024(ptr %dest, ptr %src) { call void @llvm.memmove.p0.p0.i64(ptr %dest, ptr %src, i64 1024, i1 0) @@ -177,14 +190,8 @@ define void @memmove_1024(ptr %dest, ptr %src) { ; CHECK-LABEL: memset_1024: ; NO-BULK-MEM-NOT: memory.fill ; BULK-MEM-NEXT: .functype memset_1024 (i64, i32) -> () -; BULK-MEM-NEXT: block -; BULK-MEM-NEXT: i64.const $push[[L1:[0-9]+]]=, 1024 -; BULK-MEM-NEXT: i64.eqz $push0=, $pop[[L1]] -; BULK-MEM-NEXT: br_if 0, $pop0 ; BULK-MEM-NEXT: i64.const $push[[L0:[0-9]+]]=, 1024 ; BULK-MEM-NEXT: memory.fill 0, $0, $1, $pop[[L0]] -; BULK-MEM-NEXT: .LBB{{.*}}: -; BULK-MEM-NEXT: end_block ; BULK-MEM-NEXT: return define void @memset_1024(ptr %dest, i8 %val) { call void @llvm.memset.p0.i64(ptr %dest, i8 %val, i64 1024, i1 0) @@ -207,17 +214,11 @@ define void @memset_1024(ptr %dest, i8 %val) { ; BULK-MEM-NEXT: .functype memcpy_alloca_src (i64) -> () ; BULK-MEM-NEXT: global.get $push[[L1:[0-9]+]]=, __stack_pointer ; BULK-MEM-NEXT: i64.const $push[[L0:[0-9]+]]=, 112 -; BULK-MEM-NEXT: i64.sub $[[L2:[0-9]+]]=, $pop[[L1]], $pop[[L0]] -; BULK-MEM-NEXT: block -; BULK-MEM-NEXT: i64.const $push[[L3:[0-9]+]]=, 100 -; BULK-MEM-NEXT: i64.eqz $push[[L4:[0-9]+]]=, $pop[[L3]] -; BULK-MEM-NEXT: br_if 0, $pop[[L4]] -; BULK-MEM-NEXT: i64.const $push[[L5:[0-9]+]]=, 12 -; BULK-MEM-NEXT: i64.add $push[[L6:[0-9]+]]=, $[[L2]], $pop[[L5]] -; BULK-MEM-NEXT: i64.const $push[[L7:[0-9]+]]=, 100 -; BULK-MEM-NEXT: memory.copy 0, 0, $0, $pop[[L6]], $pop[[L7]] -; BULK-MEM-NEXT: .LBB{{.*}}: -; BULK-MEM-NEXT: end_block +; BULK-MEM-NEXT: i64.sub $push[[L2:[0-9]+]]=, $pop[[L1]], $pop[[L0]] +; BULK-MEM-NEXT: i64.const $push[[L3:[0-9]+]]=, 12 +; BULK-MEM-NEXT: i64.add $push[[L4:[0-9]+]]=, $pop[[L2]], $pop[[L3]] +; BULK-MEM-NEXT: i64.const $push[[L5:[0-9]+]]=, 100 +; BULK-MEM-NEXT: memory.copy 0, 0, $0, $pop[[L4]], $pop[[L5]] ; BULK-MEM-NEXT: return define void @memcpy_alloca_src(ptr %dst) { %a = alloca [100 x i8] @@ -230,17 +231,11 @@ define void @memcpy_alloca_src(ptr %dst) { ; BULK-MEM-NEXT: .functype memcpy_alloca_dst (i64) -> () ; BULK-MEM-NEXT: global.get $push[[L1:[0-9]+]]=, __stack_pointer ; BULK-MEM-NEXT: i64.const $push[[L0:[0-9]+]]=, 112 -; BULK-MEM-NEXT: i64.sub $[[L2:[0-9]+]]=, $pop[[L1]], $pop[[L0]] -; BULK-MEM-NEXT: block -; BULK-MEM-NEXT: i64.const $push[[L3:[0-9]+]]=, 100 -; BULK-MEM-NEXT: i64.eqz $push[[L4:[0-9]+]]=, $pop[[L3]] -; BULK-MEM-NEXT: br_if 0, $pop[[L4]] -; BULK-MEM-NEXT: i64.const $push[[L5:[0-9]+]]=, 12 -; BULK-MEM-NEXT: i64.add $push[[L6:[0-9]+]]=, $[[L2]], $pop[[L5]] -; BULK-MEM-NEXT: i64.const $push[[L7:[0-9]+]]=, 100 -; BULK-MEM-NEXT: memory.copy 0, 0, $pop[[L6]], $0, $pop[[L7]] -; BULK-MEM-NEXT: .LBB{{.*}}: -; BULK-MEM-NEXT: end_block +; BULK-MEM-NEXT: i64.sub $push[[L2:[0-9]+]]=, $pop[[L1]], $pop[[L0]] +; BULK-MEM-NEXT: i64.const $push[[L3:[0-9]+]]=, 12 +; BULK-MEM-NEXT: i64.add $push[[L4:[0-9]+]]=, $pop[[L2]], $pop[[L3]] +; BULK-MEM-NEXT: i64.const $push[[L5:[0-9]+]]=, 100 +; BULK-MEM-NEXT: memory.copy 0, 0, $pop[[L4]], $0, $pop[[L5]] ; BULK-MEM-NEXT: return define void @memcpy_alloca_dst(ptr %src) { %a = alloca [100 x i8] @@ -253,17 +248,11 @@ define void @memcpy_alloca_dst(ptr %src) { ; BULK-MEM-NEXT: .functype memset_alloca (i32) -> () ; BULK-MEM-NEXT: global.get $push[[L1:[0-9]+]]=, __stack_pointer ; BULK-MEM-NEXT: i64.const $push[[L0:[0-9]+]]=, 112 -; BULK-MEM-NEXT: i64.sub $1=, $pop[[L1]], $pop[[L0]] -; BULK-MEM-NEXT: block -; BULK-MEM-NEXT: i64.const $push[[L2:[0-9]+]]=, 100 -; BULK-MEM-NEXT: i64.eqz $push[[L3:[0-9]+]]=, $pop[[L2]] -; BULK-MEM-NEXT: br_if 0, $pop[[L3]] -; BULK-MEM-NEXT: i64.const $push[[L4:[0-9]+]]=, 12 -; BULK-MEM-NEXT: i64.add $push[[L5:[0-9]+]]=, $1, $pop[[L4]] -; BULK-MEM-NEXT: i64.const $push[[L6:[0-9]+]]=, 100 -; BULK-MEM-NEXT: memory.fill 0, $pop[[L5]], $0, $pop[[L6]] -; BULK-MEM-NEXT: .LBB{{.*}}: -; BULK-MEM-NEXT: end_block +; BULK-MEM-NEXT: i64.sub $push[[L2:[0-9]+]]=, $pop[[L1]], $pop[[L0]] +; BULK-MEM-NEXT: i64.const $push[[L3:[0-9]+]]=, 12 +; BULK-MEM-NEXT: i64.add $push[[L4:[0-9]+]]=, $pop[[L2]], $pop[[L3]] +; BULK-MEM-NEXT: i64.const $push[[L5:[0-9]+]]=, 100 +; BULK-MEM-NEXT: memory.fill 0, $pop[[L4]], $0, $pop[[L5]] ; BULK-MEM-NEXT: return define void @memset_alloca(i8 %val) { %a = alloca [100 x i8] diff --git a/llvm/test/CodeGen/X86/2007-08-09-IllegalX86-64Asm.ll b/llvm/test/CodeGen/X86/2007-08-09-IllegalX86-64Asm.ll index 28b4541..7bdc4e1 100644 --- a/llvm/test/CodeGen/X86/2007-08-09-IllegalX86-64Asm.ll +++ b/llvm/test/CodeGen/X86/2007-08-09-IllegalX86-64Asm.ll @@ -44,7 +44,7 @@ define ptr @ubyte_divmod(ptr %a, ptr %b) { ; CHECK-NEXT: leaq {{[0-9]+}}(%rsp), %rsi ; CHECK-NEXT: callq __ubyte_convert_to_ctype ; CHECK-NEXT: testl %eax, %eax -; CHECK-NEXT: js LBB0_6 +; CHECK-NEXT: js LBB0_4 ; CHECK-NEXT: ## %bb.1: ## %cond_next.i ; CHECK-NEXT: leaq {{[0-9]+}}(%rsp), %rsi ; CHECK-NEXT: movq %rbx, %rdi @@ -53,84 +53,81 @@ define ptr @ubyte_divmod(ptr %a, ptr %b) { ; CHECK-NEXT: sarl $31, %ecx ; CHECK-NEXT: andl %eax, %ecx ; CHECK-NEXT: cmpl $-2, %ecx -; CHECK-NEXT: je LBB0_10 +; CHECK-NEXT: je LBB0_8 ; CHECK-NEXT: ## %bb.2: ## %cond_next.i ; CHECK-NEXT: cmpl $-1, %ecx -; CHECK-NEXT: jne LBB0_3 -; CHECK-NEXT: LBB0_8: ## %bb4 +; CHECK-NEXT: jne LBB0_6 +; CHECK-NEXT: LBB0_3: ## %bb4 ; CHECK-NEXT: movq _PyArray_API@GOTPCREL(%rip), %rax ; CHECK-NEXT: movq (%rax), %rax ; CHECK-NEXT: movq 16(%rax), %rax -; CHECK-NEXT: jmp LBB0_9 -; CHECK-NEXT: LBB0_6: ## %_ubyte_convert2_to_ctypes.exit +; CHECK-NEXT: jmp LBB0_10 +; CHECK-NEXT: LBB0_4: ## %_ubyte_convert2_to_ctypes.exit ; CHECK-NEXT: cmpl $-2, %eax -; CHECK-NEXT: je LBB0_10 -; CHECK-NEXT: ## %bb.7: ## %_ubyte_convert2_to_ctypes.exit -; CHECK-NEXT: cmpl $-1, %eax ; CHECK-NEXT: je LBB0_8 -; CHECK-NEXT: LBB0_3: ## %bb35 +; CHECK-NEXT: ## %bb.5: ## %_ubyte_convert2_to_ctypes.exit +; CHECK-NEXT: cmpl $-1, %eax +; CHECK-NEXT: je LBB0_3 +; CHECK-NEXT: LBB0_6: ## %bb35 ; CHECK-NEXT: movq _PyUFunc_API@GOTPCREL(%rip), %r14 ; CHECK-NEXT: movq (%r14), %rax ; CHECK-NEXT: callq *216(%rax) ; CHECK-NEXT: movzbl {{[0-9]+}}(%rsp), %edx ; CHECK-NEXT: testb %dl, %dl -; CHECK-NEXT: je LBB0_4 -; CHECK-NEXT: ## %bb.12: ## %cond_false.i -; CHECK-NEXT: setne %dil +; CHECK-NEXT: je LBB0_11 +; CHECK-NEXT: ## %bb.7: ## %cond_false.i ; CHECK-NEXT: movzbl {{[0-9]+}}(%rsp), %esi ; CHECK-NEXT: movzbl %sil, %ecx ; CHECK-NEXT: movl %ecx, %eax ; CHECK-NEXT: divb %dl ; CHECK-NEXT: movl %eax, %r15d ; CHECK-NEXT: testb %cl, %cl -; CHECK-NEXT: setne %al -; CHECK-NEXT: testb %dil, %al -; CHECK-NEXT: jne LBB0_5 -; CHECK-NEXT: LBB0_13: ## %cond_true.i200 -; CHECK-NEXT: testb %dl, %dl -; CHECK-NEXT: jne LBB0_15 -; CHECK-NEXT: ## %bb.14: ## %cond_true14.i -; CHECK-NEXT: movl $4, %edi -; CHECK-NEXT: callq _feraiseexcept -; CHECK-NEXT: LBB0_15: ## %ubyte_ctype_remainder.exit -; CHECK-NEXT: xorl %ebx, %ebx -; CHECK-NEXT: jmp LBB0_16 -; CHECK-NEXT: LBB0_10: ## %bb17 +; CHECK-NEXT: jne LBB0_12 +; CHECK-NEXT: jmp LBB0_14 +; CHECK-NEXT: LBB0_8: ## %bb17 ; CHECK-NEXT: callq _PyErr_Occurred ; CHECK-NEXT: testq %rax, %rax -; CHECK-NEXT: jne LBB0_23 -; CHECK-NEXT: ## %bb.11: ## %cond_next +; CHECK-NEXT: jne LBB0_27 +; CHECK-NEXT: ## %bb.9: ## %cond_next ; CHECK-NEXT: movq _PyArray_API@GOTPCREL(%rip), %rax ; CHECK-NEXT: movq (%rax), %rax ; CHECK-NEXT: movq 80(%rax), %rax -; CHECK-NEXT: LBB0_9: ## %bb4 +; CHECK-NEXT: LBB0_10: ## %bb4 ; CHECK-NEXT: movq 96(%rax), %rax ; CHECK-NEXT: movq %r14, %rdi ; CHECK-NEXT: movq %rbx, %rsi ; CHECK-NEXT: callq *40(%rax) -; CHECK-NEXT: jmp LBB0_24 -; CHECK-NEXT: LBB0_4: ## %cond_true.i +; CHECK-NEXT: jmp LBB0_28 +; CHECK-NEXT: LBB0_11: ## %cond_true.i ; CHECK-NEXT: movl $4, %edi ; CHECK-NEXT: callq _feraiseexcept ; CHECK-NEXT: movzbl {{[0-9]+}}(%rsp), %edx ; CHECK-NEXT: movzbl {{[0-9]+}}(%rsp), %esi +; CHECK-NEXT: xorl %r15d, %r15d ; CHECK-NEXT: testb %sil, %sil -; CHECK-NEXT: sete %al +; CHECK-NEXT: je LBB0_14 +; CHECK-NEXT: LBB0_12: ## %cond_false.i ; CHECK-NEXT: testb %dl, %dl -; CHECK-NEXT: sete %cl -; CHECK-NEXT: xorl %r15d, %r15d -; CHECK-NEXT: orb %al, %cl -; CHECK-NEXT: jne LBB0_13 -; CHECK-NEXT: LBB0_5: ## %cond_next17.i +; CHECK-NEXT: je LBB0_14 +; CHECK-NEXT: ## %bb.13: ## %cond_next17.i ; CHECK-NEXT: movzbl %sil, %eax ; CHECK-NEXT: divb %dl ; CHECK-NEXT: movzbl %ah, %ebx -; CHECK-NEXT: LBB0_16: ## %ubyte_ctype_remainder.exit +; CHECK-NEXT: jmp LBB0_18 +; CHECK-NEXT: LBB0_14: ## %cond_true.i200 +; CHECK-NEXT: testb %dl, %dl +; CHECK-NEXT: jne LBB0_17 +; CHECK-NEXT: ## %bb.16: ## %cond_true14.i +; CHECK-NEXT: movl $4, %edi +; CHECK-NEXT: callq _feraiseexcept +; CHECK-NEXT: LBB0_17: ## %ubyte_ctype_remainder.exit +; CHECK-NEXT: xorl %ebx, %ebx +; CHECK-NEXT: LBB0_18: ## %ubyte_ctype_remainder.exit ; CHECK-NEXT: movq (%r14), %rax ; CHECK-NEXT: callq *224(%rax) ; CHECK-NEXT: testl %eax, %eax -; CHECK-NEXT: je LBB0_19 -; CHECK-NEXT: ## %bb.17: ## %cond_true61 +; CHECK-NEXT: je LBB0_21 +; CHECK-NEXT: ## %bb.19: ## %cond_true61 ; CHECK-NEXT: movl %eax, %ebp ; CHECK-NEXT: movq (%r14), %rax ; CHECK-NEXT: movq _.str5@GOTPCREL(%rip), %rdi @@ -139,8 +136,8 @@ define ptr @ubyte_divmod(ptr %a, ptr %b) { ; CHECK-NEXT: leaq {{[0-9]+}}(%rsp), %rcx ; CHECK-NEXT: callq *200(%rax) ; CHECK-NEXT: testl %eax, %eax -; CHECK-NEXT: js LBB0_23 -; CHECK-NEXT: ## %bb.18: ## %cond_next73 +; CHECK-NEXT: js LBB0_27 +; CHECK-NEXT: ## %bb.20: ## %cond_next73 ; CHECK-NEXT: movl $1, {{[0-9]+}}(%rsp) ; CHECK-NEXT: movq (%r14), %rax ; CHECK-NEXT: movq {{[0-9]+}}(%rsp), %rsi @@ -149,13 +146,13 @@ define ptr @ubyte_divmod(ptr %a, ptr %b) { ; CHECK-NEXT: movl %ebp, %edx ; CHECK-NEXT: callq *232(%rax) ; CHECK-NEXT: testl %eax, %eax -; CHECK-NEXT: jne LBB0_23 -; CHECK-NEXT: LBB0_19: ## %cond_next89 +; CHECK-NEXT: jne LBB0_27 +; CHECK-NEXT: LBB0_21: ## %cond_next89 ; CHECK-NEXT: movl $2, %edi ; CHECK-NEXT: callq _PyTuple_New ; CHECK-NEXT: testq %rax, %rax -; CHECK-NEXT: je LBB0_23 -; CHECK-NEXT: ## %bb.20: ## %cond_next97 +; CHECK-NEXT: je LBB0_27 +; CHECK-NEXT: ## %bb.22: ## %cond_next97 ; CHECK-NEXT: movq %rax, %r14 ; CHECK-NEXT: movq _PyArray_API@GOTPCREL(%rip), %r12 ; CHECK-NEXT: movq (%r12), %rax @@ -163,8 +160,8 @@ define ptr @ubyte_divmod(ptr %a, ptr %b) { ; CHECK-NEXT: xorl %esi, %esi ; CHECK-NEXT: callq *304(%rdi) ; CHECK-NEXT: testq %rax, %rax -; CHECK-NEXT: je LBB0_21 -; CHECK-NEXT: ## %bb.25: ## %cond_next135 +; CHECK-NEXT: je LBB0_25 +; CHECK-NEXT: ## %bb.23: ## %cond_next135 ; CHECK-NEXT: movb %r15b, 16(%rax) ; CHECK-NEXT: movq %rax, 24(%r14) ; CHECK-NEXT: movq (%r12), %rax @@ -172,22 +169,22 @@ define ptr @ubyte_divmod(ptr %a, ptr %b) { ; CHECK-NEXT: xorl %esi, %esi ; CHECK-NEXT: callq *304(%rdi) ; CHECK-NEXT: testq %rax, %rax -; CHECK-NEXT: je LBB0_21 -; CHECK-NEXT: ## %bb.26: ## %cond_next182 +; CHECK-NEXT: je LBB0_25 +; CHECK-NEXT: ## %bb.24: ## %cond_next182 ; CHECK-NEXT: movb %bl, 16(%rax) ; CHECK-NEXT: movq %rax, 32(%r14) ; CHECK-NEXT: movq %r14, %rax -; CHECK-NEXT: jmp LBB0_24 -; CHECK-NEXT: LBB0_21: ## %cond_true113 +; CHECK-NEXT: jmp LBB0_28 +; CHECK-NEXT: LBB0_25: ## %cond_true113 ; CHECK-NEXT: decq (%r14) -; CHECK-NEXT: jne LBB0_23 -; CHECK-NEXT: ## %bb.22: ## %cond_true126 +; CHECK-NEXT: jne LBB0_27 +; CHECK-NEXT: ## %bb.26: ## %cond_true126 ; CHECK-NEXT: movq 8(%r14), %rax ; CHECK-NEXT: movq %r14, %rdi ; CHECK-NEXT: callq *48(%rax) -; CHECK-NEXT: LBB0_23: ## %UnifiedReturnBlock +; CHECK-NEXT: LBB0_27: ## %UnifiedReturnBlock ; CHECK-NEXT: xorl %eax, %eax -; CHECK-NEXT: LBB0_24: ## %UnifiedReturnBlock +; CHECK-NEXT: LBB0_28: ## %UnifiedReturnBlock ; CHECK-NEXT: addq $32, %rsp ; CHECK-NEXT: popq %rbx ; CHECK-NEXT: popq %r12 diff --git a/llvm/test/CodeGen/X86/absolute-symbol-kernel-code-model.ll b/llvm/test/CodeGen/X86/absolute-symbol-kernel-code-model.ll new file mode 100644 index 0000000..ce7024d --- /dev/null +++ b/llvm/test/CodeGen/X86/absolute-symbol-kernel-code-model.ll @@ -0,0 +1,34 @@ +; RUN: llc --code-model=kernel < %s -asm-verbose=0 | FileCheck %s + +target triple = "x86_64-unknown-linux-gnu" + +; CHECK-LABEL: func_no_abs_sym +define i64 @func_no_abs_sym() nounwind { + ; CHECK: movq $no_abs_sym, %rax + %1 = ptrtoint ptr @no_abs_sym to i64 + ret i64 %1 +} + +; CHECK-LABEL: func_abs_sym +define i64 @func_abs_sym() nounwind { + ; CHECK: movabsq $abs_sym, %rax + %1 = ptrtoint ptr @abs_sym to i64 + ret i64 %1 +} + +; CHECK-LABEL: func_abs_sym_in_range +define i64 @func_abs_sym_in_range() nounwind { + ;; The absolute_symbol range fits in 32 bits but we still use movabs + ;; since there's no benefit to using the sign extending instruction + ;; with absolute symbols. + ; CHECK: movabsq $abs_sym_in_range, %rax + %1 = ptrtoint ptr @abs_sym_in_range to i64 + ret i64 %1 +} + +@no_abs_sym = external hidden global [0 x i8] +@abs_sym = external hidden global [0 x i8], !absolute_symbol !0 +@abs_sym_in_range = external hidden global [0 x i8], !absolute_symbol !1 + +!0 = !{i64 -1, i64 -1} ;; Full range +!1 = !{i64 -2147483648, i64 2147483648} ;; In range diff --git a/llvm/test/CodeGen/X86/apx/cf.ll b/llvm/test/CodeGen/X86/apx/cf.ll index b2651e9..de9caa5 100644 --- a/llvm/test/CodeGen/X86/apx/cf.ll +++ b/llvm/test/CodeGen/X86/apx/cf.ll @@ -230,6 +230,24 @@ entry: ret void } +define void @and_cond(i32 %a, i1 %b) { +; CHECK-LABEL: and_cond: +; CHECK: # %bb.0: +; CHECK-NEXT: testl %edi, %edi +; CHECK-NEXT: setg %al +; CHECK-NEXT: notb %sil +; CHECK-NEXT: xorl %ecx, %ecx +; CHECK-NEXT: testb %al, %sil +; CHECK-NEXT: cfcmovnel %ecx, 0 +; CHECK-NEXT: retq + %is_pos = icmp sgt i32 %a, 0 + %not_b = xor i1 %b, true + %cond = and i1 %not_b, %is_pos + %mask = insertelement <1 x i1> zeroinitializer, i1 %cond, i64 0 + call void @llvm.masked.store.v1i32.p0(<1 x i32> zeroinitializer, ptr null, i32 1, <1 x i1> %mask) + ret void +} + define i64 @redundant_test(i64 %num, ptr %p1, i64 %in) { ; CHECK-LABEL: redundant_test: ; CHECK: # %bb.0: diff --git a/llvm/test/CodeGen/X86/call-graph-section-addrtaken.ll b/llvm/test/CodeGen/X86/call-graph-section-addrtaken.ll index 2aea9c1..632d90d 100644 --- a/llvm/test/CodeGen/X86/call-graph-section-addrtaken.ll +++ b/llvm/test/CodeGen/X86/call-graph-section-addrtaken.ll @@ -27,7 +27,7 @@ entry: !1 = !{i64 0, !"_ZTSFivE.generalized"} !2 = !{i64 0, !"_ZTSFviE.generalized"} -; CHECK: .section .callgraph,"o",@progbits,.text +; CHECK: .section .llvm.callgraph,"o",@progbits,.text ;; Version ; CHECK-NEXT: .byte 0 ;; Flags -- Potential indirect target so LSB is set to 1. Other bits are 0. diff --git a/llvm/test/CodeGen/X86/call-graph-section-assembly.ll b/llvm/test/CodeGen/X86/call-graph-section-assembly.ll index 1aabf66..ed6849a 100644 --- a/llvm/test/CodeGen/X86/call-graph-section-assembly.ll +++ b/llvm/test/CodeGen/X86/call-graph-section-assembly.ll @@ -1,8 +1,8 @@ ;; Test if temporary labels are generated for each indirect callsite. -;; Test if the .callgraph section contains the MD5 hash of callees' type (type id) +;; Test if the .llvm.callgraph section contains the MD5 hash of callees' type (type id) ;; is correctly paired with its corresponding temporary label generated for indirect ;; call sites annotated with !callee_type metadata. -;; Test if the .callgraph section contains unique direct callees. +;; Test if the .llvm.callgraph section contains unique direct callees. ; RUN: llc -mtriple=x86_64-unknown-linux --call-graph-section -o - < %s | FileCheck %s @@ -36,7 +36,7 @@ entry: !4 = !{!5} !5 = !{i64 0, !"_ZTSFPvS_E.generalized"} -; CHECK: .section .callgraph,"o",@progbits,.text +; CHECK: .section .llvm.callgraph,"o",@progbits,.text ;; Version ; CHECK-NEXT: .byte 0 ;; Flags diff --git a/llvm/test/CodeGen/X86/call-graph-section-tailcall.ll b/llvm/test/CodeGen/X86/call-graph-section-tailcall.ll index 34dc5b8..49cc335 100644 --- a/llvm/test/CodeGen/X86/call-graph-section-tailcall.ll +++ b/llvm/test/CodeGen/X86/call-graph-section-tailcall.ll @@ -1,7 +1,10 @@ -;; Tests that we store the type identifiers in .callgraph section of the object file for tailcalls. +;; Tests that we store the type identifiers in .llvm.callgraph section of the object file for tailcalls. + +; REQUIRES: x86-registered-target +; REQUIRES: arm-registered-target ; RUN: llc -mtriple=x86_64-unknown-linux --call-graph-section -filetype=obj -o - < %s | \ -; RUN: llvm-readelf -x .callgraph - | FileCheck %s +; RUN: llvm-readelf -x .llvm.callgraph - | FileCheck %s define i32 @check_tailcall(ptr %func, i8 %x) !type !0 { entry: @@ -27,7 +30,7 @@ declare !type !2 i32 @bar(i8 signext) !2 = !{i64 0, !"_ZTSFicE.generalized"} !3 = !{i64 0, !"_ZTSFiiE.generalized"} -; CHECK: Hex dump of section '.callgraph': +; CHECK: Hex dump of section '.llvm.callgraph': ; CHECK-NEXT: 0x00000000 00050000 00000000 00008e19 0b7f3326 ; CHECK-NEXT: 0x00000010 e3000154 86bc5981 4b8e3000 05000000 ;; Verify that the type id 0x308e4b8159bc8654 is in section. diff --git a/llvm/test/CodeGen/X86/call-graph-section.ll b/llvm/test/CodeGen/X86/call-graph-section.ll index c144a24..8a1c6ca 100644 --- a/llvm/test/CodeGen/X86/call-graph-section.ll +++ b/llvm/test/CodeGen/X86/call-graph-section.ll @@ -1,7 +1,10 @@ -;; Tests that we store the type identifiers in .callgraph section of the object file. +;; Tests that we store the type identifiers in .llvm.callgraph section of the object file. + +; REQUIRES: x86-registered-target +; REQUIRES: arm-registered-target ; RUN: llc -mtriple=x86_64-unknown-linux --call-graph-section -filetype=obj -o - < %s | \ -; RUN: llvm-readelf -x .callgraph - | FileCheck %s +; RUN: llvm-readelf -x .llvm.callgraph - | FileCheck %s declare !type !0 void @foo() @@ -31,7 +34,7 @@ entry: ;; Make sure following type IDs are in call graph section ;; 0x5eecb3e2444f731f, 0x814b8e305486bc59, 0xf897fd777ade6814 -; CHECK: Hex dump of section '.callgraph': +; CHECK: Hex dump of section '.llvm.callgraph': ; CHECK-NEXT: 0x00000000 00050000 00000000 00000000 00000000 ; CHECK-NEXT: 0x00000010 00000324 44f731f5 eecb3e54 86bc5981 ; CHECK-NEXT: 0x00000020 4b8e307a de6814f8 97fd77 diff --git a/llvm/test/CodeGen/X86/cpus-intel.ll b/llvm/test/CodeGen/X86/cpus-intel.ll index 40c38c2..71253c8 100644 --- a/llvm/test/CodeGen/X86/cpus-intel.ll +++ b/llvm/test/CodeGen/X86/cpus-intel.ll @@ -38,6 +38,7 @@ ; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=lunarlake 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty ; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=gracemont 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty ; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=pantherlake 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty +; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=wildcatlake 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty ; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=clearwaterforest 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty ; RUN: llc < %s -o /dev/null -mtriple=i686-unknown-unknown -mcpu=diamondrapids 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty @@ -104,6 +105,7 @@ ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=lunarlake 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=gracemont 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=pantherlake 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty +; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=wildcatlake 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=clearwaterforest 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty ; RUN: llc < %s -o /dev/null -mtriple=x86_64-unknown-unknown -mcpu=diamondrapids 2>&1 | FileCheck %s --check-prefix=CHECK-NO-ERROR --allow-empty diff --git a/llvm/test/CodeGen/X86/global-variable-partition-with-dap.ll b/llvm/test/CodeGen/X86/global-variable-partition-with-dap.ll index f3950b7..b2b0a6d 100644 --- a/llvm/test/CodeGen/X86/global-variable-partition-with-dap.ll +++ b/llvm/test/CodeGen/X86/global-variable-partition-with-dap.ll @@ -1,17 +1,101 @@ target datalayout = "e-m:e-p270:32:32-p271:32:32-p272:64:64-i64:64-i128:128-f80:128-n8:16:32:64-S128" target triple = "x86_64-unknown-linux-gnu" -;; A minimal test case. Subsequent PRs will expand on this test case -;; (e.g., with more functions, variables and profiles) and test the hotness -;; reconcillation implementation. +;; Requires asserts for -debug-only. +; REQUIRES: asserts + +; RUN: rm -rf %t && split-file %s %t && cd %t + +; RUN: llc -mtriple=x86_64-unknown-linux-gnu -relocation-model=pic \ +; RUN: -partition-static-data-sections=true \ +; RUN: -debug-only=static-data-profile-info \ +; RUN: -data-sections=true -unique-section-names=false \ +; RUN: input-with-data-access-prof-on.ll -o - 2>&1 | FileCheck %s --check-prefixes=LOG,IR + ; RUN: llc -mtriple=x86_64-unknown-linux-gnu -relocation-model=pic \ ; RUN: -partition-static-data-sections=true \ +; RUN: -debug-only=static-data-profile-info \ ; RUN: -data-sections=true -unique-section-names=false \ -; RUN: %s -o - 2>&1 | FileCheck %s --check-prefix=IR +; RUN: input-with-data-access-prof-off.ll -o - 2>&1 | FileCheck %s --check-prefixes=OFF + +; LOG: hot_bss has section prefix hot, the max from data access profiles as hot and PGO counters as hot +; LOG: data_unknown_hotness has section prefix <empty>, the max from data access profiles as <empty> and PGO counters as unlikely +; LOG: external_relro_array has section prefix unlikely, solely from data access profiles + +; IR: .type hot_bss,@object +; IR-NEXT: .section .bss.hot.,"aw" +; IR: .type data_unknown_hotness,@object +; IR-NEXT: .section .data,"aw" +; IR: .type external_relro_array,@object +; IR-NEXT: .section .data.rel.ro.unlikely.,"aw" + + +; OFF: .type hot_bss,@object +; OFF-NEXT: .section .bss.hot.,"aw" +; OFF: .type data_unknown_hotness,@object +; OFF-NEXT: .section .data.unlikely.,"aw" +;; Global variable section prefix metadata is not used when +;; module flag `EnableDataAccessProf` is 0, and @external_relro_array has +;; external linkage, so analysis based on PGO counters doesn't apply. +; OFF: .type external_relro_array,@object # @external_relro_array +; OFF-NEXT: .section .data.rel.ro,"aw" + +;--- input-with-data-access-prof-on.ll +; Internal vars +@hot_bss = internal global i32 0, !section_prefix !17 +@data_unknown_hotness = internal global i32 1 +; External vars +@external_relro_array = constant [2 x ptr] [ptr @hot_bss, ptr @data_unknown_hotness], !section_prefix !18 + +define void @cold_func() !prof !15 { + %9 = load i32, ptr @data_unknown_hotness + %11 = call i32 (...) @func_taking_arbitrary_param(i32 %9) + ret void +} + +define void @hot_func() !prof !14 { + %9 = load i32, ptr @hot_bss + %11 = call i32 (...) @func_taking_arbitrary_param(i32 %9) + ret void +} + +declare i32 @func_taking_arbitrary_param(...) -; IR: .section .bss.hot.,"aw" +!llvm.module.flags = !{!0, !1} +!0 = !{i32 2, !"EnableDataAccessProf", i32 1} +!1 = !{i32 1, !"ProfileSummary", !2} +!2 = !{!3, !4, !5, !6, !7, !8, !9, !10} +!3 = !{!"ProfileFormat", !"InstrProf"} +!4 = !{!"TotalCount", i64 1460183} +!5 = !{!"MaxCount", i64 849024} +!6 = !{!"MaxInternalCount", i64 32769} +!7 = !{!"MaxFunctionCount", i64 849024} +!8 = !{!"NumCounts", i64 23627} +!9 = !{!"NumFunctions", i64 3271} +!10 = !{!"DetailedSummary", !11} +!11 = !{!12, !13} +!12 = !{i32 990000, i64 166, i32 73} +!13 = !{i32 999999, i64 3, i32 1443} +!14 = !{!"function_entry_count", i64 100000} +!15 = !{!"function_entry_count", i64 1} +!16 = !{!"branch_weights", i32 1, i32 99999} +!17 = !{!"section_prefix", !"hot"} +!18 = !{!"section_prefix", !"unlikely"} + +;--- input-with-data-access-prof-off.ll +; Same as file above except that module flag `EnableDataAccessProf` has value 0. +; Internal vars @hot_bss = internal global i32 0, !section_prefix !17 +@data_unknown_hotness = internal global i32 1 +; External vars +@external_relro_array = constant [2 x ptr] [ptr @hot_bss, ptr @data_unknown_hotness], !section_prefix !18 + +define void @cold_func() !prof !15 { + %9 = load i32, ptr @data_unknown_hotness + %11 = call i32 (...) @func_taking_arbitrary_param(i32 %9) + ret void +} define void @hot_func() !prof !14 { %9 = load i32, ptr @hot_bss @@ -21,8 +105,9 @@ define void @hot_func() !prof !14 { declare i32 @func_taking_arbitrary_param(...) -!llvm.module.flags = !{!1} +!llvm.module.flags = !{!0, !1} +!0 = !{i32 2, !"EnableDataAccessProf", i32 0} !1 = !{i32 1, !"ProfileSummary", !2} !2 = !{!3, !4, !5, !6, !7, !8, !9, !10} !3 = !{!"ProfileFormat", !"InstrProf"} @@ -40,3 +125,4 @@ declare i32 @func_taking_arbitrary_param(...) !15 = !{!"function_entry_count", i64 1} !16 = !{!"branch_weights", i32 1, i32 99999} !17 = !{!"section_prefix", !"hot"} +!18 = !{!"section_prefix", !"unlikely"} diff --git a/llvm/test/CodeGen/X86/isel-fpclass.ll b/llvm/test/CodeGen/X86/isel-fpclass.ll index df04b67..c2b7068 100644 --- a/llvm/test/CodeGen/X86/isel-fpclass.ll +++ b/llvm/test/CodeGen/X86/isel-fpclass.ll @@ -1,10 +1,10 @@ ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5 -; RUN: llc < %s -mtriple=i686-linux | FileCheck %s -check-prefixes=X86 +; RUN: llc < %s -mtriple=i686-linux | FileCheck %s -check-prefixes=X86,X86-SDAGISEL ; RUN: llc < %s -mtriple=x86_64-linux | FileCheck %s -check-prefixes=X64,X64-SDAGISEL ; RUN: llc < %s -mtriple=i686-linux -fast-isel -fast-isel-abort=1 | FileCheck %s -check-prefixes=X86-FASTISEL ; RUN: llc < %s -mtriple=x86_64-linux -fast-isel -fast-isel-abort=1 | FileCheck %s -check-prefixes=X64,X64-FASTISEL -; RUN: llc < %s -mtriple=i686-linux -global-isel -global-isel-abort=2 | FileCheck %s -check-prefixes=X86 -; RUN: llc < %s -mtriple=x86_64-linux -global-isel -global-isel-abort=2 | FileCheck %s -check-prefixes=X64,X64-GISEL +; RUN: llc < %s -mtriple=i686-linux -global-isel -global-isel-abort=1 | FileCheck %s -check-prefixes=X86,X86-GISEL +; RUN: llc < %s -mtriple=x86_64-linux -global-isel -global-isel-abort=1 | FileCheck %s -check-prefixes=X64-GISEL define i1 @isnone_f(float %x) nounwind { ; X86-LABEL: isnone_f: @@ -23,6 +23,11 @@ define i1 @isnone_f(float %x) nounwind { ; X86-FASTISEL-NEXT: fstp %st(0) ; X86-FASTISEL-NEXT: xorl %eax, %eax ; X86-FASTISEL-NEXT: retl +; +; X64-GISEL-LABEL: isnone_f: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: xorl %eax, %eax +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 0) ret i1 %0 @@ -45,22 +50,27 @@ define i1 @isany_f(float %x) nounwind { ; X86-FASTISEL-NEXT: fstp %st(0) ; X86-FASTISEL-NEXT: movb $1, %al ; X86-FASTISEL-NEXT: retl +; +; X64-GISEL-LABEL: isany_f: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: movb $1, %al +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 1023) ret i1 %0 } define i1 @issignaling_f(float %x) nounwind { -; X86-LABEL: issignaling_f: -; X86: # %bb.0: -; X86-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF -; X86-NEXT: andl {{[0-9]+}}(%esp), %eax -; X86-NEXT: cmpl $2143289344, %eax # imm = 0x7FC00000 -; X86-NEXT: setl %cl -; X86-NEXT: cmpl $2139095041, %eax # imm = 0x7F800001 -; X86-NEXT: setge %al -; X86-NEXT: andb %cl, %al -; X86-NEXT: retl +; X86-SDAGISEL-LABEL: issignaling_f: +; X86-SDAGISEL: # %bb.0: +; X86-SDAGISEL-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-SDAGISEL-NEXT: andl {{[0-9]+}}(%esp), %eax +; X86-SDAGISEL-NEXT: cmpl $2143289344, %eax # imm = 0x7FC00000 +; X86-SDAGISEL-NEXT: setl %cl +; X86-SDAGISEL-NEXT: cmpl $2139095041, %eax # imm = 0x7F800001 +; X86-SDAGISEL-NEXT: setge %al +; X86-SDAGISEL-NEXT: andb %cl, %al +; X86-SDAGISEL-NEXT: retl ; ; X64-LABEL: issignaling_f: ; X64: # %bb.0: @@ -87,18 +97,44 @@ define i1 @issignaling_f(float %x) nounwind { ; X86-FASTISEL-NEXT: andb %cl, %al ; X86-FASTISEL-NEXT: popl %ecx ; X86-FASTISEL-NEXT: retl +; +; X86-GISEL-LABEL: issignaling_f: +; X86-GISEL: # %bb.0: +; X86-GISEL-NEXT: movl {{[0-9]+}}(%esp), %eax +; X86-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-GISEL-NEXT: xorl %ecx, %ecx +; X86-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-GISEL-NEXT: seta %dl +; X86-GISEL-NEXT: cmpl $2143289344, %eax # imm = 0x7FC00000 +; X86-GISEL-NEXT: setb %al +; X86-GISEL-NEXT: andb %dl, %al +; X86-GISEL-NEXT: orb %cl, %al +; X86-GISEL-NEXT: retl +; +; X64-GISEL-LABEL: issignaling_f: +; X64-GISEL: # %bb.0: +; X64-GISEL-NEXT: movd %xmm0, %eax +; X64-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X64-GISEL-NEXT: xorl %ecx, %ecx +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: seta %dl +; X64-GISEL-NEXT: cmpl $2143289344, %eax # imm = 0x7FC00000 +; X64-GISEL-NEXT: setb %al +; X64-GISEL-NEXT: andb %dl, %al +; X64-GISEL-NEXT: orb %cl, %al +; X64-GISEL-NEXT: retq %a0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 1) ; "snan" ret i1 %a0 } define i1 @isquiet_f(float %x) nounwind { -; X86-LABEL: isquiet_f: -; X86: # %bb.0: # %entry -; X86-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF -; X86-NEXT: andl {{[0-9]+}}(%esp), %eax -; X86-NEXT: cmpl $2143289344, %eax # imm = 0x7FC00000 -; X86-NEXT: setge %al -; X86-NEXT: retl +; X86-SDAGISEL-LABEL: isquiet_f: +; X86-SDAGISEL: # %bb.0: # %entry +; X86-SDAGISEL-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-SDAGISEL-NEXT: andl {{[0-9]+}}(%esp), %eax +; X86-SDAGISEL-NEXT: cmpl $2143289344, %eax # imm = 0x7FC00000 +; X86-SDAGISEL-NEXT: setge %al +; X86-SDAGISEL-NEXT: retl ; ; X64-LABEL: isquiet_f: ; X64: # %bb.0: # %entry @@ -119,19 +155,39 @@ define i1 @issignaling_f(float %x) nounwind { ; X86-FASTISEL-NEXT: setge %al ; X86-FASTISEL-NEXT: popl %ecx ; X86-FASTISEL-NEXT: retl +; +; X86-GISEL-LABEL: isquiet_f: +; X86-GISEL: # %bb.0: # %entry +; X86-GISEL-NEXT: movl {{[0-9]+}}(%esp), %eax +; X86-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-GISEL-NEXT: xorl %ecx, %ecx +; X86-GISEL-NEXT: cmpl $2143289344, %eax # imm = 0x7FC00000 +; X86-GISEL-NEXT: setae %al +; X86-GISEL-NEXT: orb %cl, %al +; X86-GISEL-NEXT: retl +; +; X64-GISEL-LABEL: isquiet_f: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: movd %xmm0, %eax +; X64-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X64-GISEL-NEXT: xorl %ecx, %ecx +; X64-GISEL-NEXT: cmpl $2143289344, %eax # imm = 0x7FC00000 +; X64-GISEL-NEXT: setae %al +; X64-GISEL-NEXT: orb %cl, %al +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 2) ; "qnan" ret i1 %0 } define i1 @not_isquiet_f(float %x) nounwind { -; X86-LABEL: not_isquiet_f: -; X86: # %bb.0: # %entry -; X86-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF -; X86-NEXT: andl {{[0-9]+}}(%esp), %eax -; X86-NEXT: cmpl $2143289344, %eax # imm = 0x7FC00000 -; X86-NEXT: setl %al -; X86-NEXT: retl +; X86-SDAGISEL-LABEL: not_isquiet_f: +; X86-SDAGISEL: # %bb.0: # %entry +; X86-SDAGISEL-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-SDAGISEL-NEXT: andl {{[0-9]+}}(%esp), %eax +; X86-SDAGISEL-NEXT: cmpl $2143289344, %eax # imm = 0x7FC00000 +; X86-SDAGISEL-NEXT: setl %al +; X86-SDAGISEL-NEXT: retl ; ; X64-LABEL: not_isquiet_f: ; X64: # %bb.0: # %entry @@ -152,19 +208,57 @@ define i1 @not_isquiet_f(float %x) nounwind { ; X86-FASTISEL-NEXT: setl %al ; X86-FASTISEL-NEXT: popl %ecx ; X86-FASTISEL-NEXT: retl +; +; X86-GISEL-LABEL: not_isquiet_f: +; X86-GISEL: # %bb.0: # %entry +; X86-GISEL-NEXT: movl {{[0-9]+}}(%esp), %eax +; X86-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-GISEL-NEXT: xorl %ecx, %ecx +; X86-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-GISEL-NEXT: setb %dl +; X86-GISEL-NEXT: orb %cl, %dl +; X86-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-GISEL-NEXT: sete %cl +; X86-GISEL-NEXT: orb %dl, %cl +; X86-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-GISEL-NEXT: seta %dl +; X86-GISEL-NEXT: cmpl $2143289344, %eax # imm = 0x7FC00000 +; X86-GISEL-NEXT: setb %al +; X86-GISEL-NEXT: andb %dl, %al +; X86-GISEL-NEXT: orb %cl, %al +; X86-GISEL-NEXT: retl +; +; X64-GISEL-LABEL: not_isquiet_f: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: movd %xmm0, %eax +; X64-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X64-GISEL-NEXT: xorl %ecx, %ecx +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: setb %dl +; X64-GISEL-NEXT: orb %cl, %dl +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: sete %cl +; X64-GISEL-NEXT: orb %dl, %cl +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: seta %dl +; X64-GISEL-NEXT: cmpl $2143289344, %eax # imm = 0x7FC00000 +; X64-GISEL-NEXT: setb %al +; X64-GISEL-NEXT: andb %dl, %al +; X64-GISEL-NEXT: orb %cl, %al +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 1021) ; ~"qnan" ret i1 %0 } define i1 @isinf_f(float %x) nounwind { -; X86-LABEL: isinf_f: -; X86: # %bb.0: # %entry -; X86-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF -; X86-NEXT: andl {{[0-9]+}}(%esp), %eax -; X86-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 -; X86-NEXT: sete %al -; X86-NEXT: retl +; X86-SDAGISEL-LABEL: isinf_f: +; X86-SDAGISEL: # %bb.0: # %entry +; X86-SDAGISEL-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-SDAGISEL-NEXT: andl {{[0-9]+}}(%esp), %eax +; X86-SDAGISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-SDAGISEL-NEXT: sete %al +; X86-SDAGISEL-NEXT: retl ; ; X64-LABEL: isinf_f: ; X64: # %bb.0: # %entry @@ -185,19 +279,39 @@ define i1 @isinf_f(float %x) nounwind { ; X86-FASTISEL-NEXT: sete %al ; X86-FASTISEL-NEXT: popl %ecx ; X86-FASTISEL-NEXT: retl +; +; X86-GISEL-LABEL: isinf_f: +; X86-GISEL: # %bb.0: # %entry +; X86-GISEL-NEXT: movl {{[0-9]+}}(%esp), %eax +; X86-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-GISEL-NEXT: xorl %ecx, %ecx +; X86-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-GISEL-NEXT: sete %al +; X86-GISEL-NEXT: orb %cl, %al +; X86-GISEL-NEXT: retl +; +; X64-GISEL-LABEL: isinf_f: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: movd %xmm0, %eax +; X64-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X64-GISEL-NEXT: xorl %ecx, %ecx +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: sete %al +; X64-GISEL-NEXT: orb %cl, %al +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 516) ; 0x204 = "inf" ret i1 %0 } define i1 @not_isinf_f(float %x) nounwind { -; X86-LABEL: not_isinf_f: -; X86: # %bb.0: # %entry -; X86-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF -; X86-NEXT: andl {{[0-9]+}}(%esp), %eax -; X86-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 -; X86-NEXT: setne %al -; X86-NEXT: retl +; X86-SDAGISEL-LABEL: not_isinf_f: +; X86-SDAGISEL: # %bb.0: # %entry +; X86-SDAGISEL-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-SDAGISEL-NEXT: andl {{[0-9]+}}(%esp), %eax +; X86-SDAGISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-SDAGISEL-NEXT: setne %al +; X86-SDAGISEL-NEXT: retl ; ; X64-LABEL: not_isinf_f: ; X64: # %bb.0: # %entry @@ -218,17 +332,43 @@ define i1 @not_isinf_f(float %x) nounwind { ; X86-FASTISEL-NEXT: setne %al ; X86-FASTISEL-NEXT: popl %ecx ; X86-FASTISEL-NEXT: retl +; +; X86-GISEL-LABEL: not_isinf_f: +; X86-GISEL: # %bb.0: # %entry +; X86-GISEL-NEXT: movl {{[0-9]+}}(%esp), %eax +; X86-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-GISEL-NEXT: xorl %ecx, %ecx +; X86-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-GISEL-NEXT: setb %dl +; X86-GISEL-NEXT: orb %cl, %dl +; X86-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-GISEL-NEXT: seta %al +; X86-GISEL-NEXT: orb %dl, %al +; X86-GISEL-NEXT: retl +; +; X64-GISEL-LABEL: not_isinf_f: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: movd %xmm0, %eax +; X64-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X64-GISEL-NEXT: xorl %ecx, %ecx +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: setb %dl +; X64-GISEL-NEXT: orb %cl, %dl +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: seta %al +; X64-GISEL-NEXT: orb %dl, %al +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 507) ; ~0x204 = "~inf" ret i1 %0 } define i1 @is_plus_inf_f(float %x) nounwind { -; X86-LABEL: is_plus_inf_f: -; X86: # %bb.0: # %entry -; X86-NEXT: cmpl $2139095040, {{[0-9]+}}(%esp) # imm = 0x7F800000 -; X86-NEXT: sete %al -; X86-NEXT: retl +; X86-SDAGISEL-LABEL: is_plus_inf_f: +; X86-SDAGISEL: # %bb.0: # %entry +; X86-SDAGISEL-NEXT: cmpl $2139095040, {{[0-9]+}}(%esp) # imm = 0x7F800000 +; X86-SDAGISEL-NEXT: sete %al +; X86-SDAGISEL-NEXT: retl ; ; X64-LABEL: is_plus_inf_f: ; X64: # %bb.0: # %entry @@ -246,17 +386,34 @@ define i1 @is_plus_inf_f(float %x) nounwind { ; X86-FASTISEL-NEXT: sete %al ; X86-FASTISEL-NEXT: popl %ecx ; X86-FASTISEL-NEXT: retl +; +; X86-GISEL-LABEL: is_plus_inf_f: +; X86-GISEL: # %bb.0: # %entry +; X86-GISEL-NEXT: xorl %ecx, %ecx +; X86-GISEL-NEXT: cmpl $2139095040, {{[0-9]+}}(%esp) # imm = 0x7F800000 +; X86-GISEL-NEXT: sete %al +; X86-GISEL-NEXT: orb %cl, %al +; X86-GISEL-NEXT: retl +; +; X64-GISEL-LABEL: is_plus_inf_f: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: xorl %ecx, %ecx +; X64-GISEL-NEXT: movd %xmm0, %eax +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: sete %al +; X64-GISEL-NEXT: orb %cl, %al +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 512) ; 0x200 = "+inf" ret i1 %0 } define i1 @is_minus_inf_f(float %x) nounwind { -; X86-LABEL: is_minus_inf_f: -; X86: # %bb.0: # %entry -; X86-NEXT: cmpl $-8388608, {{[0-9]+}}(%esp) # imm = 0xFF800000 -; X86-NEXT: sete %al -; X86-NEXT: retl +; X86-SDAGISEL-LABEL: is_minus_inf_f: +; X86-SDAGISEL: # %bb.0: # %entry +; X86-SDAGISEL-NEXT: cmpl $-8388608, {{[0-9]+}}(%esp) # imm = 0xFF800000 +; X86-SDAGISEL-NEXT: sete %al +; X86-SDAGISEL-NEXT: retl ; ; X64-LABEL: is_minus_inf_f: ; X64: # %bb.0: # %entry @@ -274,17 +431,34 @@ define i1 @is_minus_inf_f(float %x) nounwind { ; X86-FASTISEL-NEXT: sete %al ; X86-FASTISEL-NEXT: popl %ecx ; X86-FASTISEL-NEXT: retl +; +; X86-GISEL-LABEL: is_minus_inf_f: +; X86-GISEL: # %bb.0: # %entry +; X86-GISEL-NEXT: xorl %ecx, %ecx +; X86-GISEL-NEXT: cmpl $-8388608, {{[0-9]+}}(%esp) # imm = 0xFF800000 +; X86-GISEL-NEXT: sete %al +; X86-GISEL-NEXT: orb %cl, %al +; X86-GISEL-NEXT: retl +; +; X64-GISEL-LABEL: is_minus_inf_f: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: xorl %ecx, %ecx +; X64-GISEL-NEXT: movd %xmm0, %eax +; X64-GISEL-NEXT: cmpl $-8388608, %eax # imm = 0xFF800000 +; X64-GISEL-NEXT: sete %al +; X64-GISEL-NEXT: orb %cl, %al +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 4) ; "-inf" ret i1 %0 } define i1 @not_is_minus_inf_f(float %x) nounwind { -; X86-LABEL: not_is_minus_inf_f: -; X86: # %bb.0: # %entry -; X86-NEXT: cmpl $-8388608, {{[0-9]+}}(%esp) # imm = 0xFF800000 -; X86-NEXT: setne %al -; X86-NEXT: retl +; X86-SDAGISEL-LABEL: not_is_minus_inf_f: +; X86-SDAGISEL: # %bb.0: # %entry +; X86-SDAGISEL-NEXT: cmpl $-8388608, {{[0-9]+}}(%esp) # imm = 0xFF800000 +; X86-SDAGISEL-NEXT: setne %al +; X86-SDAGISEL-NEXT: retl ; ; X64-LABEL: not_is_minus_inf_f: ; X64: # %bb.0: # %entry @@ -302,19 +476,55 @@ define i1 @not_is_minus_inf_f(float %x) nounwind { ; X86-FASTISEL-NEXT: setne %al ; X86-FASTISEL-NEXT: popl %ecx ; X86-FASTISEL-NEXT: retl +; +; X86-GISEL-LABEL: not_is_minus_inf_f: +; X86-GISEL: # %bb.0: # %entry +; X86-GISEL-NEXT: pushl %ebx +; X86-GISEL-NEXT: movl {{[0-9]+}}(%esp), %eax +; X86-GISEL-NEXT: movl %eax, %ecx +; X86-GISEL-NEXT: andl $2147483647, %ecx # imm = 0x7FFFFFFF +; X86-GISEL-NEXT: xorl %edx, %edx +; X86-GISEL-NEXT: cmpl $2139095040, %ecx # imm = 0x7F800000 +; X86-GISEL-NEXT: setb %bl +; X86-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-GISEL-NEXT: sete %ah +; X86-GISEL-NEXT: orb %dl, %ah +; X86-GISEL-NEXT: orb %bl, %ah +; X86-GISEL-NEXT: cmpl $2139095040, %ecx # imm = 0x7F800000 +; X86-GISEL-NEXT: seta %al +; X86-GISEL-NEXT: orb %ah, %al +; X86-GISEL-NEXT: popl %ebx +; X86-GISEL-NEXT: retl +; +; X64-GISEL-LABEL: not_is_minus_inf_f: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: movd %xmm0, %eax +; X64-GISEL-NEXT: movl %eax, %ecx +; X64-GISEL-NEXT: andl $2147483647, %ecx # imm = 0x7FFFFFFF +; X64-GISEL-NEXT: xorl %edx, %edx +; X64-GISEL-NEXT: cmpl $2139095040, %ecx # imm = 0x7F800000 +; X64-GISEL-NEXT: setb %sil +; X64-GISEL-NEXT: orb %dl, %sil +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: sete %dl +; X64-GISEL-NEXT: cmpl $2139095040, %ecx # imm = 0x7F800000 +; X64-GISEL-NEXT: seta %al +; X64-GISEL-NEXT: orb %dl, %al +; X64-GISEL-NEXT: orb %sil, %al +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 1019) ; ~"-inf" ret i1 %0 } define i1 @isfinite_f(float %x) nounwind { -; X86-LABEL: isfinite_f: -; X86: # %bb.0: # %entry -; X86-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF -; X86-NEXT: andl {{[0-9]+}}(%esp), %eax -; X86-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 -; X86-NEXT: setl %al -; X86-NEXT: retl +; X86-SDAGISEL-LABEL: isfinite_f: +; X86-SDAGISEL: # %bb.0: # %entry +; X86-SDAGISEL-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-SDAGISEL-NEXT: andl {{[0-9]+}}(%esp), %eax +; X86-SDAGISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-SDAGISEL-NEXT: setl %al +; X86-SDAGISEL-NEXT: retl ; ; X64-LABEL: isfinite_f: ; X64: # %bb.0: # %entry @@ -335,19 +545,39 @@ define i1 @isfinite_f(float %x) nounwind { ; X86-FASTISEL-NEXT: setl %al ; X86-FASTISEL-NEXT: popl %ecx ; X86-FASTISEL-NEXT: retl +; +; X86-GISEL-LABEL: isfinite_f: +; X86-GISEL: # %bb.0: # %entry +; X86-GISEL-NEXT: movl {{[0-9]+}}(%esp), %eax +; X86-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-GISEL-NEXT: xorl %ecx, %ecx +; X86-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-GISEL-NEXT: setb %al +; X86-GISEL-NEXT: orb %cl, %al +; X86-GISEL-NEXT: retl +; +; X64-GISEL-LABEL: isfinite_f: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: movd %xmm0, %eax +; X64-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X64-GISEL-NEXT: xorl %ecx, %ecx +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: setb %al +; X64-GISEL-NEXT: orb %cl, %al +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 504) ; 0x1f8 = "finite" ret i1 %0 } define i1 @not_isfinite_f(float %x) nounwind { -; X86-LABEL: not_isfinite_f: -; X86: # %bb.0: # %entry -; X86-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF -; X86-NEXT: andl {{[0-9]+}}(%esp), %eax -; X86-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 -; X86-NEXT: setge %al -; X86-NEXT: retl +; X86-SDAGISEL-LABEL: not_isfinite_f: +; X86-SDAGISEL: # %bb.0: # %entry +; X86-SDAGISEL-NEXT: movl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-SDAGISEL-NEXT: andl {{[0-9]+}}(%esp), %eax +; X86-SDAGISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-SDAGISEL-NEXT: setge %al +; X86-SDAGISEL-NEXT: retl ; ; X64-LABEL: not_isfinite_f: ; X64: # %bb.0: # %entry @@ -368,17 +598,43 @@ define i1 @not_isfinite_f(float %x) nounwind { ; X86-FASTISEL-NEXT: setge %al ; X86-FASTISEL-NEXT: popl %ecx ; X86-FASTISEL-NEXT: retl +; +; X86-GISEL-LABEL: not_isfinite_f: +; X86-GISEL: # %bb.0: # %entry +; X86-GISEL-NEXT: movl {{[0-9]+}}(%esp), %eax +; X86-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X86-GISEL-NEXT: xorl %ecx, %ecx +; X86-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-GISEL-NEXT: sete %dl +; X86-GISEL-NEXT: orb %cl, %dl +; X86-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X86-GISEL-NEXT: seta %al +; X86-GISEL-NEXT: orb %dl, %al +; X86-GISEL-NEXT: retl +; +; X64-GISEL-LABEL: not_isfinite_f: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: movd %xmm0, %eax +; X64-GISEL-NEXT: andl $2147483647, %eax # imm = 0x7FFFFFFF +; X64-GISEL-NEXT: xorl %ecx, %ecx +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: sete %dl +; X64-GISEL-NEXT: orb %cl, %dl +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: seta %al +; X64-GISEL-NEXT: orb %dl, %al +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 519) ; ~0x1f8 = "~finite" ret i1 %0 } define i1 @is_plus_finite_f(float %x) nounwind { -; X86-LABEL: is_plus_finite_f: -; X86: # %bb.0: # %entry -; X86-NEXT: cmpl $2139095040, {{[0-9]+}}(%esp) # imm = 0x7F800000 -; X86-NEXT: setb %al -; X86-NEXT: retl +; X86-SDAGISEL-LABEL: is_plus_finite_f: +; X86-SDAGISEL: # %bb.0: # %entry +; X86-SDAGISEL-NEXT: cmpl $2139095040, {{[0-9]+}}(%esp) # imm = 0x7F800000 +; X86-SDAGISEL-NEXT: setb %al +; X86-SDAGISEL-NEXT: retl ; ; X64-LABEL: is_plus_finite_f: ; X64: # %bb.0: # %entry @@ -396,6 +652,23 @@ define i1 @is_plus_finite_f(float %x) nounwind { ; X86-FASTISEL-NEXT: setb %al ; X86-FASTISEL-NEXT: popl %ecx ; X86-FASTISEL-NEXT: retl +; +; X86-GISEL-LABEL: is_plus_finite_f: +; X86-GISEL: # %bb.0: # %entry +; X86-GISEL-NEXT: xorl %ecx, %ecx +; X86-GISEL-NEXT: cmpl $2139095040, {{[0-9]+}}(%esp) # imm = 0x7F800000 +; X86-GISEL-NEXT: setb %al +; X86-GISEL-NEXT: orb %cl, %al +; X86-GISEL-NEXT: retl +; +; X64-GISEL-LABEL: is_plus_finite_f: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: xorl %ecx, %ecx +; X64-GISEL-NEXT: movd %xmm0, %eax +; X64-GISEL-NEXT: cmpl $2139095040, %eax # imm = 0x7F800000 +; X64-GISEL-NEXT: setb %al +; X64-GISEL-NEXT: orb %cl, %al +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f32(float %x, i32 448) ; 0x1c0 = "+finite" ret i1 %0 @@ -418,6 +691,11 @@ define i1 @isnone_d(double %x) nounwind { ; X86-FASTISEL-NEXT: fstp %st(0) ; X86-FASTISEL-NEXT: xorl %eax, %eax ; X86-FASTISEL-NEXT: retl +; +; X64-GISEL-LABEL: isnone_d: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: xorl %eax, %eax +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f64(double %x, i32 0) ret i1 %0 @@ -440,6 +718,11 @@ define i1 @isany_d(double %x) nounwind { ; X86-FASTISEL-NEXT: fstp %st(0) ; X86-FASTISEL-NEXT: movb $1, %al ; X86-FASTISEL-NEXT: retl +; +; X64-GISEL-LABEL: isany_d: +; X64-GISEL: # %bb.0: # %entry +; X64-GISEL-NEXT: movb $1, %al +; X64-GISEL-NEXT: retq entry: %0 = tail call i1 @llvm.is.fpclass.f64(double %x, i32 1023) ret i1 %0 diff --git a/llvm/test/CodeGen/X86/masked_gather_scatter.ll b/llvm/test/CodeGen/X86/masked_gather_scatter.ll index 4cde581..caec02e 100644 --- a/llvm/test/CodeGen/X86/masked_gather_scatter.ll +++ b/llvm/test/CodeGen/X86/masked_gather_scatter.ll @@ -4765,6 +4765,66 @@ define void @scaleidx_scatter_outofrange(<8 x float> %value, ptr %base, <8 x i32 } declare void @llvm.masked.scatter.v8f32.v8p0(<8 x float>, <8 x ptr>, i32 immarg, <8 x i1>) +define <16 x i32> @pr163023_sext(ptr %a0, <16 x i32> %a1) { +; X64-LABEL: pr163023_sext: +; X64: # %bb.0: +; X64-NEXT: kxnorw %k0, %k0, %k1 +; X64-NEXT: vpxor %xmm1, %xmm1, %xmm1 +; X64-NEXT: vpgatherdd (%rdi,%zmm0), %zmm1 {%k1} +; X64-NEXT: vmovdqa64 %zmm1, %zmm0 +; X64-NEXT: retq +; +; X86-LABEL: pr163023_sext: +; X86: # %bb.0: +; X86-NEXT: movl {{[0-9]+}}(%esp), %eax +; X86-NEXT: kxnorw %k0, %k0, %k1 +; X86-NEXT: vpxor %xmm1, %xmm1, %xmm1 +; X86-NEXT: vpgatherdd (%eax,%zmm0), %zmm1 {%k1} +; X86-NEXT: vmovdqa64 %zmm1, %zmm0 +; X86-NEXT: retl + %addr.p = ptrtoint ptr %a0 to i64 + %addr.v = insertelement <1 x i64> poison, i64 %addr.p, i64 0 + %addr.splat = shufflevector <1 x i64> %addr.v, <1 x i64> poison, <16 x i32> zeroinitializer + %ofs = sext <16 x i32> %a1 to <16 x i64> + %addr = add nuw <16 x i64> %addr.splat, %ofs + %ptr = inttoptr <16 x i64> %addr to <16 x ptr> + %gather = call <16 x i32> @llvm.masked.gather.v16i32.v16p0(<16 x ptr> %ptr, i32 4, <16 x i1> splat (i1 true), <16 x i32> poison) + ret <16 x i32> %gather +} + +define <16 x i32> @pr163023_zext(ptr %a0, <16 x i32> %a1) { +; X64-LABEL: pr163023_zext: +; X64: # %bb.0: +; X64-NEXT: vpmovzxdq {{.*#+}} zmm1 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero +; X64-NEXT: vextracti64x4 $1, %zmm0, %ymm0 +; X64-NEXT: vpmovzxdq {{.*#+}} zmm0 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero +; X64-NEXT: kxnorw %k0, %k0, %k1 +; X64-NEXT: vpxor %xmm2, %xmm2, %xmm2 +; X64-NEXT: vpxor %xmm3, %xmm3, %xmm3 +; X64-NEXT: kxnorw %k0, %k0, %k2 +; X64-NEXT: vpgatherqd (%rdi,%zmm0), %ymm3 {%k2} +; X64-NEXT: vpgatherqd (%rdi,%zmm1), %ymm2 {%k1} +; X64-NEXT: vinserti64x4 $1, %ymm3, %zmm2, %zmm0 +; X64-NEXT: retq +; +; X86-LABEL: pr163023_zext: +; X86: # %bb.0: +; X86-NEXT: movl {{[0-9]+}}(%esp), %eax +; X86-NEXT: kxnorw %k0, %k0, %k1 +; X86-NEXT: vpxor %xmm1, %xmm1, %xmm1 +; X86-NEXT: vpgatherdd (%eax,%zmm0), %zmm1 {%k1} +; X86-NEXT: vmovdqa64 %zmm1, %zmm0 +; X86-NEXT: retl + %addr.p = ptrtoint ptr %a0 to i64 + %addr.v = insertelement <1 x i64> poison, i64 %addr.p, i64 0 + %addr.splat = shufflevector <1 x i64> %addr.v, <1 x i64> poison, <16 x i32> zeroinitializer + %ofs = zext <16 x i32> %a1 to <16 x i64> + %addr = add nuw <16 x i64> %addr.splat, %ofs + %ptr = inttoptr <16 x i64> %addr to <16 x ptr> + %gather = call <16 x i32> @llvm.masked.gather.v16i32.v16p0(<16 x ptr> %ptr, i32 4, <16 x i1> splat (i1 true), <16 x i32> poison) + ret <16 x i32> %gather +} + ; ; PR45906 ; This used to cause fast-isel to generate bad copy instructions that would diff --git a/llvm/test/CodeGen/X86/pr160612.ll b/llvm/test/CodeGen/X86/pr160612.ll new file mode 100644 index 0000000..6572c42 --- /dev/null +++ b/llvm/test/CodeGen/X86/pr160612.ll @@ -0,0 +1,74 @@ +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py +; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -O2 | FileCheck %s + +; Test for issue #160612: OR conditions in branches should use multiple branches +; instead of materializing booleans with SETCC when no special optimizations apply. + +declare void @subroutine_foo() +declare void @subroutine_bar() + +; Original issue: (x == 0 || y == 0) was generating SETCC + TEST + BRANCH +; instead of using two conditional branches directly. +define void @func_a(i32 noundef %x, i32 noundef %y) { +; CHECK-LABEL: func_a: +; CHECK: # %bb.0: # %entry +; CHECK-NEXT: testl %edi, %edi +; CHECK-NEXT: je subroutine_foo@PLT # TAILCALL +; CHECK-NEXT: # %bb.1: # %entry +; CHECK-NEXT: testl %esi, %esi +; CHECK-NEXT: jne subroutine_bar@PLT # TAILCALL +; CHECK-NEXT: # %bb.2: # %if.then +; CHECK-NEXT: jmp subroutine_foo@PLT # TAILCALL +entry: + %cmp = icmp eq i32 %x, 0 + %cmp1 = icmp eq i32 %y, 0 + %or.cond = or i1 %cmp, %cmp1 + br i1 %or.cond, label %if.then, label %if.else + +if.then: + tail call void @subroutine_foo() + br label %if.end + +if.else: + tail call void @subroutine_bar() + br label %if.end + +if.end: + ret void +} + +; Reference implementation that already generated optimal code. +; This should continue to generate the same optimal code. +define void @func_b(i32 noundef %x, i32 noundef %y) { +; CHECK-LABEL: func_b: +; CHECK: # %bb.0: # %entry +; CHECK-NEXT: testl %edi, %edi +; CHECK-NEXT: je subroutine_foo@PLT # TAILCALL +; CHECK-NEXT: # %bb.1: # %if.else +; CHECK-NEXT: testl %esi, %esi +; CHECK-NEXT: je subroutine_foo@PLT # TAILCALL +; CHECK-NEXT: # %bb.2: # %if.else3 +; CHECK-NEXT: jmp subroutine_bar@PLT # TAILCALL +entry: + %cmp = icmp eq i32 %x, 0 + br i1 %cmp, label %if.then, label %if.else + +if.then: + tail call void @subroutine_foo() + br label %if.end4 + +if.else: + %cmp1 = icmp eq i32 %y, 0 + br i1 %cmp1, label %if.then2, label %if.else3 + +if.then2: + tail call void @subroutine_foo() + br label %if.end4 + +if.else3: + tail call void @subroutine_bar() + br label %if.end4 + +if.end4: + ret void +} diff --git a/llvm/test/CodeGen/X86/setcc-wide-types.ll b/llvm/test/CodeGen/X86/setcc-wide-types.ll index 69abf6e..d018c53 100644 --- a/llvm/test/CodeGen/X86/setcc-wide-types.ll +++ b/llvm/test/CodeGen/X86/setcc-wide-types.ll @@ -1493,15 +1493,23 @@ define i1 @allbits_i128_load_arg(ptr %w) { } define i1 @anybits_i256_load_arg(ptr %w) { -; ANY-LABEL: anybits_i256_load_arg: -; ANY: # %bb.0: -; ANY-NEXT: movq (%rdi), %rax -; ANY-NEXT: movq 8(%rdi), %rcx -; ANY-NEXT: orq 24(%rdi), %rcx -; ANY-NEXT: orq 16(%rdi), %rax -; ANY-NEXT: orq %rcx, %rax -; ANY-NEXT: setne %al -; ANY-NEXT: retq +; SSE-LABEL: anybits_i256_load_arg: +; SSE: # %bb.0: +; SSE-NEXT: movq (%rdi), %rax +; SSE-NEXT: movq 8(%rdi), %rcx +; SSE-NEXT: orq 24(%rdi), %rcx +; SSE-NEXT: orq 16(%rdi), %rax +; SSE-NEXT: orq %rcx, %rax +; SSE-NEXT: setne %al +; SSE-NEXT: retq +; +; AVXANY-LABEL: anybits_i256_load_arg: +; AVXANY: # %bb.0: +; AVXANY-NEXT: vmovdqu (%rdi), %ymm0 +; AVXANY-NEXT: vptest %ymm0, %ymm0 +; AVXANY-NEXT: setne %al +; AVXANY-NEXT: vzeroupper +; AVXANY-NEXT: retq %ld = load i256, ptr %w %cmp = icmp ne i256 %ld, 0 ret i1 %cmp @@ -1552,21 +1560,30 @@ define i1 @allbits_i256_load_arg(ptr %w) { } define i1 @anybits_i512_load_arg(ptr %w) { -; ANY-LABEL: anybits_i512_load_arg: -; ANY: # %bb.0: -; ANY-NEXT: movq 16(%rdi), %rax -; ANY-NEXT: movq (%rdi), %rcx -; ANY-NEXT: movq 8(%rdi), %rdx -; ANY-NEXT: movq 24(%rdi), %rsi -; ANY-NEXT: orq 56(%rdi), %rsi -; ANY-NEXT: orq 40(%rdi), %rdx -; ANY-NEXT: orq %rsi, %rdx -; ANY-NEXT: orq 48(%rdi), %rax -; ANY-NEXT: orq 32(%rdi), %rcx -; ANY-NEXT: orq %rax, %rcx -; ANY-NEXT: orq %rdx, %rcx -; ANY-NEXT: setne %al -; ANY-NEXT: retq +; NO512-LABEL: anybits_i512_load_arg: +; NO512: # %bb.0: +; NO512-NEXT: movq 16(%rdi), %rax +; NO512-NEXT: movq (%rdi), %rcx +; NO512-NEXT: movq 8(%rdi), %rdx +; NO512-NEXT: movq 24(%rdi), %rsi +; NO512-NEXT: orq 56(%rdi), %rsi +; NO512-NEXT: orq 40(%rdi), %rdx +; NO512-NEXT: orq %rsi, %rdx +; NO512-NEXT: orq 48(%rdi), %rax +; NO512-NEXT: orq 32(%rdi), %rcx +; NO512-NEXT: orq %rax, %rcx +; NO512-NEXT: orq %rdx, %rcx +; NO512-NEXT: setne %al +; NO512-NEXT: retq +; +; AVX512-LABEL: anybits_i512_load_arg: +; AVX512: # %bb.0: +; AVX512-NEXT: vmovdqu64 (%rdi), %zmm0 +; AVX512-NEXT: vptestmd %zmm0, %zmm0, %k0 +; AVX512-NEXT: kortestw %k0, %k0 +; AVX512-NEXT: setne %al +; AVX512-NEXT: vzeroupper +; AVX512-NEXT: retq %ld = load i512, ptr %w %cmp = icmp ne i512 %ld, 0 ret i1 %cmp diff --git a/llvm/test/CodeGen/X86/sqrt-fastmath-mir.ll b/llvm/test/CodeGen/X86/sqrt-fastmath-mir.ll index 42617c1..18588aa 100644 --- a/llvm/test/CodeGen/X86/sqrt-fastmath-mir.ll +++ b/llvm/test/CodeGen/X86/sqrt-fastmath-mir.ll @@ -24,7 +24,7 @@ define float @sqrt_ieee_ninf(float %f) #0 { ; CHECK-NEXT: {{ $}} ; CHECK-NEXT: [[COPY:%[0-9]+]]:fr32 = COPY $xmm0 ; CHECK-NEXT: [[DEF:%[0-9]+]]:fr32 = IMPLICIT_DEF - ; CHECK-NEXT: [[VRSQRTSSr:%[0-9]+]]:fr32 = VRSQRTSSr killed [[DEF]], [[COPY]] + ; CHECK-NEXT: [[VRSQRTSSr:%[0-9]+]]:fr32 = ninf afn VRSQRTSSr killed [[DEF]], [[COPY]] ; CHECK-NEXT: [[VMULSSrr:%[0-9]+]]:fr32 = ninf afn nofpexcept VMULSSrr [[COPY]], [[VRSQRTSSr]], implicit $mxcsr ; CHECK-NEXT: [[VMOVSSrm_alt:%[0-9]+]]:fr32 = VMOVSSrm_alt $rip, 1, $noreg, %const.0, $noreg :: (load (s32) from constant-pool) ; CHECK-NEXT: [[VFMADD213SSr:%[0-9]+]]:fr32 = ninf afn nofpexcept VFMADD213SSr [[VRSQRTSSr]], killed [[VMULSSrr]], [[VMOVSSrm_alt]], implicit $mxcsr @@ -71,7 +71,7 @@ define float @sqrt_daz_ninf(float %f) #1 { ; CHECK-NEXT: {{ $}} ; CHECK-NEXT: [[COPY:%[0-9]+]]:fr32 = COPY $xmm0 ; CHECK-NEXT: [[DEF:%[0-9]+]]:fr32 = IMPLICIT_DEF - ; CHECK-NEXT: [[VRSQRTSSr:%[0-9]+]]:fr32 = VRSQRTSSr killed [[DEF]], [[COPY]] + ; CHECK-NEXT: [[VRSQRTSSr:%[0-9]+]]:fr32 = ninf afn VRSQRTSSr killed [[DEF]], [[COPY]] ; CHECK-NEXT: [[VMULSSrr:%[0-9]+]]:fr32 = ninf afn nofpexcept VMULSSrr [[COPY]], [[VRSQRTSSr]], implicit $mxcsr ; CHECK-NEXT: [[VMOVSSrm_alt:%[0-9]+]]:fr32 = VMOVSSrm_alt $rip, 1, $noreg, %const.0, $noreg :: (load (s32) from constant-pool) ; CHECK-NEXT: [[VFMADD213SSr:%[0-9]+]]:fr32 = ninf afn nofpexcept VFMADD213SSr [[VRSQRTSSr]], killed [[VMULSSrr]], [[VMOVSSrm_alt]], implicit $mxcsr diff --git a/llvm/test/CodeGen/X86/x86-shrink-wrap-unwind.ll b/llvm/test/CodeGen/X86/x86-shrink-wrap-unwind.ll index b2064b1..02d4d88 100644 --- a/llvm/test/CodeGen/X86/x86-shrink-wrap-unwind.ll +++ b/llvm/test/CodeGen/X86/x86-shrink-wrap-unwind.ll @@ -181,40 +181,38 @@ define zeroext i1 @segmentedStack(ptr readonly %vk1, ptr readonly %vk2, i64 %key ; CHECK-LABEL: segmentedStack: ; CHECK: ## %bb.0: ; CHECK-NEXT: cmpq %gs:816, %rsp -; CHECK-NEXT: jbe LBB3_6 +; CHECK-NEXT: jbe LBB3_7 ; CHECK-NEXT: LBB3_1: ## %entry ; CHECK-NEXT: pushq %rax ; CHECK-NEXT: .cfi_def_cfa_offset 16 -; CHECK-NEXT: testq %rdi, %rdi -; CHECK-NEXT: sete %al -; CHECK-NEXT: testq %rsi, %rsi -; CHECK-NEXT: sete %cl -; CHECK-NEXT: orb %al, %cl ; CHECK-NEXT: movq %rdi, %rax ; CHECK-NEXT: orq %rsi, %rax ; CHECK-NEXT: sete %al -; CHECK-NEXT: testb %cl, %cl -; CHECK-NEXT: jne LBB3_4 -; CHECK-NEXT: ## %bb.2: ## %if.end4.i +; CHECK-NEXT: testq %rdi, %rdi +; CHECK-NEXT: je LBB3_5 +; CHECK-NEXT: ## %bb.2: ## %entry +; CHECK-NEXT: testq %rsi, %rsi +; CHECK-NEXT: je LBB3_5 +; CHECK-NEXT: ## %bb.3: ## %if.end4.i ; CHECK-NEXT: movq 8(%rdi), %rdx ; CHECK-NEXT: cmpq 8(%rsi), %rdx -; CHECK-NEXT: jne LBB3_5 -; CHECK-NEXT: ## %bb.3: ## %land.rhs.i.i +; CHECK-NEXT: jne LBB3_6 +; CHECK-NEXT: ## %bb.4: ## %land.rhs.i.i ; CHECK-NEXT: movq (%rsi), %rsi ; CHECK-NEXT: movq (%rdi), %rdi ; CHECK-NEXT: callq _memcmp ; CHECK-NEXT: testl %eax, %eax ; CHECK-NEXT: sete %al -; CHECK-NEXT: LBB3_4: ## %__go_ptr_strings_equal.exit +; CHECK-NEXT: LBB3_5: ## %__go_ptr_strings_equal.exit ; CHECK-NEXT: ## kill: def $al killed $al killed $eax ; CHECK-NEXT: popq %rcx ; CHECK-NEXT: retq -; CHECK-NEXT: LBB3_5: +; CHECK-NEXT: LBB3_6: ; CHECK-NEXT: xorl %eax, %eax ; CHECK-NEXT: ## kill: def $al killed $al killed $eax ; CHECK-NEXT: popq %rcx ; CHECK-NEXT: retq -; CHECK-NEXT: LBB3_6: +; CHECK-NEXT: LBB3_7: ; CHECK-NEXT: movl $8, %r10d ; CHECK-NEXT: movl $0, %r11d ; CHECK-NEXT: callq ___morestack @@ -224,43 +222,41 @@ define zeroext i1 @segmentedStack(ptr readonly %vk1, ptr readonly %vk2, i64 %key ; NOCOMPACTUNWIND-LABEL: segmentedStack: ; NOCOMPACTUNWIND: # %bb.0: ; NOCOMPACTUNWIND-NEXT: cmpq %fs:112, %rsp -; NOCOMPACTUNWIND-NEXT: jbe .LBB3_6 +; NOCOMPACTUNWIND-NEXT: jbe .LBB3_7 ; NOCOMPACTUNWIND-NEXT: .LBB3_1: # %entry ; NOCOMPACTUNWIND-NEXT: pushq %rax ; NOCOMPACTUNWIND-NEXT: .cfi_def_cfa_offset 16 -; NOCOMPACTUNWIND-NEXT: testq %rdi, %rdi -; NOCOMPACTUNWIND-NEXT: sete %al -; NOCOMPACTUNWIND-NEXT: testq %rsi, %rsi -; NOCOMPACTUNWIND-NEXT: sete %cl -; NOCOMPACTUNWIND-NEXT: orb %al, %cl ; NOCOMPACTUNWIND-NEXT: movq %rdi, %rax ; NOCOMPACTUNWIND-NEXT: orq %rsi, %rax ; NOCOMPACTUNWIND-NEXT: sete %al -; NOCOMPACTUNWIND-NEXT: testb %cl, %cl -; NOCOMPACTUNWIND-NEXT: jne .LBB3_4 -; NOCOMPACTUNWIND-NEXT: # %bb.2: # %if.end4.i +; NOCOMPACTUNWIND-NEXT: testq %rdi, %rdi +; NOCOMPACTUNWIND-NEXT: je .LBB3_5 +; NOCOMPACTUNWIND-NEXT: # %bb.2: # %entry +; NOCOMPACTUNWIND-NEXT: testq %rsi, %rsi +; NOCOMPACTUNWIND-NEXT: je .LBB3_5 +; NOCOMPACTUNWIND-NEXT: # %bb.3: # %if.end4.i ; NOCOMPACTUNWIND-NEXT: movq 8(%rdi), %rdx ; NOCOMPACTUNWIND-NEXT: cmpq 8(%rsi), %rdx -; NOCOMPACTUNWIND-NEXT: jne .LBB3_5 -; NOCOMPACTUNWIND-NEXT: # %bb.3: # %land.rhs.i.i +; NOCOMPACTUNWIND-NEXT: jne .LBB3_6 +; NOCOMPACTUNWIND-NEXT: # %bb.4: # %land.rhs.i.i ; NOCOMPACTUNWIND-NEXT: movq (%rsi), %rsi ; NOCOMPACTUNWIND-NEXT: movq (%rdi), %rdi ; NOCOMPACTUNWIND-NEXT: callq memcmp@PLT ; NOCOMPACTUNWIND-NEXT: testl %eax, %eax ; NOCOMPACTUNWIND-NEXT: sete %al -; NOCOMPACTUNWIND-NEXT: .LBB3_4: # %__go_ptr_strings_equal.exit +; NOCOMPACTUNWIND-NEXT: .LBB3_5: # %__go_ptr_strings_equal.exit ; NOCOMPACTUNWIND-NEXT: # kill: def $al killed $al killed $eax ; NOCOMPACTUNWIND-NEXT: popq %rcx ; NOCOMPACTUNWIND-NEXT: .cfi_def_cfa_offset 8 ; NOCOMPACTUNWIND-NEXT: retq -; NOCOMPACTUNWIND-NEXT: .LBB3_5: +; NOCOMPACTUNWIND-NEXT: .LBB3_6: ; NOCOMPACTUNWIND-NEXT: .cfi_def_cfa_offset 16 ; NOCOMPACTUNWIND-NEXT: xorl %eax, %eax ; NOCOMPACTUNWIND-NEXT: # kill: def $al killed $al killed $eax ; NOCOMPACTUNWIND-NEXT: popq %rcx ; NOCOMPACTUNWIND-NEXT: .cfi_def_cfa_offset 8 ; NOCOMPACTUNWIND-NEXT: retq -; NOCOMPACTUNWIND-NEXT: .LBB3_6: +; NOCOMPACTUNWIND-NEXT: .LBB3_7: ; NOCOMPACTUNWIND-NEXT: movl $8, %r10d ; NOCOMPACTUNWIND-NEXT: movl $0, %r11d ; NOCOMPACTUNWIND-NEXT: callq __morestack diff --git a/llvm/test/MC/AArch64/data-directive-specifier.s b/llvm/test/MC/AArch64/data-directive-specifier.s index 2cb7eb3..2d1ec4f 100644 --- a/llvm/test/MC/AArch64/data-directive-specifier.s +++ b/llvm/test/MC/AArch64/data-directive-specifier.s @@ -12,6 +12,7 @@ l: # CHECK-NEXT: 0x8 R_AARCH64_PLT32 extern 0x4 # CHECK-NEXT: 0xC R_AARCH64_PLT32 g 0x8 # CHECK-NEXT: 0x10 R_AARCH64_PLT32 g 0x18 +# CHECK-NEXT: 0x14 R_AARCH64_FUNCINIT64 .text 0x0 # CHECK-NEXT: } .data .word l@plt - . @@ -21,6 +22,8 @@ l: .word g@plt - . + 8 .word g@plt - .data + 8 +.quad l@funcinit + # CHECK: Section ({{.*}}) .rela.data1 { # CHECK-NEXT: 0x0 R_AARCH64_GOTPCREL32 data1 0x0 # CHECK-NEXT: 0x4 R_AARCH64_GOTPCREL32 extern 0x4 diff --git a/llvm/test/MC/X86/verify-callgraph-section.s b/llvm/test/MC/X86/verify-callgraph-section.s index ce07228..9be5a68 100644 --- a/llvm/test/MC/X86/verify-callgraph-section.s +++ b/llvm/test/MC/X86/verify-callgraph-section.s @@ -2,7 +2,7 @@ /// (annotated by generated temporary labels .Ltmp*) are associated /// with the corresponding callee type identifiers. -// RUN: llvm-mc -triple=x86_64 -filetype=obj -o - < %s | llvm-readelf -x .callgraph - | FileCheck %s +// RUN: llvm-mc -triple=x86_64 -filetype=obj -o - < %s | llvm-readelf -x .llvm.callgraph - | FileCheck %s .text .globl ball # -- Begin function ball @@ -38,7 +38,7 @@ ball: # @ball addq $32, %rsp popq %rbx retq - .section .callgraph,"o",@progbits,.text + .section .llvm.callgraph,"o",@progbits,.text .quad 0 .quad .Lfunc_begin0 .quad 1 diff --git a/llvm/test/TableGen/directive1.td b/llvm/test/TableGen/directive1.td index 3eda077..475faf9 100644 --- a/llvm/test/TableGen/directive1.td +++ b/llvm/test/TableGen/directive1.td @@ -177,6 +177,7 @@ def TDL_DirA : Directive<[Spelling<"dira">]> { // CHECK-NEXT: static constexpr bool is_iterable = true; // CHECK-NEXT: }; // CHECK-NEXT: } // namespace llvm +// CHECK-EMPTY: // CHECK-NEXT: #endif // LLVM_Tdl_INC diff --git a/llvm/test/TableGen/directive2.td b/llvm/test/TableGen/directive2.td index a25197c..ccc0944 100644 --- a/llvm/test/TableGen/directive2.td +++ b/llvm/test/TableGen/directive2.td @@ -150,6 +150,7 @@ def TDL_DirA : Directive<[Spelling<"dira">]> { // CHECK-NEXT: static constexpr bool is_iterable = true; // CHECK-NEXT: }; // CHECK-NEXT: } // namespace llvm +// CHECK-EMPTY: // CHECK-NEXT: #endif // LLVM_Tdl_INC // IMPL: #ifdef GEN_FLANG_DIRECTIVE_CLAUSE_SETS diff --git a/llvm/test/Transforms/DFAJumpThreading/dfa-constant-propagation.ll b/llvm/test/Transforms/DFAJumpThreading/dfa-constant-propagation.ll index fdab67a..afc98ce 100644 --- a/llvm/test/Transforms/DFAJumpThreading/dfa-constant-propagation.ll +++ b/llvm/test/Transforms/DFAJumpThreading/dfa-constant-propagation.ll @@ -1,5 +1,5 @@ ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py -; RUN: opt -S -passes=dfa-jump-threading,sccp,simplifycfg %s | FileCheck %s +; RUN: opt -S -passes=dfa-jump-threading,sccp,simplifycfg -verify-dom-info=1 %s | FileCheck %s ; This test checks that a constant propagation is applied for a basic loop. ; Related to bug 44679. diff --git a/llvm/test/Transforms/DFAJumpThreading/dfa-jump-threading-analysis.ll b/llvm/test/Transforms/DFAJumpThreading/dfa-jump-threading-analysis.ll index f45798b..5076517 100644 --- a/llvm/test/Transforms/DFAJumpThreading/dfa-jump-threading-analysis.ll +++ b/llvm/test/Transforms/DFAJumpThreading/dfa-jump-threading-analysis.ll @@ -1,6 +1,6 @@ ; REQUIRES: asserts -; RUN: opt -S -passes=dfa-jump-threading -debug-only=dfa-jump-threading -disable-output %s 2>&1 | FileCheck %s -; RUN: opt -S -passes=dfa-jump-threading -print-prof-data %s -o - | FileCheck %s --check-prefix=PROFILE +; RUN: opt -S -passes=dfa-jump-threading -verify-dom-info=1 -debug-only=dfa-jump-threading -disable-output %s 2>&1 | FileCheck %s +; RUN: opt -S -passes=dfa-jump-threading -verify-dom-info=1 -print-prof-data %s -o - | FileCheck %s --check-prefix=PROFILE ; This test checks that the analysis identifies all threadable paths in a ; simple CFG. A threadable path includes a list of basic blocks, the exit diff --git a/llvm/test/Transforms/DFAJumpThreading/dfa-jump-threading-transform.ll b/llvm/test/Transforms/DFAJumpThreading/dfa-jump-threading-transform.ll index 092c854..426b51e 100644 --- a/llvm/test/Transforms/DFAJumpThreading/dfa-jump-threading-transform.ll +++ b/llvm/test/Transforms/DFAJumpThreading/dfa-jump-threading-transform.ll @@ -1,5 +1,5 @@ ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals -; RUN: opt -S -passes=dfa-jump-threading %s | FileCheck %s +; RUN: opt -S -passes=dfa-jump-threading -verify-dom-info=1 %s | FileCheck %s ; These tests check that the DFA jump threading transformation is applied ; properly to two CFGs. It checks that blocks are cloned, branches are updated, @@ -445,9 +445,67 @@ bb2: ; preds = %select.unfold unreachable } + +define i16 @DTU_update_crash() { +; CHECK-LABEL: @DTU_update_crash( +; CHECK-NEXT: entry: +; CHECK-NEXT: br label [[FOR_BODY_SELECTBLOCK:%.*]] +; CHECK: for.body.selectblock: +; CHECK-NEXT: br i1 false, label [[SWITCHBLOCK_JT0:%.*]], label [[SEL_SI_UNFOLD_FALSE_JT0:%.*]] +; CHECK: sel.si.unfold.false: +; CHECK-NEXT: br label [[SWITCHBLOCK:%.*]] +; CHECK: sel.si.unfold.false.jt0: +; CHECK-NEXT: [[DOTSI_UNFOLD_PHI_JT0:%.*]] = phi i32 [ 0, [[FOR_BODY_SELECTBLOCK]] ] +; CHECK-NEXT: br label [[SWITCHBLOCK_JT0]] +; CHECK: switchblock: +; CHECK-NEXT: [[SWITCHBLOCK_PHI:%.*]] = phi i32 [ poison, [[SEL_SI_UNFOLD_FALSE:%.*]] ] +; CHECK-NEXT: [[P_24_ADDR_3:%.*]] = phi i32 [ 0, [[SEL_SI_UNFOLD_FALSE]] ] +; CHECK-NEXT: switch i32 [[SWITCHBLOCK_PHI]], label [[CLEANUP:%.*]] [ +; CHECK-NEXT: i32 0, label [[FOR_INC:%.*]] +; CHECK-NEXT: i32 1, label [[CLEANUP]] +; CHECK-NEXT: i32 5, label [[FOR_BODY_SELECTBLOCK]] +; CHECK-NEXT: ] +; CHECK: switchblock.jt0: +; CHECK-NEXT: [[SWITCHBLOCK_PHI_JT0:%.*]] = phi i32 [ 0, [[FOR_BODY_SELECTBLOCK]] ], [ [[DOTSI_UNFOLD_PHI_JT0]], [[SEL_SI_UNFOLD_FALSE_JT0]] ] +; CHECK-NEXT: [[P_24_ADDR_3_JT0:%.*]] = phi i32 [ 0, [[FOR_BODY_SELECTBLOCK]] ], [ 0, [[SEL_SI_UNFOLD_FALSE_JT0]] ] +; CHECK-NEXT: br label [[FOR_INC]] +; CHECK: for.inc: +; CHECK-NEXT: br i1 false, label [[FOR_BODY_SELECTBLOCK]], label [[CLEANUP]] +; CHECK: cleanup: +; CHECK-NEXT: call void (...) @llvm.fake.use(i32 [[P_24_ADDR_3_JT0]]) +; CHECK-NEXT: ret i16 0 +; +entry: + br label %for.body.selectblock + +for.body.selectblock: ; preds = %for.inc, %switchblock, %entry + %sel = select i1 false, i32 0, i32 0 + br label %switchblock + +switchblock: ; preds = %for.body.selectblock + %switchblock.phi = phi i32 [ %sel, %for.body.selectblock ] + %p_24.addr.3 = phi i32 [ 0, %for.body.selectblock ] + switch i32 %switchblock.phi, label %cleanup [ + i32 0, label %for.inc + i32 1, label %cleanup + i32 5, label %for.body.selectblock + ] + +for.inc: ; preds = %switchblock + br i1 false, label %for.body.selectblock, label %cleanup + +cleanup: ; preds = %for.inc, %switchblock, %switchblock + call void (...) @llvm.fake.use(i32 %p_24.addr.3) + ret i16 0 +} + +declare void @llvm.fake.use(...) + !0 = !{!"function_entry_count", i32 10} !1 = !{!"branch_weights", i32 3, i32 5} ;. +; CHECK: attributes #[[ATTR0:[0-9]+]] = { nocallback nofree nosync nounwind willreturn memory(inaccessiblemem: readwrite) } +;. ; CHECK: [[META0:![0-9]+]] = !{!"function_entry_count", i32 10} ; CHECK: [[PROF1]] = !{!"branch_weights", i32 3, i32 5} ;. diff --git a/llvm/test/Transforms/DFAJumpThreading/dfa-unfold-select.ll b/llvm/test/Transforms/DFAJumpThreading/dfa-unfold-select.ll index de38752..95d3ffa 100644 --- a/llvm/test/Transforms/DFAJumpThreading/dfa-unfold-select.ll +++ b/llvm/test/Transforms/DFAJumpThreading/dfa-unfold-select.ll @@ -1,5 +1,5 @@ ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py -; RUN: opt -S -passes=dfa-jump-threading -dfa-early-exit-heuristic=false %s | FileCheck %s +; RUN: opt -S -passes=dfa-jump-threading -dfa-early-exit-heuristic=false -verify-dom-info=1 %s | FileCheck %s ; These tests check if selects are unfolded properly for jump threading ; opportunities. There are three different patterns to consider: diff --git a/llvm/test/Transforms/DFAJumpThreading/equivalent-states.ll b/llvm/test/Transforms/DFAJumpThreading/equivalent-states.ll index 4555dfb..71a469d 100644 --- a/llvm/test/Transforms/DFAJumpThreading/equivalent-states.ll +++ b/llvm/test/Transforms/DFAJumpThreading/equivalent-states.ll @@ -1,5 +1,5 @@ ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 3 -; RUN: opt -S -passes=dfa-jump-threading %s | FileCheck %s +; RUN: opt -S -passes=dfa-jump-threading -verify-dom-info=1 %s | FileCheck %s declare void @do_something() declare void @user(i32) diff --git a/llvm/test/Transforms/DFAJumpThreading/single_succ_switch.ll b/llvm/test/Transforms/DFAJumpThreading/single_succ_switch.ll index 00500a7..cc117e7 100644 --- a/llvm/test/Transforms/DFAJumpThreading/single_succ_switch.ll +++ b/llvm/test/Transforms/DFAJumpThreading/single_succ_switch.ll @@ -1,5 +1,5 @@ ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 3 -; RUN: opt -S -passes=dfa-jump-threading %s | FileCheck %s +; RUN: opt -S -passes=dfa-jump-threading -verify-dom-info=1 %s | FileCheck %s define void @pr60254() { ; CHECK-LABEL: define void @pr60254() { diff --git a/llvm/test/Transforms/IndVarSimplify/pointer-loop-guards.ll b/llvm/test/Transforms/IndVarSimplify/pointer-loop-guards.ll index 89b132e..9371fe2 100644 --- a/llvm/test/Transforms/IndVarSimplify/pointer-loop-guards.ll +++ b/llvm/test/Transforms/IndVarSimplify/pointer-loop-guards.ll @@ -18,7 +18,7 @@ define i64 @test_ptr_compare_guard(ptr %start, ptr %end) { ; CHECK-NEXT: br i1 [[C_1]], label %[[LOOP_LATCH]], label %[[EXIT_LOOPEXIT:.*]] ; CHECK: [[LOOP_LATCH]]: ; CHECK-NEXT: [[PTR_IV_NEXT]] = getelementptr i8, ptr [[PTR_IV]], i64 1 -; CHECK-NEXT: [[I64_IV_NEXT]] = add i64 [[I64_IV]], 1 +; CHECK-NEXT: [[I64_IV_NEXT]] = add nuw i64 [[I64_IV]], 1 ; CHECK-NEXT: [[C_2:%.*]] = icmp eq ptr [[PTR_IV_NEXT]], [[END]] ; CHECK-NEXT: br i1 [[C_2]], label %[[EXIT_LOOPEXIT]], label %[[LOOP_HEADER]] ; CHECK: [[EXIT_LOOPEXIT]]: diff --git a/llvm/test/Transforms/Inline/ML/state-accounting-skip-non-cold.ll b/llvm/test/Transforms/Inline/ML/state-accounting-skip-non-cold.ll new file mode 100644 index 0000000..0887f5e --- /dev/null +++ b/llvm/test/Transforms/Inline/ML/state-accounting-skip-non-cold.ll @@ -0,0 +1,55 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 6 +; REQUIRES: llvm_inliner_model_autogenerated && asserts +; RUN: opt -passes='default<O3>' -enable-ml-inliner=release -ml-inliner-skip-policy=if-caller-not-cold -S %s -o - | FileCheck %s +; RUN: opt -passes='default<O3>' -ml-inliner-stop-immediately -enable-ml-inliner=release -ml-inliner-skip-policy=if-caller-not-cold -S %s -o - | FileCheck %s + +declare ptr @f() + +define void @e() #0 { +; CHECK-LABEL: define void @e( +; CHECK-SAME: ) local_unnamed_addr #[[ATTR0:[0-9]+]] { +; CHECK-NEXT: tail call void @d() +; CHECK-NEXT: tail call void @g() +; CHECK-NEXT: tail call void @d() +; CHECK-NEXT: tail call void @g() +; CHECK-NEXT: tail call void @d() +; CHECK-NEXT: tail call void @g() +; CHECK-NEXT: ret void +; + call void @h() + call void @h() + call void @h() + ret void +} + +define void @d() { +; CHECK-LABEL: define void @d() local_unnamed_addr { +; CHECK-NEXT: tail call void @f() +; CHECK-NEXT: ret void +; + call void @f() + ret void +} + +define void @g() { +; CHECK-LABEL: define void @g() local_unnamed_addr { +; CHECK-NEXT: tail call void @f() +; CHECK-NEXT: ret void +; + call void @f() + ret void +} + +define void @h() #0 { +; CHECK-LABEL: define void @h( +; CHECK-SAME: ) local_unnamed_addr #[[ATTR0]] { +; CHECK-NEXT: tail call void @d() +; CHECK-NEXT: tail call void @g() +; CHECK-NEXT: ret void +; + call void @d() + call void @g() + ret void +} + +attributes #0 = { "sign-return-address"="non-leaf" "sign-return-address-key"="a_key" } diff --git a/llvm/test/Transforms/InstCombine/select-safe-impliedcond-transforms.ll b/llvm/test/Transforms/InstCombine/select-safe-impliedcond-transforms.ll index ba34930..bc988a9 100644 --- a/llvm/test/Transforms/InstCombine/select-safe-impliedcond-transforms.ll +++ b/llvm/test/Transforms/InstCombine/select-safe-impliedcond-transforms.ll @@ -1,4 +1,4 @@ -; NOTE: Assertions have been autogenerated by utils/update_test_checks.py +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals ; RUN: opt < %s -passes=instcombine -S | FileCheck %s define i1 @a_true_implies_b_true(i8 %z, i1 %X, i1 %Y) { @@ -34,15 +34,15 @@ define <2 x i1> @a_true_implies_b_true_vec(i8 %z0, <2 x i1> %X, <2 x i1> %Y) { ret <2 x i1> %res } -define i1 @a_true_implies_b_true2(i8 %z, i1 %X, i1 %Y) { +define i1 @a_true_implies_b_true2(i8 %z, i1 %X, i1 %Y) !prof !0 { ; CHECK-LABEL: @a_true_implies_b_true2( ; CHECK-NEXT: [[A:%.*]] = icmp ugt i8 [[Z:%.*]], 20 -; CHECK-NEXT: [[RES:%.*]] = select i1 [[A]], i1 [[X:%.*]], i1 false +; CHECK-NEXT: [[RES:%.*]] = select i1 [[A]], i1 [[X:%.*]], i1 false, !prof [[PROF1:![0-9]+]] ; CHECK-NEXT: ret i1 [[RES]] ; %a = icmp ugt i8 %z, 20 %b = icmp ugt i8 %z, 10 - %sel = select i1 %b, i1 %X, i1 %Y + %sel = select i1 %b, i1 %X, i1 %Y, !prof !1 %res = and i1 %a, %sel ret i1 %res } @@ -258,3 +258,10 @@ define i1 @neg_icmp_eq_implies_trunc(i8 %x, i1 %c) { %sel2 = select i1 %cmp, i1 true, i1 %sel1 ret i1 %sel2 } + +!0 = !{!"function_entry_count", i64 1000} +!1 = !{!"branch_weights", i32 2, i32 3} +;. +; CHECK: [[META0:![0-9]+]] = !{!"function_entry_count", i64 1000} +; CHECK: [[PROF1]] = !{!"branch_weights", i32 2, i32 3} +;. diff --git a/llvm/test/Transforms/InstSimplify/ptrmask.ll b/llvm/test/Transforms/InstSimplify/ptrmask.ll index 5e7c636..a3483af 100644 --- a/llvm/test/Transforms/InstSimplify/ptrmask.ll +++ b/llvm/test/Transforms/InstSimplify/ptrmask.ll @@ -158,6 +158,26 @@ define ptr addrspace(1) @ptrmask_simplify_ptrmask_i32(ptr addrspace(1) %p) { ret ptr addrspace(1) %r } +define ptr @ptrmask_simplify_ptrtoaddr(ptr %p) { +; CHECK-LABEL: define ptr @ptrmask_simplify_ptrtoaddr +; CHECK-SAME: (ptr [[P:%.*]]) { +; CHECK-NEXT: ret ptr [[P]] +; + %m = ptrtoaddr ptr %p to i64 + %r = call ptr @llvm.ptrmask.p0.i64(ptr %p, i64 %m) + ret ptr %r +} + +define ptr addrspace(1) @ptrmask_simplify_ptrtoaddr_i32(ptr addrspace(1) %p) { +; CHECK-LABEL: define ptr addrspace(1) @ptrmask_simplify_ptrtoaddr_i32 +; CHECK-SAME: (ptr addrspace(1) [[P:%.*]]) { +; CHECK-NEXT: ret ptr addrspace(1) [[P]] +; + %m = ptrtoaddr ptr addrspace(1) %p to i32 + %r = call ptr addrspace(1) @llvm.ptrmask.p1.i32(ptr addrspace(1) %p, i32 %m) + ret ptr addrspace(1) %r +} + define ptr @ptrmask_simplify_aligned_unused(ptr align 64 %p) { ; CHECK-LABEL: define ptr @ptrmask_simplify_aligned_unused ; CHECK-SAME: (ptr align 64 [[P:%.*]]) { diff --git a/llvm/test/Transforms/LoopUnroll/scevunroll.ll b/llvm/test/Transforms/LoopUnroll/scevunroll.ll index fa55eab..bc63f79 100644 --- a/llvm/test/Transforms/LoopUnroll/scevunroll.ll +++ b/llvm/test/Transforms/LoopUnroll/scevunroll.ll @@ -465,8 +465,7 @@ define void @peel_int_eq_condition(i32 %start) { ; CHECK-NEXT: br label [[LOOP:%.*]] ; CHECK: loop: ; CHECK-NEXT: [[IV:%.*]] = phi i32 [ [[IV_NEXT_PEEL]], [[ENTRY_PEEL_NEWPH]] ], [ [[IV_NEXT:%.*]], [[LOOP_LATCH:%.*]] ] -; CHECK-NEXT: [[C_0:%.*]] = icmp eq i32 [[IV]], [[START]] -; CHECK-NEXT: br i1 [[C_0]], label [[IF_THEN:%.*]], label [[LOOP_LATCH]] +; CHECK-NEXT: br i1 false, label [[IF_THEN:%.*]], label [[LOOP_LATCH]] ; CHECK: if.then: ; CHECK-NEXT: call void @fn(i32 [[IV]]) ; CHECK-NEXT: br label [[LOOP_LATCH]] diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/partial-reduce-dot-product.ll b/llvm/test/Transforms/LoopVectorize/AArch64/partial-reduce-dot-product.ll index 3a88273..56a5663 100644 --- a/llvm/test/Transforms/LoopVectorize/AArch64/partial-reduce-dot-product.ll +++ b/llvm/test/Transforms/LoopVectorize/AArch64/partial-reduce-dot-product.ll @@ -2697,4 +2697,4 @@ for.body: ; preds = %for.body.lr.ph, %fo !9 = !{!"llvm.loop.vectorize.predicate.enable", i1 true} !10 = !{!"llvm.loop.vectorize.enable", i1 true} attributes #0 = { vscale_range(1,16) "target-features"="+sve" } -attributes #1 = { vscale_range(1,16) "target-features"="+neon,+dotprod,+sve" "cpu"="neoverse-v2" } +attributes #1 = { vscale_range(1,16) "target-features"="+neon,+dotprod,+sve" "target-cpu"="neoverse-v2" } diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/partial-reduce.ll b/llvm/test/Transforms/LoopVectorize/AArch64/partial-reduce.ll index 3c2ae1c7..1e6bcb1 100644 --- a/llvm/test/Transforms/LoopVectorize/AArch64/partial-reduce.ll +++ b/llvm/test/Transforms/LoopVectorize/AArch64/partial-reduce.ll @@ -410,20 +410,32 @@ define i32 @zext_add_reduc_i8_i32_has_neon_dotprod(ptr %a) #1 { ; CHECK-INTERLEAVED-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] ; CHECK-INTERLEAVED-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[PARTIAL_REDUCE:%.*]], [[VECTOR_BODY]] ] ; CHECK-INTERLEAVED-NEXT: [[VEC_PHI1:%.*]] = phi <4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[PARTIAL_REDUCE3:%.*]], [[VECTOR_BODY]] ] +; CHECK-INTERLEAVED-NEXT: [[VEC_PHI2:%.*]] = phi <4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[PARTIAL_REDUCE8:%.*]], [[VECTOR_BODY]] ] +; CHECK-INTERLEAVED-NEXT: [[VEC_PHI3:%.*]] = phi <4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[PARTIAL_REDUCE9:%.*]], [[VECTOR_BODY]] ] ; CHECK-INTERLEAVED-NEXT: [[TMP1:%.*]] = getelementptr i8, ptr [[A]], i64 [[INDEX]] ; CHECK-INTERLEAVED-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[TMP1]], i32 16 +; CHECK-INTERLEAVED-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[TMP1]], i32 32 +; CHECK-INTERLEAVED-NEXT: [[TMP8:%.*]] = getelementptr i8, ptr [[TMP1]], i32 48 ; CHECK-INTERLEAVED-NEXT: [[WIDE_LOAD:%.*]] = load <16 x i8>, ptr [[TMP1]], align 1 ; CHECK-INTERLEAVED-NEXT: [[WIDE_LOAD2:%.*]] = load <16 x i8>, ptr [[TMP3]], align 1 +; CHECK-INTERLEAVED-NEXT: [[WIDE_LOAD5:%.*]] = load <16 x i8>, ptr [[TMP2]], align 1 +; CHECK-INTERLEAVED-NEXT: [[WIDE_LOAD6:%.*]] = load <16 x i8>, ptr [[TMP8]], align 1 ; CHECK-INTERLEAVED-NEXT: [[TMP4:%.*]] = zext <16 x i8> [[WIDE_LOAD]] to <16 x i32> ; CHECK-INTERLEAVED-NEXT: [[TMP5:%.*]] = zext <16 x i8> [[WIDE_LOAD2]] to <16 x i32> +; CHECK-INTERLEAVED-NEXT: [[TMP10:%.*]] = zext <16 x i8> [[WIDE_LOAD5]] to <16 x i32> +; CHECK-INTERLEAVED-NEXT: [[TMP7:%.*]] = zext <16 x i8> [[WIDE_LOAD6]] to <16 x i32> ; CHECK-INTERLEAVED-NEXT: [[PARTIAL_REDUCE]] = call <4 x i32> @llvm.vector.partial.reduce.add.v4i32.v16i32(<4 x i32> [[VEC_PHI]], <16 x i32> [[TMP4]]) ; CHECK-INTERLEAVED-NEXT: [[PARTIAL_REDUCE3]] = call <4 x i32> @llvm.vector.partial.reduce.add.v4i32.v16i32(<4 x i32> [[VEC_PHI1]], <16 x i32> [[TMP5]]) -; CHECK-INTERLEAVED-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 32 +; CHECK-INTERLEAVED-NEXT: [[PARTIAL_REDUCE8]] = call <4 x i32> @llvm.vector.partial.reduce.add.v4i32.v16i32(<4 x i32> [[VEC_PHI2]], <16 x i32> [[TMP10]]) +; CHECK-INTERLEAVED-NEXT: [[PARTIAL_REDUCE9]] = call <4 x i32> @llvm.vector.partial.reduce.add.v4i32.v16i32(<4 x i32> [[VEC_PHI3]], <16 x i32> [[TMP7]]) +; CHECK-INTERLEAVED-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 64 ; CHECK-INTERLEAVED-NEXT: [[TMP6:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; CHECK-INTERLEAVED-NEXT: br i1 [[TMP6]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; CHECK-INTERLEAVED: middle.block: ; CHECK-INTERLEAVED-NEXT: [[BIN_RDX:%.*]] = add <4 x i32> [[PARTIAL_REDUCE3]], [[PARTIAL_REDUCE]] -; CHECK-INTERLEAVED-NEXT: [[TMP7:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX]]) +; CHECK-INTERLEAVED-NEXT: [[BIN_RDX10:%.*]] = add <4 x i32> [[PARTIAL_REDUCE8]], [[BIN_RDX]] +; CHECK-INTERLEAVED-NEXT: [[BIN_RDX11:%.*]] = add <4 x i32> [[PARTIAL_REDUCE9]], [[BIN_RDX10]] +; CHECK-INTERLEAVED-NEXT: [[TMP9:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX11]]) ; CHECK-INTERLEAVED-NEXT: br label [[SCALAR_PH:%.*]] ; CHECK-INTERLEAVED: scalar.ph: ; @@ -432,25 +444,20 @@ define i32 @zext_add_reduc_i8_i32_has_neon_dotprod(ptr %a) #1 { ; CHECK-MAXBW-NEXT: entry: ; CHECK-MAXBW-NEXT: br label [[VECTOR_PH:%.*]] ; CHECK-MAXBW: vector.ph: -; CHECK-MAXBW-NEXT: [[TMP2:%.*]] = call i64 @llvm.vscale.i64() -; CHECK-MAXBW-NEXT: [[TMP3:%.*]] = mul nuw i64 [[TMP2]], 16 -; CHECK-MAXBW-NEXT: [[N_MOD_VF:%.*]] = urem i64 1025, [[TMP3]] -; CHECK-MAXBW-NEXT: [[N_VEC:%.*]] = sub i64 1025, [[N_MOD_VF]] ; CHECK-MAXBW-NEXT: br label [[VECTOR_BODY:%.*]] ; CHECK-MAXBW: vector.body: ; CHECK-MAXBW-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] -; CHECK-MAXBW-NEXT: [[VEC_PHI:%.*]] = phi <vscale x 4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[PARTIAL_REDUCE:%.*]], [[VECTOR_BODY]] ] +; CHECK-MAXBW-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[PARTIAL_REDUCE:%.*]], [[VECTOR_BODY]] ] ; CHECK-MAXBW-NEXT: [[TMP7:%.*]] = getelementptr i8, ptr [[A]], i64 [[INDEX]] -; CHECK-MAXBW-NEXT: [[WIDE_LOAD:%.*]] = load <vscale x 16 x i8>, ptr [[TMP7]], align 1 -; CHECK-MAXBW-NEXT: [[TMP9:%.*]] = zext <vscale x 16 x i8> [[WIDE_LOAD]] to <vscale x 16 x i32> -; CHECK-MAXBW-NEXT: [[PARTIAL_REDUCE]] = call <vscale x 4 x i32> @llvm.vector.partial.reduce.add.nxv4i32.nxv16i32(<vscale x 4 x i32> [[VEC_PHI]], <vscale x 16 x i32> [[TMP9]]) -; CHECK-MAXBW-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], [[TMP3]] -; CHECK-MAXBW-NEXT: [[TMP10:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] -; CHECK-MAXBW-NEXT: br i1 [[TMP10]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] +; CHECK-MAXBW-NEXT: [[WIDE_LOAD:%.*]] = load <16 x i8>, ptr [[TMP7]], align 1 +; CHECK-MAXBW-NEXT: [[TMP1:%.*]] = zext <16 x i8> [[WIDE_LOAD]] to <16 x i32> +; CHECK-MAXBW-NEXT: [[PARTIAL_REDUCE]] = call <4 x i32> @llvm.vector.partial.reduce.add.v4i32.v16i32(<4 x i32> [[VEC_PHI]], <16 x i32> [[TMP1]]) +; CHECK-MAXBW-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 16 +; CHECK-MAXBW-NEXT: [[TMP2:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 +; CHECK-MAXBW-NEXT: br i1 [[TMP2]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] ; CHECK-MAXBW: middle.block: -; CHECK-MAXBW-NEXT: [[TMP11:%.*]] = call i32 @llvm.vector.reduce.add.nxv4i32(<vscale x 4 x i32> [[PARTIAL_REDUCE]]) -; CHECK-MAXBW-NEXT: [[CMP_N:%.*]] = icmp eq i64 1025, [[N_VEC]] -; CHECK-MAXBW-NEXT: br i1 [[CMP_N]], label [[FOR_EXIT:%.*]], label [[SCALAR_PH:%.*]] +; CHECK-MAXBW-NEXT: [[TMP3:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[PARTIAL_REDUCE]]) +; CHECK-MAXBW-NEXT: br label [[SCALAR_PH:%.*]] ; CHECK-MAXBW: scalar.ph: ; entry: @@ -693,20 +700,32 @@ define i32 @zext_sub_reduc_i8_i32_has_neon_dotprod(ptr %a) #1 { ; CHECK-INTERLEAVED-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] ; CHECK-INTERLEAVED-NEXT: [[VEC_PHI:%.*]] = phi <16 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP6:%.*]], [[VECTOR_BODY]] ] ; CHECK-INTERLEAVED-NEXT: [[VEC_PHI1:%.*]] = phi <16 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP7:%.*]], [[VECTOR_BODY]] ] +; CHECK-INTERLEAVED-NEXT: [[VEC_PHI2:%.*]] = phi <16 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP10:%.*]], [[VECTOR_BODY]] ] +; CHECK-INTERLEAVED-NEXT: [[VEC_PHI3:%.*]] = phi <16 x i32> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP11:%.*]], [[VECTOR_BODY]] ] ; CHECK-INTERLEAVED-NEXT: [[TMP1:%.*]] = getelementptr i8, ptr [[A]], i64 [[INDEX]] ; CHECK-INTERLEAVED-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[TMP1]], i32 16 +; CHECK-INTERLEAVED-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[TMP1]], i32 32 +; CHECK-INTERLEAVED-NEXT: [[TMP9:%.*]] = getelementptr i8, ptr [[TMP1]], i32 48 ; CHECK-INTERLEAVED-NEXT: [[WIDE_LOAD:%.*]] = load <16 x i8>, ptr [[TMP1]], align 1 ; CHECK-INTERLEAVED-NEXT: [[WIDE_LOAD2:%.*]] = load <16 x i8>, ptr [[TMP3]], align 1 +; CHECK-INTERLEAVED-NEXT: [[WIDE_LOAD5:%.*]] = load <16 x i8>, ptr [[TMP2]], align 1 +; CHECK-INTERLEAVED-NEXT: [[WIDE_LOAD6:%.*]] = load <16 x i8>, ptr [[TMP9]], align 1 ; CHECK-INTERLEAVED-NEXT: [[TMP4:%.*]] = zext <16 x i8> [[WIDE_LOAD]] to <16 x i32> ; CHECK-INTERLEAVED-NEXT: [[TMP5:%.*]] = zext <16 x i8> [[WIDE_LOAD2]] to <16 x i32> +; CHECK-INTERLEAVED-NEXT: [[TMP12:%.*]] = zext <16 x i8> [[WIDE_LOAD5]] to <16 x i32> +; CHECK-INTERLEAVED-NEXT: [[TMP14:%.*]] = zext <16 x i8> [[WIDE_LOAD6]] to <16 x i32> ; CHECK-INTERLEAVED-NEXT: [[TMP6]] = sub <16 x i32> [[VEC_PHI]], [[TMP4]] ; CHECK-INTERLEAVED-NEXT: [[TMP7]] = sub <16 x i32> [[VEC_PHI1]], [[TMP5]] -; CHECK-INTERLEAVED-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 32 +; CHECK-INTERLEAVED-NEXT: [[TMP10]] = sub <16 x i32> [[VEC_PHI2]], [[TMP12]] +; CHECK-INTERLEAVED-NEXT: [[TMP11]] = sub <16 x i32> [[VEC_PHI3]], [[TMP14]] +; CHECK-INTERLEAVED-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 64 ; CHECK-INTERLEAVED-NEXT: [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], 1024 ; CHECK-INTERLEAVED-NEXT: br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP13:![0-9]+]] ; CHECK-INTERLEAVED: middle.block: ; CHECK-INTERLEAVED-NEXT: [[BIN_RDX:%.*]] = add <16 x i32> [[TMP7]], [[TMP6]] -; CHECK-INTERLEAVED-NEXT: [[TMP9:%.*]] = call i32 @llvm.vector.reduce.add.v16i32(<16 x i32> [[BIN_RDX]]) +; CHECK-INTERLEAVED-NEXT: [[BIN_RDX7:%.*]] = add <16 x i32> [[TMP10]], [[BIN_RDX]] +; CHECK-INTERLEAVED-NEXT: [[BIN_RDX8:%.*]] = add <16 x i32> [[TMP11]], [[BIN_RDX7]] +; CHECK-INTERLEAVED-NEXT: [[TMP13:%.*]] = call i32 @llvm.vector.reduce.add.v16i32(<16 x i32> [[BIN_RDX8]]) ; CHECK-INTERLEAVED-NEXT: br label [[SCALAR_PH:%.*]] ; CHECK-INTERLEAVED: scalar.ph: ; @@ -1093,9 +1112,124 @@ exit: ret i32 %add.lcssa } +define i64 @sext_reduction_i32_to_i64(ptr %arr, i64 %n) #1 { +; CHECK-INTERLEAVE1-LABEL: define i64 @sext_reduction_i32_to_i64( +; CHECK-INTERLEAVE1-SAME: ptr [[ARR:%.*]], i64 [[N:%.*]]) #[[ATTR2]] { +; CHECK-INTERLEAVE1-NEXT: entry: +; CHECK-INTERLEAVE1-NEXT: [[UMAX:%.*]] = call i64 @llvm.umax.i64(i64 [[N]], i64 1) +; CHECK-INTERLEAVE1-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[UMAX]], 2 +; CHECK-INTERLEAVE1-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]] +; CHECK-INTERLEAVE1: vector.ph: +; CHECK-INTERLEAVE1-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[UMAX]], 2 +; CHECK-INTERLEAVE1-NEXT: [[N_VEC:%.*]] = sub i64 [[UMAX]], [[N_MOD_VF]] +; CHECK-INTERLEAVE1-NEXT: br label [[VECTOR_BODY:%.*]] +; CHECK-INTERLEAVE1: vector.body: +; CHECK-INTERLEAVE1-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] +; CHECK-INTERLEAVE1-NEXT: [[VEC_PHI:%.*]] = phi <2 x i64> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP2:%.*]], [[VECTOR_BODY]] ] +; CHECK-INTERLEAVE1-NEXT: [[TMP4:%.*]] = getelementptr inbounds i32, ptr [[ARR]], i64 [[INDEX]] +; CHECK-INTERLEAVE1-NEXT: [[WIDE_LOAD:%.*]] = load <2 x i32>, ptr [[TMP4]], align 4 +; CHECK-INTERLEAVE1-NEXT: [[TMP1:%.*]] = sext <2 x i32> [[WIDE_LOAD]] to <2 x i64> +; CHECK-INTERLEAVE1-NEXT: [[TMP2]] = add <2 x i64> [[VEC_PHI]], [[TMP1]] +; CHECK-INTERLEAVE1-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 +; CHECK-INTERLEAVE1-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] +; CHECK-INTERLEAVE1-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP21:![0-9]+]] +; CHECK-INTERLEAVE1: middle.block: +; CHECK-INTERLEAVE1-NEXT: [[TMP5:%.*]] = call i64 @llvm.vector.reduce.add.v2i64(<2 x i64> [[TMP2]]) +; CHECK-INTERLEAVE1-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[UMAX]], [[N_VEC]] +; CHECK-INTERLEAVE1-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-INTERLEAVE1: scalar.ph: +; +; CHECK-INTERLEAVED-LABEL: define i64 @sext_reduction_i32_to_i64( +; CHECK-INTERLEAVED-SAME: ptr [[ARR:%.*]], i64 [[N:%.*]]) #[[ATTR2]] { +; CHECK-INTERLEAVED-NEXT: entry: +; CHECK-INTERLEAVED-NEXT: [[UMAX:%.*]] = call i64 @llvm.umax.i64(i64 [[N]], i64 1) +; CHECK-INTERLEAVED-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[UMAX]], 8 +; CHECK-INTERLEAVED-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]] +; CHECK-INTERLEAVED: vector.ph: +; CHECK-INTERLEAVED-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[UMAX]], 8 +; CHECK-INTERLEAVED-NEXT: [[N_VEC:%.*]] = sub i64 [[UMAX]], [[N_MOD_VF]] +; CHECK-INTERLEAVED-NEXT: br label [[VECTOR_BODY:%.*]] +; CHECK-INTERLEAVED: vector.body: +; CHECK-INTERLEAVED-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] +; CHECK-INTERLEAVED-NEXT: [[VEC_PHI:%.*]] = phi <2 x i64> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP8:%.*]], [[VECTOR_BODY]] ] +; CHECK-INTERLEAVED-NEXT: [[VEC_PHI1:%.*]] = phi <2 x i64> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP9:%.*]], [[VECTOR_BODY]] ] +; CHECK-INTERLEAVED-NEXT: [[VEC_PHI2:%.*]] = phi <2 x i64> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP10:%.*]], [[VECTOR_BODY]] ] +; CHECK-INTERLEAVED-NEXT: [[VEC_PHI3:%.*]] = phi <2 x i64> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP11:%.*]], [[VECTOR_BODY]] ] +; CHECK-INTERLEAVED-NEXT: [[TMP4:%.*]] = getelementptr inbounds i32, ptr [[ARR]], i64 [[INDEX]] +; CHECK-INTERLEAVED-NEXT: [[TMP1:%.*]] = getelementptr inbounds i32, ptr [[TMP4]], i32 2 +; CHECK-INTERLEAVED-NEXT: [[TMP2:%.*]] = getelementptr inbounds i32, ptr [[TMP4]], i32 4 +; CHECK-INTERLEAVED-NEXT: [[TMP3:%.*]] = getelementptr inbounds i32, ptr [[TMP4]], i32 6 +; CHECK-INTERLEAVED-NEXT: [[WIDE_LOAD:%.*]] = load <2 x i32>, ptr [[TMP4]], align 4 +; CHECK-INTERLEAVED-NEXT: [[WIDE_LOAD4:%.*]] = load <2 x i32>, ptr [[TMP1]], align 4 +; CHECK-INTERLEAVED-NEXT: [[WIDE_LOAD5:%.*]] = load <2 x i32>, ptr [[TMP2]], align 4 +; CHECK-INTERLEAVED-NEXT: [[WIDE_LOAD6:%.*]] = load <2 x i32>, ptr [[TMP3]], align 4 +; CHECK-INTERLEAVED-NEXT: [[TMP14:%.*]] = sext <2 x i32> [[WIDE_LOAD]] to <2 x i64> +; CHECK-INTERLEAVED-NEXT: [[TMP5:%.*]] = sext <2 x i32> [[WIDE_LOAD4]] to <2 x i64> +; CHECK-INTERLEAVED-NEXT: [[TMP6:%.*]] = sext <2 x i32> [[WIDE_LOAD5]] to <2 x i64> +; CHECK-INTERLEAVED-NEXT: [[TMP7:%.*]] = sext <2 x i32> [[WIDE_LOAD6]] to <2 x i64> +; CHECK-INTERLEAVED-NEXT: [[TMP8]] = add <2 x i64> [[VEC_PHI]], [[TMP14]] +; CHECK-INTERLEAVED-NEXT: [[TMP9]] = add <2 x i64> [[VEC_PHI1]], [[TMP5]] +; CHECK-INTERLEAVED-NEXT: [[TMP10]] = add <2 x i64> [[VEC_PHI2]], [[TMP6]] +; CHECK-INTERLEAVED-NEXT: [[TMP11]] = add <2 x i64> [[VEC_PHI3]], [[TMP7]] +; CHECK-INTERLEAVED-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 8 +; CHECK-INTERLEAVED-NEXT: [[TMP12:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] +; CHECK-INTERLEAVED-NEXT: br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP21:![0-9]+]] +; CHECK-INTERLEAVED: middle.block: +; CHECK-INTERLEAVED-NEXT: [[BIN_RDX:%.*]] = add <2 x i64> [[TMP9]], [[TMP8]] +; CHECK-INTERLEAVED-NEXT: [[BIN_RDX7:%.*]] = add <2 x i64> [[TMP10]], [[BIN_RDX]] +; CHECK-INTERLEAVED-NEXT: [[BIN_RDX8:%.*]] = add <2 x i64> [[TMP11]], [[BIN_RDX7]] +; CHECK-INTERLEAVED-NEXT: [[TMP13:%.*]] = call i64 @llvm.vector.reduce.add.v2i64(<2 x i64> [[BIN_RDX8]]) +; CHECK-INTERLEAVED-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[UMAX]], [[N_VEC]] +; CHECK-INTERLEAVED-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-INTERLEAVED: scalar.ph: +; +; CHECK-MAXBW-LABEL: define i64 @sext_reduction_i32_to_i64( +; CHECK-MAXBW-SAME: ptr [[ARR:%.*]], i64 [[N:%.*]]) #[[ATTR2]] { +; CHECK-MAXBW-NEXT: entry: +; CHECK-MAXBW-NEXT: [[UMAX:%.*]] = call i64 @llvm.umax.i64(i64 [[N]], i64 1) +; CHECK-MAXBW-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[UMAX]], 2 +; CHECK-MAXBW-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]] +; CHECK-MAXBW: vector.ph: +; CHECK-MAXBW-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[UMAX]], 2 +; CHECK-MAXBW-NEXT: [[N_VEC:%.*]] = sub i64 [[UMAX]], [[N_MOD_VF]] +; CHECK-MAXBW-NEXT: br label [[VECTOR_BODY:%.*]] +; CHECK-MAXBW: vector.body: +; CHECK-MAXBW-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ] +; CHECK-MAXBW-NEXT: [[VEC_PHI:%.*]] = phi <2 x i64> [ zeroinitializer, [[VECTOR_PH]] ], [ [[TMP2:%.*]], [[VECTOR_BODY]] ] +; CHECK-MAXBW-NEXT: [[TMP4:%.*]] = getelementptr inbounds i32, ptr [[ARR]], i64 [[INDEX]] +; CHECK-MAXBW-NEXT: [[WIDE_LOAD:%.*]] = load <2 x i32>, ptr [[TMP4]], align 4 +; CHECK-MAXBW-NEXT: [[TMP1:%.*]] = sext <2 x i32> [[WIDE_LOAD]] to <2 x i64> +; CHECK-MAXBW-NEXT: [[TMP2]] = add <2 x i64> [[VEC_PHI]], [[TMP1]] +; CHECK-MAXBW-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2 +; CHECK-MAXBW-NEXT: [[TMP7:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] +; CHECK-MAXBW-NEXT: br i1 [[TMP7]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP21:![0-9]+]] +; CHECK-MAXBW: middle.block: +; CHECK-MAXBW-NEXT: [[TMP5:%.*]] = call i64 @llvm.vector.reduce.add.v2i64(<2 x i64> [[TMP2]]) +; CHECK-MAXBW-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[UMAX]], [[N_VEC]] +; CHECK-MAXBW-NEXT: br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]] +; CHECK-MAXBW: scalar.ph: +; +entry: + br label %loop + +loop: + %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ] + %acc = phi i64 [ 0, %entry ], [ %add, %loop ] + %gep = getelementptr inbounds i32, ptr %arr, i64 %iv + %load = load i32, ptr %gep + %sext = sext i32 %load to i64 + %add = add i64 %acc, %sext + %iv.next = add i64 %iv, 1 + %cmp = icmp ult i64 %iv.next, %n + br i1 %cmp, label %loop, label %exit + +exit: + ret i64 %add +} + !0 = distinct !{!0, !1} !1 = !{!"llvm.loop.vectorize.predicate.enable", i1 true} attributes #0 = { vscale_range(1,16) "target-features"="+sve" } -attributes #1 = { vscale_range(1,16) "target-features"="+neon,+dotprod,+sve" "cpu"="neoverse-v2" } +attributes #1 = { vscale_range(1,16) "target-features"="+neon,+dotprod,+sve" "target-cpu"="neoverse-v2" } attributes #2 = { "target-features"="+neon,+dotprod" } diff --git a/llvm/test/Transforms/LoopVectorize/AArch64/replicating-load-store-costs.ll b/llvm/test/Transforms/LoopVectorize/AArch64/replicating-load-store-costs.ll index ab9b48f..aff2c4c 100644 --- a/llvm/test/Transforms/LoopVectorize/AArch64/replicating-load-store-costs.ll +++ b/llvm/test/Transforms/LoopVectorize/AArch64/replicating-load-store-costs.ll @@ -153,17 +153,20 @@ define void @uniform_gep_for_replicating_gep(ptr %dst) { ; CHECK-NEXT: [[VEC_IND:%.*]] = phi <2 x i32> [ <i32 0, i32 1>, %[[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], %[[VECTOR_BODY]] ] ; CHECK-NEXT: [[STEP_ADD:%.*]] = add <2 x i32> [[VEC_IND]], splat (i32 2) ; CHECK-NEXT: [[TMP2:%.*]] = add i32 [[INDEX]], 2 -; CHECK-NEXT: [[TMP5:%.*]] = icmp eq <2 x i32> [[STEP_ADD]], zeroinitializer +; CHECK-NEXT: [[TMP5:%.*]] = icmp eq <2 x i32> [[VEC_IND]], zeroinitializer +; CHECK-NEXT: [[TMP3:%.*]] = icmp eq <2 x i32> [[STEP_ADD]], zeroinitializer ; CHECK-NEXT: [[TMP8:%.*]] = lshr i32 [[INDEX]], 1 ; CHECK-NEXT: [[TMP9:%.*]] = lshr i32 [[TMP2]], 1 ; CHECK-NEXT: [[TMP11:%.*]] = zext <2 x i1> [[TMP5]] to <2 x i8> +; CHECK-NEXT: [[TMP6:%.*]] = zext <2 x i1> [[TMP3]] to <2 x i8> ; CHECK-NEXT: [[TMP14:%.*]] = zext i32 [[TMP8]] to i64 ; CHECK-NEXT: [[TMP15:%.*]] = zext i32 [[TMP9]] to i64 ; CHECK-NEXT: [[TMP18:%.*]] = getelementptr i64, ptr [[DST]], i64 [[TMP14]] ; CHECK-NEXT: [[TMP19:%.*]] = getelementptr i64, ptr [[DST]], i64 [[TMP15]] ; CHECK-NEXT: [[TMP22:%.*]] = extractelement <2 x i8> [[TMP11]], i32 1 +; CHECK-NEXT: [[TMP12:%.*]] = extractelement <2 x i8> [[TMP6]], i32 1 ; CHECK-NEXT: store i8 [[TMP22]], ptr [[TMP18]], align 1 -; CHECK-NEXT: store i8 [[TMP22]], ptr [[TMP19]], align 1 +; CHECK-NEXT: store i8 [[TMP12]], ptr [[TMP19]], align 1 ; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4 ; CHECK-NEXT: [[VEC_IND_NEXT]] = add <2 x i32> [[STEP_ADD]], splat (i32 2) ; CHECK-NEXT: [[TMP24:%.*]] = icmp eq i32 [[INDEX_NEXT]], 128 diff --git a/llvm/test/Transforms/LoopVectorize/narrow-to-single-scalar.ll b/llvm/test/Transforms/LoopVectorize/narrow-to-single-scalar.ll index 1533906..53dad3a 100644 --- a/llvm/test/Transforms/LoopVectorize/narrow-to-single-scalar.ll +++ b/llvm/test/Transforms/LoopVectorize/narrow-to-single-scalar.ll @@ -74,8 +74,7 @@ exit: ret void } -; FIXME: Currently this mis-compiled when interleaving; all stores store the -; last lane of the last part, instead of the last lane per part. +; Check each unrolled store stores the last lane of the corresponding part. ; Test case for https://github.com/llvm/llvm-project/issues/162498. define void @narrow_to_single_scalar_store_address_not_uniform_across_all_parts(ptr %dst) { ; VF4IC1-LABEL: define void @narrow_to_single_scalar_store_address_not_uniform_across_all_parts( @@ -121,13 +120,15 @@ define void @narrow_to_single_scalar_store_address_not_uniform_across_all_parts( ; VF2IC2-NEXT: br label %[[VECTOR_BODY:.*]] ; VF2IC2: [[VECTOR_BODY]]: ; VF2IC2-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; VF2IC2-NEXT: [[TMP7:%.*]] = add i32 [[INDEX]], 0 +; VF2IC2-NEXT: [[TMP8:%.*]] = add i32 [[INDEX]], 1 ; VF2IC2-NEXT: [[TMP0:%.*]] = add i32 [[INDEX]], 2 ; VF2IC2-NEXT: [[TMP1:%.*]] = add i32 [[INDEX]], 3 -; VF2IC2-NEXT: [[TMP2:%.*]] = lshr i32 [[INDEX]], 1 +; VF2IC2-NEXT: [[TMP2:%.*]] = lshr i32 [[TMP7]], 1 ; VF2IC2-NEXT: [[TMP3:%.*]] = lshr i32 [[TMP0]], 1 ; VF2IC2-NEXT: [[TMP4:%.*]] = getelementptr i32, ptr [[DST]], i32 [[TMP2]] ; VF2IC2-NEXT: [[TMP5:%.*]] = getelementptr i32, ptr [[DST]], i32 [[TMP3]] -; VF2IC2-NEXT: store i32 [[TMP1]], ptr [[TMP4]], align 4 +; VF2IC2-NEXT: store i32 [[TMP8]], ptr [[TMP4]], align 4 ; VF2IC2-NEXT: store i32 [[TMP1]], ptr [[TMP5]], align 4 ; VF2IC2-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4 ; VF2IC2-NEXT: [[TMP6:%.*]] = icmp eq i32 [[INDEX_NEXT]], 100 diff --git a/llvm/test/Transforms/LoopVectorize/reduction-minmax-users-and-predicated.ll b/llvm/test/Transforms/LoopVectorize/reduction-minmax-users-and-predicated.ll new file mode 100644 index 0000000..e4322cf --- /dev/null +++ b/llvm/test/Transforms/LoopVectorize/reduction-minmax-users-and-predicated.ll @@ -0,0 +1,588 @@ +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals none --version 6 +; RUN: opt -p loop-vectorize -force-vector-width=4 -S %s | FileCheck %s + +define i32 @umax_phi_used_outside(ptr %src, i32 %n) { +; CHECK-LABEL: define i32 @umax_phi_used_outside( +; CHECK-SAME: ptr [[SRC:%.*]], i32 [[N:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: br label %[[LOOP:.*]] +; CHECK: [[LOOP]]: +; CHECK-NEXT: [[IV:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[MAX:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[SPEC_SELECT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[GEP_SRC:%.*]] = getelementptr inbounds i8, ptr [[SRC]], i32 [[IV]] +; CHECK-NEXT: [[L:%.*]] = load i8, ptr [[GEP_SRC]], align 1 +; CHECK-NEXT: [[L_EXT:%.*]] = zext i8 [[L]] to i32 +; CHECK-NEXT: [[SPEC_SELECT]] = tail call i32 @llvm.umax.i32(i32 [[MAX]], i32 [[L_EXT]]) +; CHECK-NEXT: [[IV_NEXT]] = add i32 [[IV]], 1 +; CHECK-NEXT: [[EC:%.*]] = icmp eq i32 [[IV]], [[N]] +; CHECK-NEXT: br i1 [[EC]], label %[[EXIT:.*]], label %[[LOOP]] +; CHECK: [[EXIT]]: +; CHECK-NEXT: [[MAX_LCSSA:%.*]] = phi i32 [ [[MAX]], %[[LOOP]] ] +; CHECK-NEXT: ret i32 [[MAX_LCSSA]] +; +entry: + br label %loop + +loop: + %iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ] + %max = phi i32 [ 0, %entry ], [ %spec.select, %loop ] + %gep.src = getelementptr inbounds i8, ptr %src, i32 %iv + %l = load i8, ptr %gep.src + %l.ext = zext i8 %l to i32 + %spec.select = tail call i32 @llvm.umax.i32(i32 %max, i32 %l.ext) + %iv.next = add i32 %iv, 1 + %ec = icmp eq i32 %iv, %n + br i1 %ec, label %exit, label %loop + +exit: + ret i32 %max +} + +define i32 @chained_smax(i32 %x, ptr %src) { +; CHECK-LABEL: define i32 @chained_smax( +; CHECK-SAME: i32 [[X:%.*]], ptr [[SRC:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: br label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[X]], i64 0 +; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[PRED_LOAD_CONTINUE6:.*]] ] +; CHECK-NEXT: [[VEC_IND:%.*]] = phi <4 x i8> [ <i8 0, i8 1, i8 2, i8 3>, %[[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], %[[PRED_LOAD_CONTINUE6]] ] +; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP26:%.*]], %[[PRED_LOAD_CONTINUE6]] ] +; CHECK-NEXT: [[TMP0:%.*]] = icmp ule <4 x i8> [[VEC_IND]], splat (i8 1) +; CHECK-NEXT: [[TMP1:%.*]] = call <4 x i32> @llvm.smax.v4i32(<4 x i32> [[BROADCAST_SPLAT]], <4 x i32> [[VEC_PHI]]) +; CHECK-NEXT: [[TMP2:%.*]] = extractelement <4 x i1> [[TMP0]], i32 0 +; CHECK-NEXT: br i1 [[TMP2]], label %[[PRED_LOAD_IF:.*]], label %[[PRED_LOAD_CONTINUE:.*]] +; CHECK: [[PRED_LOAD_IF]]: +; CHECK-NEXT: [[TMP3:%.*]] = add i64 [[INDEX]], 0 +; CHECK-NEXT: [[TMP4:%.*]] = getelementptr [3 x i32], ptr [[SRC]], i64 [[TMP3]] +; CHECK-NEXT: [[TMP5:%.*]] = load i32, ptr [[TMP4]], align 4 +; CHECK-NEXT: [[TMP6:%.*]] = insertelement <4 x i32> poison, i32 [[TMP5]], i32 0 +; CHECK-NEXT: br label %[[PRED_LOAD_CONTINUE]] +; CHECK: [[PRED_LOAD_CONTINUE]]: +; CHECK-NEXT: [[TMP7:%.*]] = phi <4 x i32> [ poison, %[[VECTOR_BODY]] ], [ [[TMP6]], %[[PRED_LOAD_IF]] ] +; CHECK-NEXT: [[TMP8:%.*]] = extractelement <4 x i1> [[TMP0]], i32 1 +; CHECK-NEXT: br i1 [[TMP8]], label %[[PRED_LOAD_IF1:.*]], label %[[PRED_LOAD_CONTINUE2:.*]] +; CHECK: [[PRED_LOAD_IF1]]: +; CHECK-NEXT: [[TMP9:%.*]] = add i64 [[INDEX]], 1 +; CHECK-NEXT: [[TMP10:%.*]] = getelementptr [3 x i32], ptr [[SRC]], i64 [[TMP9]] +; CHECK-NEXT: [[TMP11:%.*]] = load i32, ptr [[TMP10]], align 4 +; CHECK-NEXT: [[TMP12:%.*]] = insertelement <4 x i32> [[TMP7]], i32 [[TMP11]], i32 1 +; CHECK-NEXT: br label %[[PRED_LOAD_CONTINUE2]] +; CHECK: [[PRED_LOAD_CONTINUE2]]: +; CHECK-NEXT: [[TMP13:%.*]] = phi <4 x i32> [ [[TMP7]], %[[PRED_LOAD_CONTINUE]] ], [ [[TMP12]], %[[PRED_LOAD_IF1]] ] +; CHECK-NEXT: [[TMP14:%.*]] = extractelement <4 x i1> [[TMP0]], i32 2 +; CHECK-NEXT: br i1 [[TMP14]], label %[[PRED_LOAD_IF3:.*]], label %[[PRED_LOAD_CONTINUE4:.*]] +; CHECK: [[PRED_LOAD_IF3]]: +; CHECK-NEXT: [[TMP15:%.*]] = add i64 [[INDEX]], 2 +; CHECK-NEXT: [[TMP16:%.*]] = getelementptr [3 x i32], ptr [[SRC]], i64 [[TMP15]] +; CHECK-NEXT: [[TMP17:%.*]] = load i32, ptr [[TMP16]], align 4 +; CHECK-NEXT: [[TMP18:%.*]] = insertelement <4 x i32> [[TMP13]], i32 [[TMP17]], i32 2 +; CHECK-NEXT: br label %[[PRED_LOAD_CONTINUE4]] +; CHECK: [[PRED_LOAD_CONTINUE4]]: +; CHECK-NEXT: [[TMP19:%.*]] = phi <4 x i32> [ [[TMP13]], %[[PRED_LOAD_CONTINUE2]] ], [ [[TMP18]], %[[PRED_LOAD_IF3]] ] +; CHECK-NEXT: [[TMP20:%.*]] = extractelement <4 x i1> [[TMP0]], i32 3 +; CHECK-NEXT: br i1 [[TMP20]], label %[[PRED_LOAD_IF5:.*]], label %[[PRED_LOAD_CONTINUE6]] +; CHECK: [[PRED_LOAD_IF5]]: +; CHECK-NEXT: [[TMP21:%.*]] = add i64 [[INDEX]], 3 +; CHECK-NEXT: [[TMP22:%.*]] = getelementptr [3 x i32], ptr [[SRC]], i64 [[TMP21]] +; CHECK-NEXT: [[TMP23:%.*]] = load i32, ptr [[TMP22]], align 4 +; CHECK-NEXT: [[TMP24:%.*]] = insertelement <4 x i32> [[TMP19]], i32 [[TMP23]], i32 3 +; CHECK-NEXT: br label %[[PRED_LOAD_CONTINUE6]] +; CHECK: [[PRED_LOAD_CONTINUE6]]: +; CHECK-NEXT: [[TMP25:%.*]] = phi <4 x i32> [ [[TMP19]], %[[PRED_LOAD_CONTINUE4]] ], [ [[TMP24]], %[[PRED_LOAD_IF5]] ] +; CHECK-NEXT: [[TMP26]] = call <4 x i32> @llvm.smax.v4i32(<4 x i32> [[TMP25]], <4 x i32> [[TMP1]]) +; CHECK-NEXT: [[TMP27:%.*]] = select <4 x i1> [[TMP0]], <4 x i32> [[TMP26]], <4 x i32> [[VEC_PHI]] +; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4 +; CHECK-NEXT: [[VEC_IND_NEXT]] = add <4 x i8> [[VEC_IND]], splat (i8 4) +; CHECK-NEXT: br i1 true, label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] +; CHECK: [[MIDDLE_BLOCK]]: +; CHECK-NEXT: [[TMP28:%.*]] = call i32 @llvm.vector.reduce.smax.v4i32(<4 x i32> [[TMP27]]) +; CHECK-NEXT: br label %[[EXIT:.*]] +; CHECK: [[EXIT]]: +; CHECK-NEXT: ret i32 [[TMP28]] +; +entry: + br label %loop + +loop: ; preds = %loop, %entry + %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ] + %max = phi i32 [ 0, %entry ], [ %max.next, %loop ] + %gep.src = getelementptr [3 x i32], ptr %src, i64 %iv + %max.1 = tail call i32 @llvm.smax.i32(i32 %x, i32 %max) + %l = load i32, ptr %gep.src, align 4 + %max.next = tail call i32 @llvm.smax.i32(i32 %l, i32 %max.1) + %iv.next = add i64 %iv, 1 + %ec = icmp eq i64 %iv, 1 + br i1 %ec, label %exit, label %loop + +exit: + ret i32 %max.next +} + +define void @smax_with_invariant_store_user(ptr noalias %src, ptr %dst, i64 %n) { +; CHECK-LABEL: define void @smax_with_invariant_store_user( +; CHECK-SAME: ptr noalias [[SRC:%.*]], ptr [[DST:%.*]], i64 [[N:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[N]], 1 +; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP0]], 4 +; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[TMP0]], 4 +; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[TMP0]], [[N_MOD_VF]] +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i64 [[INDEX]] +; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x i32>, ptr [[TMP1]], align 4 +; CHECK-NEXT: [[TMP2]] = call <4 x i32> @llvm.smax.v4i32(<4 x i32> [[VEC_PHI]], <4 x i32> [[WIDE_LOAD]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4 +; CHECK-NEXT: [[TMP3:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] +; CHECK-NEXT: br i1 [[TMP3]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]] +; CHECK: [[MIDDLE_BLOCK]]: +; CHECK-NEXT: [[TMP4:%.*]] = call i32 @llvm.vector.reduce.smax.v4i32(<4 x i32> [[TMP2]]) +; CHECK-NEXT: store i32 [[TMP4]], ptr [[DST]], align 4 +; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[TMP0]], [[N_VEC]] +; CHECK-NEXT: br i1 [[CMP_N]], label %[[EXIT:.*]], label %[[SCALAR_PH]] +; CHECK: [[SCALAR_PH]]: +; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ] +; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ [[TMP4]], %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ] +; CHECK-NEXT: br label %[[LOOP:.*]] +; CHECK: [[LOOP]]: +; CHECK-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[MAX:%.*]] = phi i32 [ [[BC_MERGE_RDX]], %[[SCALAR_PH]] ], [ [[MAX_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[GEP_SRC:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i64 [[IV]] +; CHECK-NEXT: [[L:%.*]] = load i32, ptr [[GEP_SRC]], align 4 +; CHECK-NEXT: [[MAX_NEXT]] = tail call i32 @llvm.smax.i32(i32 [[MAX]], i32 [[L]]) +; CHECK-NEXT: store i32 [[MAX_NEXT]], ptr [[DST]], align 4 +; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], 1 +; CHECK-NEXT: [[EC:%.*]] = icmp eq i64 [[IV]], [[N]] +; CHECK-NEXT: br i1 [[EC]], label %[[EXIT]], label %[[LOOP]], !llvm.loop [[LOOP4:![0-9]+]] +; CHECK: [[EXIT]]: +; CHECK-NEXT: ret void +; +entry: + br label %loop + +loop: + %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ] + %max = phi i32 [ 0, %entry ], [ %max.next, %loop ] + %gep.src = getelementptr inbounds i32, ptr %src, i64 %iv + %l = load i32, ptr %gep.src, align 4 + %max.next = tail call i32 @llvm.smax.i32(i32 %max, i32 %l) + store i32 %max.next, ptr %dst, align 4 + %iv.next = add i64 %iv, 1 + %ec = icmp eq i64 %iv, %n + br i1 %ec, label %exit, label %loop + +exit: + ret void +} + +define void @smax_with_multiple_invariant_store_user_same_addr(ptr noalias %src, ptr %dst, i64 %n) { +; CHECK-LABEL: define void @smax_with_multiple_invariant_store_user_same_addr( +; CHECK-SAME: ptr noalias [[SRC:%.*]], ptr [[DST:%.*]], i64 [[N:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[N]], 1 +; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP0]], 4 +; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[TMP0]], 4 +; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[TMP0]], [[N_MOD_VF]] +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i64 [[INDEX]] +; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x i32>, ptr [[TMP1]], align 4 +; CHECK-NEXT: [[TMP2]] = call <4 x i32> @llvm.smax.v4i32(<4 x i32> [[VEC_PHI]], <4 x i32> [[WIDE_LOAD]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4 +; CHECK-NEXT: [[TMP3:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] +; CHECK-NEXT: br i1 [[TMP3]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]] +; CHECK: [[MIDDLE_BLOCK]]: +; CHECK-NEXT: [[TMP4:%.*]] = call i32 @llvm.vector.reduce.smax.v4i32(<4 x i32> [[TMP2]]) +; CHECK-NEXT: store i32 [[TMP4]], ptr [[DST]], align 4 +; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[TMP0]], [[N_VEC]] +; CHECK-NEXT: br i1 [[CMP_N]], label %[[EXIT:.*]], label %[[SCALAR_PH]] +; CHECK: [[SCALAR_PH]]: +; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ] +; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ [[TMP4]], %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ] +; CHECK-NEXT: br label %[[LOOP:.*]] +; CHECK: [[LOOP]]: +; CHECK-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[MAX:%.*]] = phi i32 [ [[BC_MERGE_RDX]], %[[SCALAR_PH]] ], [ [[MAX_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[GEP_SRC:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i64 [[IV]] +; CHECK-NEXT: [[L:%.*]] = load i32, ptr [[GEP_SRC]], align 4 +; CHECK-NEXT: [[MAX_NEXT]] = tail call i32 @llvm.smax.i32(i32 [[MAX]], i32 [[L]]) +; CHECK-NEXT: store i32 [[MAX_NEXT]], ptr [[DST]], align 4 +; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], 1 +; CHECK-NEXT: store i32 [[MAX_NEXT]], ptr [[DST]], align 4 +; CHECK-NEXT: [[EC:%.*]] = icmp eq i64 [[IV]], [[N]] +; CHECK-NEXT: br i1 [[EC]], label %[[EXIT]], label %[[LOOP]], !llvm.loop [[LOOP6:![0-9]+]] +; CHECK: [[EXIT]]: +; CHECK-NEXT: ret void +; +entry: + br label %loop + +loop: + %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ] + %max = phi i32 [ 0, %entry ], [ %max.next, %loop ] + %gep.src = getelementptr inbounds i32, ptr %src, i64 %iv + %l = load i32, ptr %gep.src, align 4 + %max.next = tail call i32 @llvm.smax.i32(i32 %max, i32 %l) + store i32 %max.next, ptr %dst, align 4 + %iv.next = add i64 %iv, 1 + store i32 %max.next, ptr %dst, align 4 + %ec = icmp eq i64 %iv, %n + br i1 %ec, label %exit, label %loop + +exit: + ret void +} + +define void @smax_with_multiple_invariant_store_user_same_addr2(ptr noalias %src, ptr %dst, i64 %n) { +; CHECK-LABEL: define void @smax_with_multiple_invariant_store_user_same_addr2( +; CHECK-SAME: ptr noalias [[SRC:%.*]], ptr [[DST:%.*]], i64 [[N:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: br label %[[LOOP:.*]] +; CHECK: [[LOOP]]: +; CHECK-NEXT: [[IV:%.*]] = phi i64 [ 0, %[[ENTRY]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[MAX:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[MAX_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[GEP_SRC:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i64 [[IV]] +; CHECK-NEXT: [[L:%.*]] = load i32, ptr [[GEP_SRC]], align 4 +; CHECK-NEXT: [[MAX_NEXT]] = tail call i32 @llvm.smax.i32(i32 [[MAX]], i32 [[L]]) +; CHECK-NEXT: store i32 [[MAX_NEXT]], ptr [[DST]], align 4 +; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], 1 +; CHECK-NEXT: store i32 0, ptr [[DST]], align 4 +; CHECK-NEXT: [[EC:%.*]] = icmp eq i64 [[IV]], [[N]] +; CHECK-NEXT: br i1 [[EC]], label %[[EXIT:.*]], label %[[LOOP]] +; CHECK: [[EXIT]]: +; CHECK-NEXT: ret void +; +entry: + br label %loop + +loop: + %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ] + %max = phi i32 [ 0, %entry ], [ %max.next, %loop ] + %gep.src = getelementptr inbounds i32, ptr %src, i64 %iv + %l = load i32, ptr %gep.src, align 4 + %max.next = tail call i32 @llvm.smax.i32(i32 %max, i32 %l) + store i32 %max.next, ptr %dst, align 4 + %iv.next = add i64 %iv, 1 + store i32 0, ptr %dst, align 4 + %ec = icmp eq i64 %iv, %n + br i1 %ec, label %exit, label %loop + +exit: + ret void +} + +define void @smax_with_multiple_invariant_store_user_same_addr3(ptr noalias %src, ptr %dst, i64 %n) { +; CHECK-LABEL: define void @smax_with_multiple_invariant_store_user_same_addr3( +; CHECK-SAME: ptr noalias [[SRC:%.*]], ptr [[DST:%.*]], i64 [[N:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: [[TMP0:%.*]] = add i64 [[N]], 1 +; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP0]], 4 +; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[TMP0]], 4 +; CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[TMP0]], [[N_MOD_VF]] +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP1:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i64 [[INDEX]] +; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x i32>, ptr [[TMP1]], align 4 +; CHECK-NEXT: [[TMP2]] = call <4 x i32> @llvm.smax.v4i32(<4 x i32> [[VEC_PHI]], <4 x i32> [[WIDE_LOAD]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4 +; CHECK-NEXT: [[TMP3:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]] +; CHECK-NEXT: br i1 [[TMP3]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP7:![0-9]+]] +; CHECK: [[MIDDLE_BLOCK]]: +; CHECK-NEXT: [[TMP4:%.*]] = call i32 @llvm.vector.reduce.smax.v4i32(<4 x i32> [[TMP2]]) +; CHECK-NEXT: store i32 [[TMP4]], ptr [[DST]], align 4 +; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[TMP0]], [[N_VEC]] +; CHECK-NEXT: br i1 [[CMP_N]], label %[[EXIT:.*]], label %[[SCALAR_PH]] +; CHECK: [[SCALAR_PH]]: +; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ] +; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ [[TMP4]], %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ] +; CHECK-NEXT: br label %[[LOOP:.*]] +; CHECK: [[LOOP]]: +; CHECK-NEXT: [[IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[MAX:%.*]] = phi i32 [ [[BC_MERGE_RDX]], %[[SCALAR_PH]] ], [ [[MAX_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[GEP_SRC:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i64 [[IV]] +; CHECK-NEXT: [[L:%.*]] = load i32, ptr [[GEP_SRC]], align 4 +; CHECK-NEXT: [[MAX_NEXT]] = tail call i32 @llvm.smax.i32(i32 [[MAX]], i32 [[L]]) +; CHECK-NEXT: store i32 0, ptr [[DST]], align 4 +; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], 1 +; CHECK-NEXT: store i32 [[MAX_NEXT]], ptr [[DST]], align 4 +; CHECK-NEXT: [[EC:%.*]] = icmp eq i64 [[IV]], [[N]] +; CHECK-NEXT: br i1 [[EC]], label %[[EXIT]], label %[[LOOP]], !llvm.loop [[LOOP8:![0-9]+]] +; CHECK: [[EXIT]]: +; CHECK-NEXT: ret void +; +entry: + br label %loop + +loop: + %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ] + %max = phi i32 [ 0, %entry ], [ %max.next, %loop ] + %gep.src = getelementptr inbounds i32, ptr %src, i64 %iv + %l = load i32, ptr %gep.src, align 4 + %max.next = tail call i32 @llvm.smax.i32(i32 %max, i32 %l) + store i32 0, ptr %dst, align 4 + %iv.next = add i64 %iv, 1 + store i32 %max.next, ptr %dst, align 4 + %ec = icmp eq i64 %iv, %n + br i1 %ec, label %exit, label %loop + +exit: + ret void +} + +define void @smax_with_multiple_invariant_store_user_different_addr(ptr noalias %src, ptr noalias %dst, ptr noalias %dst.2, i64 %n) { +; CHECK-LABEL: define void @smax_with_multiple_invariant_store_user_different_addr( +; CHECK-SAME: ptr noalias [[SRC:%.*]], ptr noalias [[DST:%.*]], ptr noalias [[DST_2:%.*]], i64 [[N:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: br label %[[LOOP:.*]] +; CHECK: [[LOOP]]: +; CHECK-NEXT: [[IV:%.*]] = phi i64 [ 0, %[[ENTRY]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[MAX:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[MAX_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[GEP_SRC:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i64 [[IV]] +; CHECK-NEXT: [[L:%.*]] = load i32, ptr [[GEP_SRC]], align 4 +; CHECK-NEXT: [[MAX_NEXT]] = tail call i32 @llvm.smax.i32(i32 [[MAX]], i32 [[L]]) +; CHECK-NEXT: store i32 [[MAX_NEXT]], ptr [[DST]], align 4 +; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], 1 +; CHECK-NEXT: store i32 [[MAX_NEXT]], ptr [[DST_2]], align 4 +; CHECK-NEXT: [[EC:%.*]] = icmp eq i64 [[IV]], [[N]] +; CHECK-NEXT: br i1 [[EC]], label %[[EXIT:.*]], label %[[LOOP]] +; CHECK: [[EXIT]]: +; CHECK-NEXT: ret void +; +entry: + br label %loop + +loop: + %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ] + %max = phi i32 [ 0, %entry ], [ %max.next, %loop ] + %gep.src = getelementptr inbounds i32, ptr %src, i64 %iv + %l = load i32, ptr %gep.src, align 4 + %max.next = tail call i32 @llvm.smax.i32(i32 %max, i32 %l) + store i32 %max.next, ptr %dst, align 4 + %iv.next = add i64 %iv, 1 + store i32 %max.next, ptr %dst.2, align 4 + %ec = icmp eq i64 %iv, %n + br i1 %ec, label %exit, label %loop + +exit: + ret void +} + +define i32 @chained_instructions_feeding_max1(i32 %x, ptr %src) { +; CHECK-LABEL: define i32 @chained_instructions_feeding_max1( +; CHECK-SAME: i32 [[X:%.*]], ptr [[SRC:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: br label %[[LOOP:.*]] +; CHECK: [[LOOP]]: +; CHECK-NEXT: [[IV:%.*]] = phi i64 [ 0, %[[ENTRY]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[MAX:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[MAX_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[GEP_SRC:%.*]] = getelementptr [3 x i32], ptr [[SRC]], i64 [[IV]] +; CHECK-NEXT: [[L:%.*]] = load i32, ptr [[GEP_SRC]], align 4 +; CHECK-NEXT: [[ADD:%.*]] = add i32 [[MAX]], [[L]] +; CHECK-NEXT: [[MAX_NEXT]] = tail call i32 @llvm.smax.i32(i32 [[ADD]], i32 [[L]]) +; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], 1 +; CHECK-NEXT: [[EC:%.*]] = icmp eq i64 [[IV]], 1 +; CHECK-NEXT: br i1 [[EC]], label %[[EXIT:.*]], label %[[LOOP]] +; CHECK: [[EXIT]]: +; CHECK-NEXT: [[MAX_NEXT_LCSSA:%.*]] = phi i32 [ [[MAX_NEXT]], %[[LOOP]] ] +; CHECK-NEXT: ret i32 [[MAX_NEXT_LCSSA]] +; +entry: + br label %loop + +loop: ; preds = %loop, %entry + %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ] + %max = phi i32 [ 0, %entry ], [ %max.next, %loop ] + %gep.src = getelementptr [3 x i32], ptr %src, i64 %iv + %l = load i32, ptr %gep.src, align 4 + %add = add i32 %max, %l + %max.next = tail call i32 @llvm.smax.i32(i32 %add, i32 %l) + %iv.next = add i64 %iv, 1 + %ec = icmp eq i64 %iv, 1 + br i1 %ec, label %exit, label %loop + +exit: + ret i32 %max.next +} + +define i32 @chained_instructions_feeding_max2(i32 %x, ptr %src) { +; CHECK-LABEL: define i32 @chained_instructions_feeding_max2( +; CHECK-SAME: i32 [[X:%.*]], ptr [[SRC:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: br label %[[LOOP:.*]] +; CHECK: [[LOOP]]: +; CHECK-NEXT: [[IV:%.*]] = phi i64 [ 0, %[[ENTRY]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[MAX:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[MAX_NEXT:%.*]], %[[LOOP]] ] +; CHECK-NEXT: [[GEP_SRC:%.*]] = getelementptr [3 x i32], ptr [[SRC]], i64 [[IV]] +; CHECK-NEXT: [[MAX_1:%.*]] = tail call i32 @llvm.smax.i32(i32 [[X]], i32 [[MAX]]) +; CHECK-NEXT: [[L:%.*]] = load i32, ptr [[GEP_SRC]], align 4 +; CHECK-NEXT: [[ADD:%.*]] = add i32 [[L]], [[MAX_1]] +; CHECK-NEXT: [[MAX_NEXT]] = tail call i32 @llvm.smax.i32(i32 [[ADD]], i32 100) +; CHECK-NEXT: [[IV_NEXT]] = add i64 [[IV]], 1 +; CHECK-NEXT: [[EC:%.*]] = icmp eq i64 [[IV]], 1 +; CHECK-NEXT: br i1 [[EC]], label %[[EXIT:.*]], label %[[LOOP]] +; CHECK: [[EXIT]]: +; CHECK-NEXT: [[MAX_NEXT_LCSSA:%.*]] = phi i32 [ [[MAX_NEXT]], %[[LOOP]] ] +; CHECK-NEXT: ret i32 [[MAX_NEXT_LCSSA]] +; +entry: + br label %loop + +loop: ; preds = %loop, %entry + %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop ] + %max = phi i32 [ 0, %entry ], [ %max.next, %loop ] + %gep.src = getelementptr [3 x i32], ptr %src, i64 %iv + %max.1 = tail call i32 @llvm.smax.i32(i32 %x, i32 %max) + %l = load i32, ptr %gep.src, align 4 + %add = add i32 %l, %max.1 + %max.next = tail call i32 @llvm.smax.i32(i32 %add, i32 100) + %iv.next = add i64 %iv, 1 + %ec = icmp eq i64 %iv, 1 + br i1 %ec, label %exit, label %loop + +exit: + ret i32 %max.next +} + + +define i32 @test_predicated_smin(ptr %src) { +; CHECK-LABEL: define i32 @test_predicated_smin( +; CHECK-SAME: ptr [[SRC:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*:]] +; CHECK-NEXT: br label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[PREDPHI:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[TMP0:%.*]] = getelementptr float, ptr [[SRC]], i64 [[INDEX]] +; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x float>, ptr [[TMP0]], align 4 +; CHECK-NEXT: [[TMP1:%.*]] = fcmp une <4 x float> [[WIDE_LOAD]], zeroinitializer +; CHECK-NEXT: [[TMP2:%.*]] = fdiv <4 x float> [[WIDE_LOAD]], splat (float 3.000000e+00) +; CHECK-NEXT: [[TMP3:%.*]] = fptosi <4 x float> [[TMP2]] to <4 x i32> +; CHECK-NEXT: [[TMP4:%.*]] = call <4 x i32> @llvm.smin.v4i32(<4 x i32> [[VEC_PHI]], <4 x i32> [[TMP3]]) +; CHECK-NEXT: [[PREDPHI]] = select <4 x i1> [[TMP1]], <4 x i32> [[TMP4]], <4 x i32> [[VEC_PHI]] +; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 4 +; CHECK-NEXT: [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 112 +; CHECK-NEXT: br i1 [[TMP5]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP9:![0-9]+]] +; CHECK: [[MIDDLE_BLOCK]]: +; CHECK-NEXT: [[TMP6:%.*]] = call i32 @llvm.vector.reduce.smin.v4i32(<4 x i32> [[PREDPHI]]) +; CHECK-NEXT: br label %[[EXIT:.*]] +; CHECK: [[EXIT]]: +; CHECK-NEXT: ret i32 [[TMP6]] +; +entry: + br label %loop.header + +loop.header: + %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop.latch ] + %min = phi i32 [ 0, %entry ], [ %min.merge, %loop.latch ] + %gep.src = getelementptr float, ptr %src, i64 %iv + %l = load float, ptr %gep.src, align 4 + %c = fcmp une float %l, 0.0 + br i1 %c, label %then, label %loop.latch + +then: + %div = fdiv float %l, 3.0 + %div.i32 = fptosi float %div to i32 + %min.next = tail call i32 @llvm.smin.i32(i32 %min, i32 %div.i32) + br label %loop.latch + +loop.latch: + %min.merge = phi i32 [ %min.next, %then ], [ %min, %loop.header ] + %iv.next = add i64 %iv, 1 + %ec = icmp eq i64 %iv, 111 + br i1 %ec, label %exit, label %loop.header + +exit: + ret i32 %min.merge +} + +define i32 @smax_reduction_multiple_incoming(ptr %src, i32 %n, i1 %cond) { +; CHECK-LABEL: define i32 @smax_reduction_multiple_incoming( +; CHECK-SAME: ptr [[SRC:%.*]], i32 [[N:%.*]], i1 [[COND:%.*]]) { +; CHECK-NEXT: [[ENTRY:.*]]: +; CHECK-NEXT: br i1 [[COND]], label %[[LOOP_HEADER_PREHEADER:.*]], label %[[ELSE:.*]] +; CHECK: [[ELSE]]: +; CHECK-NEXT: br label %[[LOOP_HEADER_PREHEADER]] +; CHECK: [[LOOP_HEADER_PREHEADER]]: +; CHECK-NEXT: [[IV_PH:%.*]] = phi i32 [ 10, %[[ELSE]] ], [ 0, %[[ENTRY]] ] +; CHECK-NEXT: [[MAX_PH:%.*]] = phi i32 [ 5, %[[ELSE]] ], [ 0, %[[ENTRY]] ] +; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N]], 1 +; CHECK-NEXT: [[TMP1:%.*]] = sub i32 [[TMP0]], [[IV_PH]] +; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[TMP1]], 4 +; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label %[[SCALAR_PH:.*]], label %[[VECTOR_SCEVCHECK:.*]] +; CHECK: [[VECTOR_SCEVCHECK]]: +; CHECK-NEXT: [[TMP2:%.*]] = icmp slt i32 [[N]], [[IV_PH]] +; CHECK-NEXT: br i1 [[TMP2]], label %[[SCALAR_PH]], label %[[VECTOR_PH:.*]] +; CHECK: [[VECTOR_PH]]: +; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i32 [[TMP1]], 4 +; CHECK-NEXT: [[N_VEC:%.*]] = sub i32 [[TMP1]], [[N_MOD_VF]] +; CHECK-NEXT: [[TMP3:%.*]] = add i32 [[IV_PH]], [[N_VEC]] +; CHECK-NEXT: [[BROADCAST_SPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[MAX_PH]], i64 0 +; CHECK-NEXT: [[BROADCAST_SPLAT:%.*]] = shufflevector <4 x i32> [[BROADCAST_SPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer +; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] +; CHECK: [[VECTOR_BODY]]: +; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[VEC_PHI:%.*]] = phi <4 x i32> [ [[BROADCAST_SPLAT]], %[[VECTOR_PH]] ], [ [[TMP5:%.*]], %[[VECTOR_BODY]] ] +; CHECK-NEXT: [[OFFSET_IDX:%.*]] = add i32 [[IV_PH]], [[INDEX]] +; CHECK-NEXT: [[TMP4:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i32 [[OFFSET_IDX]] +; CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x i32>, ptr [[TMP4]], align 4 +; CHECK-NEXT: [[TMP5]] = call <4 x i32> @llvm.smax.v4i32(<4 x i32> [[VEC_PHI]], <4 x i32> [[WIDE_LOAD]]) +; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4 +; CHECK-NEXT: [[TMP6:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]] +; CHECK-NEXT: br i1 [[TMP6]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP10:![0-9]+]] +; CHECK: [[MIDDLE_BLOCK]]: +; CHECK-NEXT: [[TMP7:%.*]] = call i32 @llvm.vector.reduce.smax.v4i32(<4 x i32> [[TMP5]]) +; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 [[TMP1]], [[N_VEC]] +; CHECK-NEXT: br i1 [[CMP_N]], label %[[EXIT:.*]], label %[[SCALAR_PH]] +; CHECK: [[SCALAR_PH]]: +; CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i32 [ [[TMP3]], %[[MIDDLE_BLOCK]] ], [ [[IV_PH]], %[[LOOP_HEADER_PREHEADER]] ], [ [[IV_PH]], %[[VECTOR_SCEVCHECK]] ] +; CHECK-NEXT: [[BC_MERGE_RDX:%.*]] = phi i32 [ [[TMP7]], %[[MIDDLE_BLOCK]] ], [ [[MAX_PH]], %[[LOOP_HEADER_PREHEADER]] ], [ [[MAX_PH]], %[[VECTOR_SCEVCHECK]] ] +; CHECK-NEXT: br label %[[LOOP_HEADER:.*]] +; CHECK: [[LOOP_HEADER]]: +; CHECK-NEXT: [[IV:%.*]] = phi i32 [ [[IV_NEXT:%.*]], %[[LOOP_HEADER]] ], [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ] +; CHECK-NEXT: [[MAX:%.*]] = phi i32 [ [[MAX_NEXT:%.*]], %[[LOOP_HEADER]] ], [ [[BC_MERGE_RDX]], %[[SCALAR_PH]] ] +; CHECK-NEXT: [[GEP_SRC:%.*]] = getelementptr inbounds i32, ptr [[SRC]], i32 [[IV]] +; CHECK-NEXT: [[L:%.*]] = load i32, ptr [[GEP_SRC]], align 4 +; CHECK-NEXT: [[MAX_NEXT]] = tail call i32 @llvm.smax.i32(i32 [[MAX]], i32 [[L]]) +; CHECK-NEXT: [[IV_NEXT]] = add i32 [[IV]], 1 +; CHECK-NEXT: [[EC:%.*]] = icmp eq i32 [[IV]], [[N]] +; CHECK-NEXT: br i1 [[EC]], label %[[EXIT]], label %[[LOOP_HEADER]], !llvm.loop [[LOOP11:![0-9]+]] +; CHECK: [[EXIT]]: +; CHECK-NEXT: [[MAX_NEXT_LCSSA:%.*]] = phi i32 [ [[MAX_NEXT]], %[[LOOP_HEADER]] ], [ [[TMP7]], %[[MIDDLE_BLOCK]] ] +; CHECK-NEXT: ret i32 [[MAX_NEXT_LCSSA]] +; +entry: + br i1 %cond, label %loop.header, label %else + +else: + br label %loop.header + +loop.header: + %iv = phi i32 [ 0, %entry ], [ 10, %else ], [ %iv.next, %loop.header ] + %max = phi i32 [ 0, %entry ], [ 5, %else ], [ %max.next, %loop.header ] + %gep.src = getelementptr inbounds i32, ptr %src, i32 %iv + %l = load i32, ptr %gep.src, align 4 + %max.next = tail call i32 @llvm.smax.i32(i32 %max, i32 %l) + %iv.next = add i32 %iv, 1 + %ec = icmp eq i32 %iv, %n + br i1 %ec, label %exit, label %loop.header + +exit: + ret i32 %max.next +} diff --git a/llvm/test/Transforms/SROA/phi-and-select.ll b/llvm/test/Transforms/SROA/phi-and-select.ll index 616617b..5d5a610 100644 --- a/llvm/test/Transforms/SROA/phi-and-select.ll +++ b/llvm/test/Transforms/SROA/phi-and-select.ll @@ -1,4 +1,4 @@ -; NOTE: Assertions have been autogenerated by utils/update_test_checks.py +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals ; RUN: opt < %s -passes='sroa<preserve-cfg>' -S | FileCheck %s --check-prefixes=CHECK,CHECK-PRESERVE-CFG ; RUN: opt < %s -passes='sroa<modify-cfg>' -S | FileCheck %s --check-prefixes=CHECK,CHECK-MODIFY-CFG target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-n8:16:32:64" @@ -36,11 +36,11 @@ exit: ret i32 %result } -define i32 @test2() { +define i32 @test2() !prof !0 { ; CHECK-LABEL: @test2( ; CHECK-NEXT: entry: ; CHECK-NEXT: [[COND:%.*]] = icmp sle i32 0, 1 -; CHECK-NEXT: [[RESULT_SROA_SPECULATED:%.*]] = select i1 [[COND]], i32 1, i32 0 +; CHECK-NEXT: [[RESULT_SROA_SPECULATED:%.*]] = select i1 [[COND]], i32 1, i32 0, !prof [[PROF1:![0-9]+]] ; CHECK-NEXT: ret i32 [[RESULT_SROA_SPECULATED]] ; entry: @@ -53,7 +53,7 @@ entry: %v1 = load i32, ptr %a1 %cond = icmp sle i32 %v0, %v1 - %select = select i1 %cond, ptr %a1, ptr %a + %select = select i1 %cond, ptr %a1, ptr %a, !prof !1 %result = load i32, ptr %select ret i32 %result @@ -870,3 +870,17 @@ define i8 @volatile_select(ptr %p, i1 %b) { %v2 = load i8, ptr %px ret i8 %v2 } + +!0 = !{!"function_entry_count", i32 10} +!1 = !{!"branch_weights", i32 3, i32 5} +;. +; CHECK-PRESERVE-CFG: attributes #[[ATTR0:[0-9]+]] = { sanitize_address } +;. +; CHECK-MODIFY-CFG: attributes #[[ATTR0:[0-9]+]] = { sanitize_address } +;. +; CHECK-PRESERVE-CFG: [[META0:![0-9]+]] = !{!"function_entry_count", i32 10} +; CHECK-PRESERVE-CFG: [[PROF1]] = !{!"branch_weights", i32 3, i32 5} +;. +; CHECK-MODIFY-CFG: [[META0:![0-9]+]] = !{!"function_entry_count", i32 10} +; CHECK-MODIFY-CFG: [[PROF1]] = !{!"branch_weights", i32 3, i32 5} +;. diff --git a/llvm/test/Transforms/SROA/phi-gep.ll b/llvm/test/Transforms/SROA/phi-gep.ll index 776624c..45c3bbd 100644 --- a/llvm/test/Transforms/SROA/phi-gep.ll +++ b/llvm/test/Transforms/SROA/phi-gep.ll @@ -1,9 +1,12 @@ -; NOTE: Assertions have been autogenerated by utils/update_test_checks.py +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals ; RUN: opt -S -passes='sroa<preserve-cfg>' < %s | FileCheck %s --check-prefixes=CHECK,CHECK-PRESERVE-CFG ; RUN: opt -S -passes='sroa<modify-cfg>' < %s | FileCheck %s --check-prefixes=CHECK,CHECK-MODIFY-CFG %pair = type { i32, i32 } +;. +; CHECK: @g = global %pair zeroinitializer, align 4 +;. define i32 @test_sroa_phi_gep(i1 %cond) { ; CHECK-LABEL: @test_sroa_phi_gep( ; CHECK-NEXT: entry: @@ -334,18 +337,18 @@ exit: unreachable } -define void @test_sroa_gep_phi_select_same_block(i1 %c1, i1 %c2, ptr %ptr) { +define void @test_sroa_gep_phi_select_same_block(i1 %c1, i1 %c2, ptr %ptr) !prof !0 { ; CHECK-LABEL: @test_sroa_gep_phi_select_same_block( ; CHECK-NEXT: entry: ; CHECK-NEXT: [[ALLOCA:%.*]] = alloca [[PAIR:%.*]], align 8 ; CHECK-NEXT: br label [[WHILE_BODY:%.*]] ; CHECK: while.body: ; CHECK-NEXT: [[PHI:%.*]] = phi ptr [ [[ALLOCA]], [[ENTRY:%.*]] ], [ [[SELECT:%.*]], [[WHILE_BODY]] ] -; CHECK-NEXT: [[SELECT]] = select i1 [[C1:%.*]], ptr [[PHI]], ptr [[PTR:%.*]] +; CHECK-NEXT: [[SELECT]] = select i1 [[C1:%.*]], ptr [[PHI]], ptr [[PTR:%.*]], !prof [[PROF1:![0-9]+]] ; CHECK-NEXT: [[PHI_SROA_GEP:%.*]] = getelementptr inbounds [[PAIR]], ptr [[PHI]], i64 1 ; CHECK-NEXT: [[PTR_SROA_GEP:%.*]] = getelementptr inbounds [[PAIR]], ptr [[PTR]], i64 1 -; CHECK-NEXT: [[SELECT_SROA_SEL:%.*]] = select i1 [[C1]], ptr [[PHI_SROA_GEP]], ptr [[PTR_SROA_GEP]] -; CHECK-NEXT: br i1 [[C2:%.*]], label [[EXIT:%.*]], label [[WHILE_BODY]] +; CHECK-NEXT: [[SELECT_SROA_SEL:%.*]] = select i1 [[C1]], ptr [[PHI_SROA_GEP]], ptr [[PTR_SROA_GEP]], !prof [[PROF1]] +; CHECK-NEXT: br i1 [[C2:%.*]], label [[EXIT:%.*]], label [[WHILE_BODY]], !prof [[PROF2:![0-9]+]] ; CHECK: exit: ; CHECK-NEXT: ret void ; @@ -355,9 +358,9 @@ entry: while.body: %phi = phi ptr [ %alloca, %entry ], [ %select, %while.body ] - %select = select i1 %c1, ptr %phi, ptr %ptr + %select = select i1 %c1, ptr %phi, ptr %ptr, !prof !1 %gep = getelementptr inbounds %pair, ptr %select, i64 1 - br i1 %c2, label %exit, label %while.body + br i1 %c2, label %exit, label %while.body, !prof !2 exit: ret void @@ -747,6 +750,18 @@ declare ptr @foo() declare i32 @__gxx_personality_v0(...) declare void @llvm.memcpy.p0.p0.i64(ptr noalias nocapture writeonly, ptr noalias nocapture readonly, i64, i1 immarg) + +!0 = !{!"function_entry_count", i32 10} +!1 = !{!"branch_weights", i32 3, i32 5} +!2 = !{!"branch_weights", i32 7, i32 11} + +;. +; CHECK: attributes #[[ATTR0:[0-9]+]] = { nocallback nofree nounwind willreturn memory(argmem: readwrite) } +;. +; CHECK: [[META0:![0-9]+]] = !{!"function_entry_count", i32 10} +; CHECK: [[PROF1]] = !{!"branch_weights", i32 3, i32 5} +; CHECK: [[PROF2]] = !{!"branch_weights", i32 7, i32 11} +;. ;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line: ; CHECK-MODIFY-CFG: {{.*}} ; CHECK-PRESERVE-CFG: {{.*}} diff --git a/llvm/test/Transforms/SROA/select-gep.ll b/llvm/test/Transforms/SROA/select-gep.ll index b48b0f7..a701d78 100644 --- a/llvm/test/Transforms/SROA/select-gep.ll +++ b/llvm/test/Transforms/SROA/select-gep.ll @@ -1,4 +1,4 @@ -; NOTE: Assertions have been autogenerated by utils/update_test_checks.py +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals ; RUN: opt -S -passes='sroa<preserve-cfg>' < %s | FileCheck %s --check-prefixes=CHECK,CHECK-PRESERVE-CFG ; RUN: opt -S -passes='sroa<modify-cfg>' < %s | FileCheck %s --check-prefixes=CHECK,CHECK-MODIFY-CFG @@ -203,10 +203,10 @@ define i32 @test_select_idx_mem2reg(i1 %c) { ; Test gep with a select-like zext index unfolding on an alloca that is ; splittable and promotable. -define i64 @test_select_like_zext_idx_mem2reg(i1 %c) { +define i64 @test_select_like_zext_idx_mem2reg(i1 %c) !prof !0 { ; CHECK-LABEL: @test_select_like_zext_idx_mem2reg( ; CHECK-NEXT: [[IDX:%.*]] = zext i1 [[C:%.*]] to i64 -; CHECK-NEXT: [[RES:%.*]] = select i1 [[C]], i64 2, i64 1 +; CHECK-NEXT: [[RES:%.*]] = select i1 [[C]], i64 2, i64 1, !prof [[PROF1:![0-9]+]] ; CHECK-NEXT: ret i64 [[RES]] ; %alloca = alloca [2 x i64], align 8 @@ -352,3 +352,16 @@ define i32 @test_select_idx_not_constant3(i1 %c, ptr %p, i64 %arg) { %res = load i32, ptr %gep, align 4 ret i32 %res } + +!0 = !{!"function_entry_count", i32 10} +;. +; CHECK-PRESERVE-CFG: attributes #[[ATTR0:[0-9]+]] = { nocallback nofree nounwind willreturn memory(argmem: readwrite) } +;. +; CHECK-MODIFY-CFG: attributes #[[ATTR0:[0-9]+]] = { nocallback nofree nounwind willreturn memory(argmem: readwrite) } +;. +; CHECK-PRESERVE-CFG: [[META0:![0-9]+]] = !{!"function_entry_count", i32 10} +; CHECK-PRESERVE-CFG: [[PROF1]] = !{!"unknown", !"sroa"} +;. +; CHECK-MODIFY-CFG: [[META0:![0-9]+]] = !{!"function_entry_count", i32 10} +; CHECK-MODIFY-CFG: [[PROF1]] = !{!"unknown", !"sroa"} +;. diff --git a/llvm/test/Transforms/SROA/slice-width.ll b/llvm/test/Transforms/SROA/slice-width.ll index eabb697..3b77e49 100644 --- a/llvm/test/Transforms/SROA/slice-width.ll +++ b/llvm/test/Transforms/SROA/slice-width.ll @@ -1,4 +1,4 @@ -; NOTE: Assertions have been autogenerated by utils/update_test_checks.py +; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals ; RUN: opt < %s -passes='sroa<preserve-cfg>' -S | FileCheck %s --check-prefixes=CHECK,CHECK-PRESERVE-CFG ; RUN: opt < %s -passes='sroa<modify-cfg>' -S | FileCheck %s --check-prefixes=CHECK,CHECK-MODIFY-CFG target datalayout = "e-p:64:64:64-p1:16:16:16-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:64:64-f80:128-v64:64:64-v128:128:128-a0:0:64-n8:16:32:64" @@ -8,6 +8,10 @@ declare void @llvm.memset.p0.i32(ptr nocapture, i8, i32, i1) nounwind declare void @llvm.memset.p0.i64(ptr nocapture, i8, i64, i1) nounwind ; This tests that allocas are not split into slices that are not byte width multiple +;. +; CHECK: @foo_copy_source = external constant %union.Foo +; CHECK: @i64_sink = global i64 0 +;. define void @no_split_on_non_byte_width(i32) { ; CHECK-LABEL: @no_split_on_non_byte_width( ; CHECK-NEXT: [[ARG_SROA_0:%.*]] = alloca i8, align 8 @@ -92,12 +96,12 @@ declare i32 @memcpy_vec3float_helper(ptr) ; PR18726: Check that SROA does not rewrite a 12-byte memcpy into a 16-byte ; vector store, hence accidentally putting gibberish onto the stack. -define i32 @memcpy_vec3float_widening(ptr %x) { +define i32 @memcpy_vec3float_widening(ptr %x) !prof !0 { ; CHECK-LABEL: @memcpy_vec3float_widening( ; CHECK-NEXT: entry: ; CHECK-NEXT: [[TMP1_SROA_0_0_COPYLOAD:%.*]] = load <3 x float>, ptr [[X:%.*]], align 4 ; CHECK-NEXT: [[TMP1_SROA_0_0_VEC_EXPAND:%.*]] = shufflevector <3 x float> [[TMP1_SROA_0_0_COPYLOAD]], <3 x float> poison, <4 x i32> <i32 0, i32 1, i32 2, i32 poison> -; CHECK-NEXT: [[TMP1_SROA_0_0_VECBLEND:%.*]] = select <4 x i1> <i1 true, i1 true, i1 true, i1 false>, <4 x float> [[TMP1_SROA_0_0_VEC_EXPAND]], <4 x float> undef +; CHECK-NEXT: [[TMP1_SROA_0_0_VECBLEND:%.*]] = select <4 x i1> <i1 true, i1 true, i1 true, i1 false>, <4 x float> [[TMP1_SROA_0_0_VEC_EXPAND]], <4 x float> undef, !prof [[PROF1:![0-9]+]] ; CHECK-NEXT: [[TMP2:%.*]] = alloca [[S_VEC3FLOAT:%.*]], align 4 ; CHECK-NEXT: [[TMP1_SROA_0_0_VEC_EXTRACT:%.*]] = shufflevector <4 x float> [[TMP1_SROA_0_0_VECBLEND]], <4 x float> poison, <3 x i32> <i32 0, i32 1, i32 2> ; CHECK-NEXT: store <3 x float> [[TMP1_SROA_0_0_VEC_EXTRACT]], ptr [[TMP2]], align 4 @@ -158,6 +162,15 @@ define i1 @presplit_overlarge_load() { %L2 = load i1, ptr %A ret i1 %L2 } +!0 = !{!"function_entry_count", i32 10} + +;. +; CHECK: attributes #[[ATTR0:[0-9]+]] = { nocallback nofree nounwind willreturn memory(argmem: readwrite) } +; CHECK: attributes #[[ATTR1:[0-9]+]] = { nocallback nofree nounwind willreturn memory(argmem: write) } +;. +; CHECK: [[META0:![0-9]+]] = !{!"function_entry_count", i32 10} +; CHECK: [[PROF1]] = !{!"unknown", !"sroa"} +;. ;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line: ; CHECK-MODIFY-CFG: {{.*}} ; CHECK-PRESERVE-CFG: {{.*}} diff --git a/llvm/test/tools/llvm-objcopy/DXContainer/dump-section-errors.yaml b/llvm/test/tools/llvm-objcopy/DXContainer/dump-section-errors.yaml new file mode 100644 index 0000000..e748eecf --- /dev/null +++ b/llvm/test/tools/llvm-objcopy/DXContainer/dump-section-errors.yaml @@ -0,0 +1,27 @@ +# RUN: yaml2obj %s -o %t.dxbc +# RUN: not llvm-objcopy --dump-section=FKE0=%t.fek0 %t.dxbc 2>&1 | FileCheck %s --check-prefix=CHECK-ZEROSIZE -DFILE=%t.fek0 +# RUN: not llvm-objcopy --dump-section=FKE3=%t.fek1 %t.dxbc 2>&1 | FileCheck %s --check-prefix=CHECK-MISSING -DFILE=%t.fek1 +# RUN: not llvm-objcopy --dump-section=FKE2=%t/does_not_exist/.fek2 %t.dxbc 2>&1 | FileCheck %s --check-prefix=CHECK-BAD-PATH -DFILE=%t/does_not_exist/.fek2 -DMSG=%errc_ENOENT + +# CHECK-ZEROSIZE: error: '[[FILE]]': part 'FKE0' is empty +# CHECK-MISSING: error: '[[FILE]]': part 'FKE3' not found +# CHECK-BAD-PATH: error: '[[FILE]]': [[MSG]] + +--- !dxcontainer +Header: + Hash: [ 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 ] + Version: + Major: 1 + Minor: 0 + FileSize: 108 + PartCount: 3 + PartOffsets: [ 60, 68, 76 ] +Parts: + - Name: FKE0 + Size: 0 + - Name: FKE1 + Size: 0 + - Name: FKE2 + Size: 8 +... diff --git a/llvm/test/tools/llvm-objcopy/DXContainer/dump-section.yaml b/llvm/test/tools/llvm-objcopy/DXContainer/dump-section.yaml new file mode 100644 index 0000000..7d80a2c --- /dev/null +++ b/llvm/test/tools/llvm-objcopy/DXContainer/dump-section.yaml @@ -0,0 +1,278 @@ +# RUN: yaml2obj %s -o %t.dxbc +# RUN: llvm-objcopy --dump-section=DXIL=%t.bc %t.dxbc +# RUN: llvm-dis %t.bc -o - | FileCheck %s --check-prefix=BITCODE +# RUN: wc -c %t.bc | FileCheck %s --check-prefix=DXIL-SIZE + +## Verify that when dumping the DXIL part we get a valid bitcode file. +# BITCODE: define void @main() +## Verify the size of the bitcode data. +# DXIL-SIZE: 1708 + +## Dump the PSV0 part and verify its size. +# RUN: llvm-objcopy --dump-section=PSV0=%t.psv0 %t.dxbc +# RUN: wc -c %t.psv0 | FileCheck %s --check-prefix=PSV0-SIZE +# RUN: od -v -Ax -t x1 %t.psv0 | FileCheck %s --check-prefix=PSV0-CONTENTS +# PSV0-SIZE: 76 + +# For a compute shader the structure size is encoded followed by a bunch of 00'd +# bytes until you get to the unused wave size min and max (0xffff), followed by +# the shader stage (5 for compute). +# TODO: Update this test to use objdump or obj2yaml once we support +# --add-section in objcopy. See issue: +# https://github.com/llvm/llvm-project/issues/162159. +# PSV0-CONTENTS: 0000 34 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 +# PSV0-CONTENTS: 0010 00 00 00 00 00 00 00 00 ff ff ff ff 05 00 00 00 + +--- !dxcontainer +Header: + Hash: [ 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, 0x0, 0x0 ] + Version: + Major: 1 + Minor: 0 + FileSize: 1872 + PartCount: 2 + PartOffsets: [ 40, 1780 ] +Parts: + - Name: DXIL + Size: 1732 + Program: + MajorVersion: 6 + MinorVersion: 0 + ShaderKind: 5 + Size: 433 + DXILMajorVersion: 1 + DXILMinorVersion: 0 + DXILSize: 1708 + DXIL: [ 0x42, 0x43, 0xC0, 0xDE, 0x21, 0xC, 0x0, 0x0, 0xA8, + 0x1, 0x0, 0x0, 0xB, 0x82, 0x20, 0x0, 0x2, 0x0, + 0x0, 0x0, 0x13, 0x0, 0x0, 0x0, 0x7, 0x81, 0x23, + 0x91, 0x41, 0xC8, 0x4, 0x49, 0x6, 0x10, 0x32, + 0x39, 0x92, 0x1, 0x84, 0xC, 0x25, 0x5, 0x8, 0x19, + 0x1E, 0x4, 0x8B, 0x62, 0x80, 0x10, 0x45, 0x2, + 0x42, 0x92, 0xB, 0x42, 0x84, 0x10, 0x32, 0x14, + 0x38, 0x8, 0x18, 0x4B, 0xA, 0x32, 0x42, 0x88, + 0x48, 0x90, 0x14, 0x20, 0x43, 0x46, 0x88, 0xA5, + 0x0, 0x19, 0x32, 0x42, 0xE4, 0x48, 0xE, 0x90, + 0x11, 0x22, 0xC4, 0x50, 0x41, 0x51, 0x81, 0x8C, + 0xE1, 0x83, 0xE5, 0x8A, 0x4, 0x21, 0x46, 0x6, + 0x51, 0x18, 0x0, 0x0, 0x4, 0x0, 0x0, 0x0, 0x1B, + 0x90, 0xE0, 0xFF, 0xFF, 0xFF, 0xFF, 0x7, 0xC0, + 0x1, 0x24, 0x80, 0x2, 0x0, 0x0, 0x0, 0x49, 0x18, + 0x0, 0x0, 0x1, 0x0, 0x0, 0x0, 0x13, 0x82, 0x0, + 0x0, 0x89, 0x20, 0x0, 0x0, 0x11, 0x0, 0x0, 0x0, + 0x32, 0x22, 0x8, 0x9, 0x20, 0x64, 0x85, 0x4, 0x13, + 0x22, 0xA4, 0x84, 0x4, 0x13, 0x22, 0xE3, 0x84, + 0xA1, 0x90, 0x14, 0x12, 0x4C, 0x88, 0x8C, 0xB, + 0x84, 0x84, 0x4C, 0x10, 0x20, 0x73, 0x4, 0x8, + 0xC1, 0x65, 0xC3, 0x85, 0x2C, 0xE8, 0x3, 0x40, + 0x14, 0x91, 0x4E, 0xD1, 0x4A, 0x48, 0x44, 0x54, + 0x11, 0xC3, 0x9, 0x30, 0xC4, 0x18, 0x1, 0x30, + 0x2, 0x50, 0x82, 0x21, 0x1A, 0x8, 0x98, 0x23, + 0x0, 0x3, 0x0, 0x13, 0x14, 0x72, 0xC0, 0x87, 0x74, + 0x60, 0x87, 0x36, 0x68, 0x87, 0x79, 0x68, 0x3, + 0x72, 0xC0, 0x87, 0xD, 0xAE, 0x50, 0xE, 0x6D, + 0xD0, 0xE, 0x7A, 0x50, 0xE, 0x6D, 0x0, 0xF, 0x7A, + 0x30, 0x7, 0x72, 0xA0, 0x7, 0x73, 0x20, 0x7, 0x6D, + 0x90, 0xE, 0x71, 0xA0, 0x7, 0x73, 0x20, 0x7, 0x6D, + 0x90, 0xE, 0x78, 0xA0, 0x7, 0x78, 0xD0, 0x6, 0xE9, + 0x10, 0x7, 0x76, 0xA0, 0x7, 0x71, 0x60, 0x7, 0x6D, + 0x90, 0xE, 0x73, 0x20, 0x7, 0x7A, 0x30, 0x7, 0x72, + 0xD0, 0x6, 0xE9, 0x60, 0x7, 0x74, 0xA0, 0x7, 0x76, + 0x40, 0x7, 0x6D, 0x60, 0xE, 0x71, 0x60, 0x7, 0x7A, + 0x10, 0x7, 0x76, 0xD0, 0x6, 0xE6, 0x30, 0x7, 0x72, + 0xA0, 0x7, 0x73, 0x20, 0x7, 0x6D, 0x60, 0xE, 0x76, + 0x40, 0x7, 0x7A, 0x60, 0x7, 0x74, 0xD0, 0x6, 0xEE, + 0x80, 0x7, 0x7A, 0x10, 0x7, 0x76, 0xA0, 0x7, 0x73, + 0x20, 0x7, 0x7A, 0x60, 0x7, 0x74, 0x30, 0xE4, + 0x21, 0x0, 0x0, 0x8, 0x0, 0x0, 0x0, 0x2, 0x0, + 0x0, 0x0, 0x20, 0xB, 0x4, 0x7, 0x0, 0x0, 0x0, + 0x32, 0x1E, 0x98, 0xC, 0x19, 0x11, 0x4C, 0x90, + 0x8C, 0x9, 0x26, 0x47, 0xC6, 0x4, 0x43, 0xBA, + 0x12, 0x28, 0x88, 0x62, 0x18, 0x1, 0x28, 0x84, + 0x22, 0x0, 0x0, 0x0, 0x79, 0x18, 0x0, 0x0, 0xE5, + 0x0, 0x0, 0x0, 0x33, 0x8, 0x80, 0x1C, 0xC4, 0xE1, + 0x1C, 0x66, 0x14, 0x1, 0x3D, 0x88, 0x43, 0x38, + 0x84, 0xC3, 0x8C, 0x42, 0x80, 0x7, 0x79, 0x78, + 0x7, 0x73, 0x98, 0x71, 0xC, 0xE6, 0x0, 0xF, 0xED, + 0x10, 0xE, 0xF4, 0x80, 0xE, 0x33, 0xC, 0x42, 0x1E, + 0xC2, 0xC1, 0x1D, 0xCE, 0xA1, 0x1C, 0x66, 0x30, + 0x5, 0x3D, 0x88, 0x43, 0x38, 0x84, 0x83, 0x1B, + 0xCC, 0x3, 0x3D, 0xC8, 0x43, 0x3D, 0x8C, 0x3, + 0x3D, 0xCC, 0x78, 0x8C, 0x74, 0x70, 0x7, 0x7B, + 0x8, 0x7, 0x79, 0x48, 0x87, 0x70, 0x70, 0x7, 0x7A, + 0x70, 0x3, 0x76, 0x78, 0x87, 0x70, 0x20, 0x87, + 0x19, 0xCC, 0x11, 0xE, 0xEC, 0x90, 0xE, 0xE1, + 0x30, 0xF, 0x6E, 0x30, 0xF, 0xE3, 0xF0, 0xE, 0xF0, + 0x50, 0xE, 0x33, 0x10, 0xC4, 0x1D, 0xDE, 0x21, + 0x1C, 0xD8, 0x21, 0x1D, 0xC2, 0x61, 0x1E, 0x66, + 0x30, 0x89, 0x3B, 0xBC, 0x83, 0x3B, 0xD0, 0x43, + 0x39, 0xB4, 0x3, 0x3C, 0xBC, 0x83, 0x3C, 0x84, + 0x3, 0x3B, 0xCC, 0xF0, 0x14, 0x76, 0x60, 0x7, + 0x7B, 0x68, 0x7, 0x37, 0x68, 0x87, 0x72, 0x68, + 0x7, 0x37, 0x80, 0x87, 0x70, 0x90, 0x87, 0x70, + 0x60, 0x7, 0x76, 0x28, 0x7, 0x76, 0xF8, 0x5, 0x76, + 0x78, 0x87, 0x77, 0x80, 0x87, 0x5F, 0x8, 0x87, + 0x71, 0x18, 0x87, 0x72, 0x98, 0x87, 0x79, 0x98, + 0x81, 0x2C, 0xEE, 0xF0, 0xE, 0xEE, 0xE0, 0xE, + 0xF5, 0xC0, 0xE, 0xEC, 0x30, 0x3, 0x62, 0xC8, + 0xA1, 0x1C, 0xE4, 0xA1, 0x1C, 0xCC, 0xA1, 0x1C, + 0xE4, 0xA1, 0x1C, 0xDC, 0x61, 0x1C, 0xCA, 0x21, + 0x1C, 0xC4, 0x81, 0x1D, 0xCA, 0x61, 0x6, 0xD6, + 0x90, 0x43, 0x39, 0xC8, 0x43, 0x39, 0x98, 0x43, + 0x39, 0xC8, 0x43, 0x39, 0xB8, 0xC3, 0x38, 0x94, + 0x43, 0x38, 0x88, 0x3, 0x3B, 0x94, 0xC3, 0x2F, + 0xBC, 0x83, 0x3C, 0xFC, 0x82, 0x3B, 0xD4, 0x3, + 0x3B, 0xB0, 0xC3, 0xC, 0xC7, 0x69, 0x87, 0x70, + 0x58, 0x87, 0x72, 0x70, 0x83, 0x74, 0x68, 0x7, + 0x78, 0x60, 0x87, 0x74, 0x18, 0x87, 0x74, 0xA0, + 0x87, 0x19, 0xCE, 0x53, 0xF, 0xEE, 0x0, 0xF, 0xF2, + 0x50, 0xE, 0xE4, 0x90, 0xE, 0xE3, 0x40, 0xF, 0xE1, + 0x20, 0xE, 0xEC, 0x50, 0xE, 0x33, 0x20, 0x28, + 0x1D, 0xDC, 0xC1, 0x1E, 0xC2, 0x41, 0x1E, 0xD2, + 0x21, 0x1C, 0xDC, 0x81, 0x1E, 0xDC, 0xE0, 0x1C, + 0xE4, 0xE1, 0x1D, 0xEA, 0x1, 0x1E, 0x66, 0x18, + 0x51, 0x38, 0xB0, 0x43, 0x3A, 0x9C, 0x83, 0x3B, + 0xCC, 0x50, 0x24, 0x76, 0x60, 0x7, 0x7B, 0x68, + 0x7, 0x37, 0x60, 0x87, 0x77, 0x78, 0x7, 0x78, + 0x98, 0x51, 0x4C, 0xF4, 0x90, 0xF, 0xF0, 0x50, + 0xE, 0x33, 0x1E, 0x6A, 0x1E, 0xCA, 0x61, 0x1C, + 0xE8, 0x21, 0x1D, 0xDE, 0xC1, 0x1D, 0x7E, 0x1, + 0x1E, 0xE4, 0xA1, 0x1C, 0xCC, 0x21, 0x1D, 0xF0, + 0x61, 0x6, 0x54, 0x85, 0x83, 0x38, 0xCC, 0xC3, + 0x3B, 0xB0, 0x43, 0x3D, 0xD0, 0x43, 0x39, 0xFC, + 0xC2, 0x3C, 0xE4, 0x43, 0x3B, 0x88, 0xC3, 0x3B, + 0xB0, 0xC3, 0x8C, 0xC5, 0xA, 0x87, 0x79, 0x98, + 0x87, 0x77, 0x18, 0x87, 0x74, 0x8, 0x7, 0x7A, + 0x28, 0x7, 0x72, 0x98, 0x81, 0x5C, 0xE3, 0x10, + 0xE, 0xEC, 0xC0, 0xE, 0xE5, 0x50, 0xE, 0xF3, 0x30, + 0x23, 0xC1, 0xD2, 0x41, 0x1E, 0xE4, 0xE1, 0x17, + 0xD8, 0xE1, 0x1D, 0xDE, 0x1, 0x1E, 0x66, 0x48, + 0x19, 0x3B, 0xB0, 0x83, 0x3D, 0xB4, 0x83, 0x1B, + 0x84, 0xC3, 0x38, 0x8C, 0x43, 0x39, 0xCC, 0xC3, + 0x3C, 0xB8, 0xC1, 0x39, 0xC8, 0xC3, 0x3B, 0xD4, + 0x3, 0x3C, 0xCC, 0x48, 0xB4, 0x71, 0x8, 0x7, 0x76, + 0x60, 0x7, 0x71, 0x8, 0x87, 0x71, 0x58, 0x87, + 0x19, 0xDB, 0xC6, 0xE, 0xEC, 0x60, 0xF, 0xED, + 0xE0, 0x6, 0xF0, 0x20, 0xF, 0xE5, 0x30, 0xF, 0xE5, + 0x20, 0xF, 0xF6, 0x50, 0xE, 0x6E, 0x10, 0xE, 0xE3, + 0x30, 0xE, 0xE5, 0x30, 0xF, 0xF3, 0xE0, 0x6, 0xE9, + 0xE0, 0xE, 0xE4, 0x50, 0xE, 0xF8, 0x30, 0x23, + 0xE2, 0xEC, 0x61, 0x1C, 0xC2, 0x81, 0x1D, 0xD8, + 0xE1, 0x17, 0xEC, 0x21, 0x1D, 0xE6, 0x21, 0x1D, + 0xC4, 0x21, 0x1D, 0xD8, 0x21, 0x1D, 0xE8, 0x21, + 0x1F, 0x66, 0x20, 0x9D, 0x3B, 0xBC, 0x43, 0x3D, + 0xB8, 0x3, 0x39, 0x94, 0x83, 0x39, 0xCC, 0x58, + 0xBC, 0x70, 0x70, 0x7, 0x77, 0x78, 0x7, 0x7A, + 0x8, 0x7, 0x7A, 0x48, 0x87, 0x77, 0x70, 0x87, + 0x19, 0xCB, 0xE7, 0xE, 0xEF, 0x30, 0xF, 0xE1, + 0xE0, 0xE, 0xE9, 0x40, 0xF, 0xE9, 0xA0, 0xF, 0xE5, + 0x30, 0xC3, 0x1, 0x3, 0x73, 0xA8, 0x7, 0x77, 0x18, + 0x87, 0x5F, 0x98, 0x87, 0x70, 0x70, 0x87, 0x74, + 0xA0, 0x87, 0x74, 0xD0, 0x87, 0x72, 0x98, 0x81, + 0x84, 0x41, 0x39, 0xE0, 0xC3, 0x38, 0xB0, 0x43, + 0x3D, 0x90, 0x43, 0x39, 0xCC, 0x40, 0xC4, 0xA0, + 0x1D, 0xCA, 0xA1, 0x1D, 0xE0, 0x41, 0x1E, 0xDE, + 0xC1, 0x1C, 0x66, 0x24, 0x63, 0x30, 0xE, 0xE1, + 0xC0, 0xE, 0xEC, 0x30, 0xF, 0xE9, 0x40, 0xF, 0xE5, + 0x30, 0x43, 0x21, 0x83, 0x75, 0x18, 0x7, 0x73, + 0x48, 0x87, 0x5F, 0xA0, 0x87, 0x7C, 0x80, 0x87, + 0x72, 0x98, 0xB1, 0x94, 0x1, 0x3C, 0x8C, 0xC3, + 0x3C, 0x94, 0xC3, 0x38, 0xD0, 0x43, 0x3A, 0xBC, + 0x83, 0x3B, 0xCC, 0xC3, 0x8C, 0xC5, 0xC, 0x48, + 0x21, 0x15, 0x42, 0x61, 0x1E, 0xE6, 0x21, 0x1D, + 0xCE, 0xC1, 0x1D, 0x52, 0x81, 0x14, 0x66, 0x4C, + 0x67, 0x30, 0xE, 0xEF, 0x20, 0xF, 0xEF, 0xE0, + 0x6, 0xEF, 0x50, 0xF, 0xF4, 0x30, 0xF, 0xE9, 0x40, + 0xE, 0xE5, 0xE0, 0x6, 0xE6, 0x20, 0xF, 0xE1, 0xD0, + 0xE, 0xE5, 0x30, 0xA3, 0x40, 0x83, 0x76, 0x68, + 0x7, 0x79, 0x8, 0x87, 0x19, 0x52, 0x1A, 0xB8, + 0xC3, 0x3B, 0x84, 0x3, 0x3B, 0xA4, 0x43, 0x38, + 0xCC, 0x83, 0x1B, 0x84, 0x3, 0x39, 0x90, 0x83, + 0x3C, 0xCC, 0x3, 0x3C, 0x84, 0xC3, 0x38, 0x94, + 0xC3, 0xC, 0x46, 0xD, 0xC6, 0x21, 0x1C, 0xD8, + 0x81, 0x1D, 0xCA, 0xA1, 0x1C, 0x7E, 0x81, 0x1E, + 0xF2, 0x1, 0x1E, 0xCA, 0x61, 0xC6, 0xB1, 0x6, + 0xEE, 0xF0, 0xE, 0xE6, 0x20, 0xF, 0xE5, 0x50, + 0xE, 0x33, 0x1C, 0x36, 0x20, 0x7, 0x7C, 0x70, + 0x3, 0x77, 0x78, 0x7, 0x77, 0xA8, 0x7, 0x77, 0x48, + 0x7, 0x73, 0x78, 0x7, 0x79, 0x68, 0x87, 0x19, + 0x55, 0x1B, 0x90, 0x3, 0x3E, 0xB8, 0xC1, 0x38, + 0xBC, 0x83, 0x3B, 0xD0, 0x83, 0x3C, 0xBC, 0x3, + 0x3B, 0x98, 0x3, 0x3B, 0xBC, 0xC3, 0x3D, 0xB8, + 0x1, 0x3A, 0xA4, 0x83, 0x3B, 0xD0, 0xC3, 0x3C, + 0xCC, 0x58, 0xDC, 0x80, 0x1C, 0xF0, 0xC1, 0xD, + 0xE0, 0x41, 0x1E, 0xCA, 0x61, 0x1C, 0xD2, 0x61, + 0x1E, 0xCA, 0x1, 0x0, 0x79, 0x28, 0x0, 0x0, 0x52, + 0x0, 0x0, 0x0, 0xC2, 0x3C, 0x90, 0x40, 0x86, 0x10, + 0x19, 0x32, 0xE2, 0x64, 0x90, 0x40, 0x46, 0x2, + 0x19, 0x23, 0x23, 0x46, 0x2, 0x13, 0x24, 0xC6, + 0x0, 0x13, 0x74, 0xCE, 0x61, 0x8C, 0x2D, 0xCC, + 0xED, 0xC, 0xC4, 0xAE, 0x4C, 0x6E, 0x2E, 0xED, + 0xCD, 0xD, 0x44, 0x46, 0xC6, 0x5, 0xC6, 0x5, 0xE6, + 0x2C, 0x8D, 0xE, 0x4, 0xE5, 0x2C, 0x8D, 0xE, 0xE8, + 0x2C, 0x8D, 0xE, 0xAD, 0x4E, 0xCC, 0x65, 0xEC, + 0xAD, 0x4D, 0x87, 0x8D, 0xCD, 0xAE, 0xED, 0x85, + 0x8D, 0xCD, 0xAE, 0xAD, 0x5, 0x4E, 0xEE, 0x4D, + 0xAD, 0x6C, 0x8C, 0xCE, 0xE5, 0x2C, 0x8D, 0xE, + 0xA4, 0xEC, 0xC6, 0x86, 0xA6, 0x2C, 0x26, 0x7, + 0xA6, 0xAC, 0xC, 0x26, 0x26, 0xE7, 0x46, 0x6C, + 0x2C, 0xA6, 0xC, 0x66, 0xA6, 0x6C, 0xC6, 0x4C, + 0x6, 0x86, 0x6C, 0x4C, 0x6, 0x46, 0xCC, 0x66, + 0x6C, 0x2C, 0xC, 0x27, 0x46, 0x6C, 0x86, 0x6C, + 0x2C, 0xE5, 0x8, 0x63, 0x73, 0x87, 0x68, 0xB, + 0x4B, 0x73, 0x3B, 0xCA, 0xDD, 0x18, 0x5A, 0x98, + 0xDC, 0xD7, 0x5C, 0x9A, 0x5E, 0xD9, 0x69, 0xCC, + 0xE4, 0xC2, 0xDA, 0xCA, 0x5A, 0xE0, 0xDE, 0xD2, + 0xDC, 0xE8, 0xCA, 0xE4, 0x86, 0x20, 0x1C, 0xC1, + 0x10, 0x84, 0x43, 0x18, 0x82, 0x70, 0xC, 0x43, + 0x10, 0xE, 0x62, 0x8, 0x42, 0x1, 0xC, 0x41, 0x38, + 0x8A, 0x21, 0x8, 0x87, 0x31, 0x6, 0xC1, 0x38, + 0xC6, 0x10, 0x4, 0x63, 0x18, 0x4, 0x24, 0x19, + 0x83, 0x60, 0x24, 0x63, 0x18, 0xC, 0xC3, 0x18, + 0x83, 0xB0, 0x44, 0x63, 0x28, 0x94, 0x1, 0x0, + 0xA4, 0x31, 0xC, 0x6, 0xB1, 0x8C, 0x61, 0x60, + 0xA, 0xC6, 0x24, 0x64, 0x78, 0x2E, 0x76, 0x61, + 0x6C, 0x76, 0x65, 0x72, 0x43, 0x9, 0x18, 0xA3, + 0xB0, 0xB1, 0xD9, 0xB5, 0xB9, 0xA4, 0x91, 0x95, + 0xB9, 0xD1, 0xD, 0x25, 0x68, 0x8C, 0x43, 0x86, + 0xE7, 0x32, 0x87, 0x16, 0x46, 0x56, 0x26, 0xD7, + 0xF4, 0x46, 0x56, 0xC6, 0x36, 0x94, 0xC0, 0x31, + 0xA, 0x19, 0x9E, 0x8B, 0x5D, 0x99, 0xDC, 0x5C, + 0xDA, 0x9B, 0xDB, 0x50, 0x82, 0xC7, 0x38, 0x64, + 0x78, 0x2E, 0x65, 0x6E, 0x74, 0x72, 0x79, 0x50, + 0x6F, 0x69, 0x6E, 0x74, 0x73, 0x43, 0x9, 0x24, + 0x13, 0xB1, 0xB1, 0xD9, 0xB5, 0xB9, 0xB4, 0xBD, + 0x91, 0xD5, 0xB1, 0x95, 0xB9, 0x98, 0xB1, 0x85, + 0x9D, 0xCD, 0xD, 0x45, 0x98, 0x28, 0x0, 0x71, + 0x20, 0x0, 0x0, 0x2, 0x0, 0x0, 0x0, 0x6, 0x40, + 0x30, 0x0, 0xD2, 0x0, 0x0, 0x0, 0x61, 0x20, 0x0, + 0x0, 0x6, 0x0, 0x0, 0x0, 0x13, 0x4, 0x1, 0x86, + 0x3, 0x1, 0x0, 0x0, 0x2, 0x0, 0x0, 0x0, 0x7, 0x50, + 0x10, 0xCD, 0x14, 0x61, 0x0, 0x0, 0x0, 0x0, 0x0, + 0x0, 0x0, 0x0, 0x0, 0x0 ] + - Name: PSV0 + Size: 76 + PSVInfo: + Version: 3 + ShaderStage: 5 + MinimumWaveLaneCount: 0 + MaximumWaveLaneCount: 4294967295 + UsesViewID: 0 + SigInputVectors: 0 + SigOutputVectors: [ 0, 0, 0, 0 ] + NumThreadsX: 1 + NumThreadsY: 1 + NumThreadsZ: 1 + EntryName: main + ResourceStride: 24 + Resources: [] + SigInputElements: [] + SigOutputElements: [] + SigPatchOrPrimElements: [] + InputOutputMap: + - [ ] + - [ ] + - [ ] + - [ ] +... diff --git a/llvm/test/tools/llvm-readobj/ELF/bb-addr-map.test b/llvm/test/tools/llvm-readobj/ELF/bb-addr-map.test index fd1492f..bcffd40 100644 --- a/llvm/test/tools/llvm-readobj/ELF/bb-addr-map.test +++ b/llvm/test/tools/llvm-readobj/ELF/bb-addr-map.test @@ -34,6 +34,7 @@ # CHECK-NEXT: { # CHECK-NEXT: ID: 0 # CHECK-NEXT: Offset: 0x0 +# CHECK-NEXT: Hash: 0x0 # CHECK-NEXT: Size: 0x1 # CHECK-NEXT: HasReturn: No # CHECK-NEXT: HasTailCall: Yes @@ -50,6 +51,7 @@ # CHECK-NEXT: ID: 2 # CHECK-NEXT: Offset: 0x3 # CHECK-NEXT: Callsite End Offsets: [1, 3] +# CHECK-NEXT: Hash: 0x123 # CHECK-NEXT: Size: 0x7 # CHECK-NEXT: HasReturn: Yes # CHECK-NEXT: HasTailCall: No @@ -144,8 +146,8 @@ Sections: ShSize: [[SIZE=<none>]] Link: .text Entries: - - Version: 3 - Feature: 0x28 + - Version: 4 + Feature: 0x68 BBRanges: - BaseAddress: [[ADDR=0x11111]] BBEntries: @@ -160,6 +162,7 @@ Sections: Size: 0x4 Metadata: 0x15 CallsiteEndOffsets: [ 0x1 , 0x2 ] + Hash: 0x123 - Version: 2 BBRanges: - BaseAddress: 0x22222 diff --git a/llvm/test/tools/obj2yaml/ELF/bb-addr-map.yaml b/llvm/test/tools/obj2yaml/ELF/bb-addr-map.yaml index dc14025..7a22efe 100644 --- a/llvm/test/tools/obj2yaml/ELF/bb-addr-map.yaml +++ b/llvm/test/tools/obj2yaml/ELF/bb-addr-map.yaml @@ -162,6 +162,92 @@ Sections: BBRanges: - BaseAddress: 0x20 +## Check that obj2yaml can dump basic block hash in the .llvm_bb_addr_map section. + +# RUN: yaml2obj --docnum=4 %s -o %t4 +# RUN: obj2yaml %t4 | FileCheck %s --check-prefix=BBHASH + +# BBHASH: --- !ELF +# BBHASH-NEXT: FileHeader: +# BBHASH-NEXT: Class: ELFCLASS64 +# BBHASH-NEXT: Data: ELFDATA2LSB +# BBHASH-NEXT: Type: ET_EXEC +# BBHASH-NEXT: Sections: +# BBHASH-NEXT: - Name: .llvm_bb_addr_map +# BBHASH-NEXT: Type: SHT_LLVM_BB_ADDR_MAP +# BBHASH-NEXT: Entries: +# BBHASH-NEXT: - Version: 4 +# BBHASH-NEXT: Feature: 0x40 +# BBHASH-NEXT: BBRanges: +# BBHASH-NEXT: - BBEntries: +# BBHASH-NEXT: - ID: 0 +# BBHASH-NEXT: AddressOffset: 0x1 +# BBHASH-NEXT: Size: 0x2 +# BBHASH-NEXT: Metadata: 0x3 +# BBHASH-NEXT: Hash: 0x1 +# BBHASH-NEXT: - ID: 2 +# BBHASH-NEXT: AddressOffset: 0x4 +# BBHASH-NEXT: Size: 0x5 +# BBHASH-NEXT: Metadata: 0x6 +# BBHASH-NEXT: Hash: 0x2 +# BBHASH-NEXT: - ID: 4 +# BBHASH-NEXT: AddressOffset: 0xFFFFFFFFFFFFFFF7 +# BBHASH-NEXT: Size: 0xFFFFFFFFFFFFFFF8 +# BBHASH-NEXT: Metadata: 0xFFFFFFFFFFFFFFF9 +# BBHASH-NEXT: Hash: 0x3 +# BBHASH-NEXT: - Version: 4 +# BBHASH-NEXT: Feature: 0x68 +# BBHASH-NEXT: BBRanges: +# BBHASH-NEXT: - BaseAddress: 0xFFFFFFFFFFFFFF20 +# BBHASH-NEXT: BBEntries: +# BBHASH-NEXT: - ID: 6 +# BBHASH-NEXT: AddressOffset: 0xA +# BBHASH-NEXT: Size: 0xB +# BBHASH-NEXT: Metadata: 0xC +# BBHASH-NEXT: CallsiteEndOffsets: [ 0x1, 0x2 ] +# BBHASH-NEXT: Hash: 0x123 + +--- !ELF +FileHeader: + Class: ELFCLASS64 + Data: ELFDATA2LSB + Type: ET_EXEC +Sections: + - Name: .llvm_bb_addr_map + Type: SHT_LLVM_BB_ADDR_MAP + Entries: + - Version: 4 + Feature: 0x40 + BBRanges: + - BaseAddress: 0x0 + BBEntries: + - ID: 0 + AddressOffset: 0x1 + Size: 0x2 + Metadata: 0x3 + Hash: 0x1 + - ID: 2 + AddressOffset: 0x4 + Size: 0x5 + Metadata: 0x6 + Hash: 0x2 + - ID: 4 + AddressOffset: 0xFFFFFFFFFFFFFFF7 + Size: 0xFFFFFFFFFFFFFFF8 + Metadata: 0xFFFFFFFFFFFFFFF9 + Hash: 0x3 + - Version: 4 + Feature: 0x68 + BBRanges: + - BaseAddress: 0xFFFFFFFFFFFFFF20 + BBEntries: + - ID: 6 + AddressOffset: 0xA + Size: 0xB + Metadata: 0xC + CallsiteEndOffsets: [ 0x1, 0x2 ] + Hash: 0x123 + ## Check that obj2yaml uses the "Content" tag to describe an .llvm_bb_addr_map section ## when it can't extract the entries, for example, when the section is truncated, or ## when an invalid 'NumBlocks' or 'NumBBRanges` field is specified. diff --git a/llvm/test/tools/yaml2obj/ELF/bb-addr-map.yaml b/llvm/test/tools/yaml2obj/ELF/bb-addr-map.yaml index 418f90f..339e419 100644 --- a/llvm/test/tools/yaml2obj/ELF/bb-addr-map.yaml +++ b/llvm/test/tools/yaml2obj/ELF/bb-addr-map.yaml @@ -72,6 +72,13 @@ # CHECK-NEXT: 0000: 03202000 00000000 0000010E 01000203 # CHECK-NEXT: ) +# Case 10: Specify basic block hash. +# CHECK: Name: .llvm_bb_addr_map (1) +# CHECK: SectionData ( +# CHECK-NEXT: 0000: 04602000 00000000 0000010E 01000203 +# CHECK-NEXT: 0010: 23010000 00000000 +# CHECK-NEXT: ) + --- !ELF FileHeader: @@ -176,6 +183,22 @@ Sections: Metadata: 0x00000003 CallsiteEndOffsets: [] +## 10) We can produce a SHT_LLVM_BB_ADDR_MAP section with basic block hash. + - Name: '.llvm_bb_addr_map (10)' + Type: SHT_LLVM_BB_ADDR_MAP + Entries: + - Version: 4 + Feature: 0x60 + BBRanges: + - BaseAddress: 0x0000000000000020 + BBEntries: + - ID: 14 + AddressOffset: 0x00000001 + Size: 0x00000002 + Metadata: 0x00000003 + CallsiteEndOffsets: [] + Hash: 0x123 + ## Check we can't use Entries at the same time as either Content or Size. # RUN: not yaml2obj --docnum=2 -DCONTENT="00" %s 2>&1 | FileCheck %s --check-prefix=INVALID # RUN: not yaml2obj --docnum=2 -DSIZE="0" %s 2>&1 | FileCheck %s --check-prefix=INVALID @@ -197,7 +220,7 @@ Sections: ## Check that yaml2obj generates a warning when we use unsupported versions. # RUN: yaml2obj --docnum=3 %s 2>&1 | FileCheck %s --check-prefix=INVALID-VERSION -# INVALID-VERSION: warning: unsupported SHT_LLVM_BB_ADDR_MAP version: 4; encoding using the most recent version +# INVALID-VERSION: warning: unsupported SHT_LLVM_BB_ADDR_MAP version: 5; encoding using the most recent version --- !ELF FileHeader: @@ -209,4 +232,4 @@ Sections: Type: SHT_LLVM_BB_ADDR_MAP Entries: ## Specify unsupported version - - Version: 4 + - Version: 5 diff --git a/llvm/tools/llc/llc.cpp b/llvm/tools/llc/llc.cpp index f04b256..8b03db3 100644 --- a/llvm/tools/llc/llc.cpp +++ b/llvm/tools/llc/llc.cpp @@ -218,13 +218,12 @@ static cl::opt<std::string> PassPipeline( static cl::alias PassPipeline2("p", cl::aliasopt(PassPipeline), cl::desc("Alias for -passes")); -namespace { - -std::vector<std::string> &getRunPassNames() { +static std::vector<std::string> &getRunPassNames() { static std::vector<std::string> RunPassNames; return RunPassNames; } +namespace { struct RunPassOption { void operator=(const std::string &Val) const { if (Val.empty()) diff --git a/llvm/tools/lli/lli.cpp b/llvm/tools/lli/lli.cpp index 875ec1b..7fee06b 100644 --- a/llvm/tools/lli/lli.cpp +++ b/llvm/tools/lli/lli.cpp @@ -92,206 +92,202 @@ static codegen::RegisterCodeGenFlags CGF; #define DEBUG_TYPE "lli" namespace { - - enum class JITKind { MCJIT, Orc, OrcLazy }; - enum class JITLinkerKind { Default, RuntimeDyld, JITLink }; - - cl::opt<std::string> - InputFile(cl::desc("<input bitcode>"), cl::Positional, cl::init("-")); - - cl::list<std::string> - InputArgv(cl::ConsumeAfter, cl::desc("<program arguments>...")); - - cl::opt<bool> ForceInterpreter("force-interpreter", - cl::desc("Force interpretation: disable JIT"), - cl::init(false)); - - cl::opt<JITKind> UseJITKind( - "jit-kind", cl::desc("Choose underlying JIT kind."), - cl::init(JITKind::Orc), - cl::values(clEnumValN(JITKind::MCJIT, "mcjit", "MCJIT"), - clEnumValN(JITKind::Orc, "orc", "Orc JIT"), - clEnumValN(JITKind::OrcLazy, "orc-lazy", - "Orc-based lazy JIT."))); - - cl::opt<JITLinkerKind> - JITLinker("jit-linker", cl::desc("Choose the dynamic linker/loader."), - cl::init(JITLinkerKind::Default), - cl::values(clEnumValN(JITLinkerKind::Default, "default", - "Default for platform and JIT-kind"), - clEnumValN(JITLinkerKind::RuntimeDyld, "rtdyld", - "RuntimeDyld"), - clEnumValN(JITLinkerKind::JITLink, "jitlink", - "Orc-specific linker"))); - cl::opt<std::string> OrcRuntime("orc-runtime", - cl::desc("Use ORC runtime from given path"), - cl::init("")); - - cl::opt<unsigned> - LazyJITCompileThreads("compile-threads", - cl::desc("Choose the number of compile threads " - "(jit-kind=orc-lazy only)"), - cl::init(0)); - - cl::list<std::string> - ThreadEntryPoints("thread-entry", - cl::desc("calls the given entry-point on a new thread " - "(jit-kind=orc-lazy only)")); - - cl::opt<bool> PerModuleLazy( - "per-module-lazy", - cl::desc("Performs lazy compilation on whole module boundaries " - "rather than individual functions"), - cl::init(false)); - - cl::list<std::string> - JITDylibs("jd", - cl::desc("Specifies the JITDylib to be used for any subsequent " - "-extra-module arguments.")); - - cl::list<std::string> - Dylibs("dlopen", cl::desc("Dynamic libraries to load before linking")); - - // The MCJIT supports building for a target address space separate from - // the JIT compilation process. Use a forked process and a copying - // memory manager with IPC to execute using this functionality. - cl::opt<bool> RemoteMCJIT("remote-mcjit", - cl::desc("Execute MCJIT'ed code in a separate process."), +enum class JITKind { MCJIT, Orc, OrcLazy }; +enum class JITLinkerKind { Default, RuntimeDyld, JITLink }; +} // namespace + +static cl::opt<std::string> InputFile(cl::desc("<input bitcode>"), + cl::Positional, cl::init("-")); + +static cl::list<std::string> InputArgv(cl::ConsumeAfter, + cl::desc("<program arguments>...")); + +static cl::opt<bool> + ForceInterpreter("force-interpreter", + cl::desc("Force interpretation: disable JIT"), + cl::init(false)); + +static cl::opt<JITKind> + UseJITKind("jit-kind", cl::desc("Choose underlying JIT kind."), + cl::init(JITKind::Orc), + cl::values(clEnumValN(JITKind::MCJIT, "mcjit", "MCJIT"), + clEnumValN(JITKind::Orc, "orc", "Orc JIT"), + clEnumValN(JITKind::OrcLazy, "orc-lazy", + "Orc-based lazy JIT."))); + +static cl::opt<JITLinkerKind> JITLinker( + "jit-linker", cl::desc("Choose the dynamic linker/loader."), + cl::init(JITLinkerKind::Default), + cl::values(clEnumValN(JITLinkerKind::Default, "default", + "Default for platform and JIT-kind"), + clEnumValN(JITLinkerKind::RuntimeDyld, "rtdyld", "RuntimeDyld"), + clEnumValN(JITLinkerKind::JITLink, "jitlink", + "Orc-specific linker"))); +static cl::opt<std::string> + OrcRuntime("orc-runtime", cl::desc("Use ORC runtime from given path"), + cl::init("")); + +static cl::opt<unsigned> + LazyJITCompileThreads("compile-threads", + cl::desc("Choose the number of compile threads " + "(jit-kind=orc-lazy only)"), + cl::init(0)); + +static cl::list<std::string> + ThreadEntryPoints("thread-entry", + cl::desc("calls the given entry-point on a new thread " + "(jit-kind=orc-lazy only)")); + +static cl::opt<bool> PerModuleLazy( + "per-module-lazy", + cl::desc("Performs lazy compilation on whole module boundaries " + "rather than individual functions"), cl::init(false)); - // Manually specify the child process for remote execution. This overrides - // the simulated remote execution that allocates address space for child - // execution. The child process will be executed and will communicate with - // lli via stdin/stdout pipes. - cl::opt<std::string> - ChildExecPath("mcjit-remote-process", - cl::desc("Specify the filename of the process to launch " - "for remote MCJIT execution. If none is specified," - "\n\tremote execution will be simulated in-process."), - cl::value_desc("filename"), cl::init("")); - - // Determine optimization level. - cl::opt<char> OptLevel("O", - cl::desc("Optimization level. [-O0, -O1, -O2, or -O3] " - "(default = '-O2')"), - cl::Prefix, cl::init('2')); - - cl::opt<std::string> - TargetTriple("mtriple", cl::desc("Override target triple for module")); - - cl::opt<std::string> - EntryFunc("entry-function", - cl::desc("Specify the entry function (default = 'main') " - "of the executable"), - cl::value_desc("function"), - cl::init("main")); - - cl::list<std::string> - ExtraModules("extra-module", - cl::desc("Extra modules to be loaded"), - cl::value_desc("input bitcode")); - - cl::list<std::string> - ExtraObjects("extra-object", - cl::desc("Extra object files to be loaded"), - cl::value_desc("input object")); - - cl::list<std::string> - ExtraArchives("extra-archive", - cl::desc("Extra archive files to be loaded"), - cl::value_desc("input archive")); - - cl::opt<bool> - EnableCacheManager("enable-cache-manager", - cl::desc("Use cache manager to save/load modules"), - cl::init(false)); - - cl::opt<std::string> - ObjectCacheDir("object-cache-dir", - cl::desc("Directory to store cached object files " - "(must be user writable)"), - cl::init("")); - - cl::opt<std::string> - FakeArgv0("fake-argv0", - cl::desc("Override the 'argv[0]' value passed into the executing" - " program"), cl::value_desc("executable")); - - cl::opt<bool> - DisableCoreFiles("disable-core-files", cl::Hidden, - cl::desc("Disable emission of core files if possible")); - - cl::opt<bool> - NoLazyCompilation("disable-lazy-compilation", - cl::desc("Disable JIT lazy compilation"), - cl::init(false)); - - cl::opt<bool> - GenerateSoftFloatCalls("soft-float", - cl::desc("Generate software floating point library calls"), +static cl::list<std::string> + JITDylibs("jd", + cl::desc("Specifies the JITDylib to be used for any subsequent " + "-extra-module arguments.")); + +static cl::list<std::string> + Dylibs("dlopen", cl::desc("Dynamic libraries to load before linking")); + +// The MCJIT supports building for a target address space separate from +// the JIT compilation process. Use a forked process and a copying +// memory manager with IPC to execute using this functionality. +static cl::opt<bool> + RemoteMCJIT("remote-mcjit", + cl::desc("Execute MCJIT'ed code in a separate process."), + cl::init(false)); + +// Manually specify the child process for remote execution. This overrides +// the simulated remote execution that allocates address space for child +// execution. The child process will be executed and will communicate with +// lli via stdin/stdout pipes. +static cl::opt<std::string> ChildExecPath( + "mcjit-remote-process", + cl::desc("Specify the filename of the process to launch " + "for remote MCJIT execution. If none is specified," + "\n\tremote execution will be simulated in-process."), + cl::value_desc("filename"), cl::init("")); + +// Determine optimization level. +static cl::opt<char> + OptLevel("O", + cl::desc("Optimization level. [-O0, -O1, -O2, or -O3] " + "(default = '-O2')"), + cl::Prefix, cl::init('2')); + +static cl::opt<std::string> + TargetTriple("mtriple", cl::desc("Override target triple for module")); + +static cl::opt<std::string> + EntryFunc("entry-function", + cl::desc("Specify the entry function (default = 'main') " + "of the executable"), + cl::value_desc("function"), cl::init("main")); + +static cl::list<std::string> + ExtraModules("extra-module", cl::desc("Extra modules to be loaded"), + cl::value_desc("input bitcode")); + +static cl::list<std::string> + ExtraObjects("extra-object", cl::desc("Extra object files to be loaded"), + cl::value_desc("input object")); + +static cl::list<std::string> + ExtraArchives("extra-archive", cl::desc("Extra archive files to be loaded"), + cl::value_desc("input archive")); + +static cl::opt<bool> + EnableCacheManager("enable-cache-manager", + cl::desc("Use cache manager to save/load modules"), + cl::init(false)); + +static cl::opt<std::string> + ObjectCacheDir("object-cache-dir", + cl::desc("Directory to store cached object files " + "(must be user writable)"), + cl::init("")); + +static cl::opt<std::string> + FakeArgv0("fake-argv0", + cl::desc("Override the 'argv[0]' value passed into the executing" + " program"), + cl::value_desc("executable")); + +static cl::opt<bool> + DisableCoreFiles("disable-core-files", cl::Hidden, + cl::desc("Disable emission of core files if possible")); + +static cl::opt<bool> NoLazyCompilation("disable-lazy-compilation", + cl::desc("Disable JIT lazy compilation"), + cl::init(false)); + +static cl::opt<bool> GenerateSoftFloatCalls( + "soft-float", cl::desc("Generate software floating point library calls"), cl::init(false)); - cl::opt<bool> NoProcessSymbols( - "no-process-syms", - cl::desc("Do not resolve lli process symbols in JIT'd code"), - cl::init(false)); - - enum class LLJITPlatform { Inactive, Auto, ExecutorNative, GenericIR }; - - cl::opt<LLJITPlatform> Platform( - "lljit-platform", cl::desc("Platform to use with LLJIT"), - cl::init(LLJITPlatform::Auto), - cl::values(clEnumValN(LLJITPlatform::Auto, "Auto", - "Like 'ExecutorNative' if ORC runtime " - "provided, otherwise like 'GenericIR'"), - clEnumValN(LLJITPlatform::ExecutorNative, "ExecutorNative", - "Use the native platform for the executor." - "Requires -orc-runtime"), - clEnumValN(LLJITPlatform::GenericIR, "GenericIR", - "Use LLJITGenericIRPlatform"), - clEnumValN(LLJITPlatform::Inactive, "Inactive", - "Disable platform support explicitly")), - cl::Hidden); - - enum class DumpKind { - NoDump, - DumpFuncsToStdOut, - DumpModsToStdOut, - DumpModsToDisk, - DumpDebugDescriptor, - DumpDebugObjects, - }; +static cl::opt<bool> NoProcessSymbols( + "no-process-syms", + cl::desc("Do not resolve lli process symbols in JIT'd code"), + cl::init(false)); - cl::opt<DumpKind> OrcDumpKind( - "orc-lazy-debug", cl::desc("Debug dumping for the orc-lazy JIT."), - cl::init(DumpKind::NoDump), - cl::values( - clEnumValN(DumpKind::NoDump, "no-dump", "Don't dump anything."), - clEnumValN(DumpKind::DumpFuncsToStdOut, "funcs-to-stdout", - "Dump function names to stdout."), - clEnumValN(DumpKind::DumpModsToStdOut, "mods-to-stdout", - "Dump modules to stdout."), - clEnumValN(DumpKind::DumpModsToDisk, "mods-to-disk", - "Dump modules to the current " - "working directory. (WARNING: " - "will overwrite existing files)."), - clEnumValN(DumpKind::DumpDebugDescriptor, "jit-debug-descriptor", - "Dump __jit_debug_descriptor contents to stdout"), - clEnumValN(DumpKind::DumpDebugObjects, "jit-debug-objects", - "Dump __jit_debug_descriptor in-memory debug " - "objects as tool output")), - cl::Hidden); - - ExitOnError ExitOnErr; -} +enum class LLJITPlatform { Inactive, Auto, ExecutorNative, GenericIR }; + +static cl::opt<LLJITPlatform> Platform( + "lljit-platform", cl::desc("Platform to use with LLJIT"), + cl::init(LLJITPlatform::Auto), + cl::values(clEnumValN(LLJITPlatform::Auto, "Auto", + "Like 'ExecutorNative' if ORC runtime " + "provided, otherwise like 'GenericIR'"), + clEnumValN(LLJITPlatform::ExecutorNative, "ExecutorNative", + "Use the native platform for the executor." + "Requires -orc-runtime"), + clEnumValN(LLJITPlatform::GenericIR, "GenericIR", + "Use LLJITGenericIRPlatform"), + clEnumValN(LLJITPlatform::Inactive, "Inactive", + "Disable platform support explicitly")), + cl::Hidden); + +enum class DumpKind { + NoDump, + DumpFuncsToStdOut, + DumpModsToStdOut, + DumpModsToDisk, + DumpDebugDescriptor, + DumpDebugObjects, +}; -LLVM_ATTRIBUTE_USED void linkComponents() { +static cl::opt<DumpKind> OrcDumpKind( + "orc-lazy-debug", cl::desc("Debug dumping for the orc-lazy JIT."), + cl::init(DumpKind::NoDump), + cl::values(clEnumValN(DumpKind::NoDump, "no-dump", "Don't dump anything."), + clEnumValN(DumpKind::DumpFuncsToStdOut, "funcs-to-stdout", + "Dump function names to stdout."), + clEnumValN(DumpKind::DumpModsToStdOut, "mods-to-stdout", + "Dump modules to stdout."), + clEnumValN(DumpKind::DumpModsToDisk, "mods-to-disk", + "Dump modules to the current " + "working directory. (WARNING: " + "will overwrite existing files)."), + clEnumValN(DumpKind::DumpDebugDescriptor, "jit-debug-descriptor", + "Dump __jit_debug_descriptor contents to stdout"), + clEnumValN(DumpKind::DumpDebugObjects, "jit-debug-objects", + "Dump __jit_debug_descriptor in-memory debug " + "objects as tool output")), + cl::Hidden); + +static ExitOnError ExitOnErr; + +LLVM_ATTRIBUTE_USED static void linkComponents() { errs() << (void *)&llvm_orc_registerEHFrameSectionAllocAction << (void *)&llvm_orc_deregisterEHFrameSectionAllocAction << (void *)&llvm_orc_registerJITLoaderGDBWrapper << (void *)&llvm_orc_registerJITLoaderGDBAllocAction; } +namespace { //===----------------------------------------------------------------------===// // Object cache // @@ -367,6 +363,7 @@ private: return true; } }; +} // namespace // On Mingw and Cygwin, an external symbol named '__main' is called from the // generated 'main' function to allow static initialization. To avoid linking @@ -400,7 +397,7 @@ static void addCygMingExtraModule(ExecutionEngine &EE, LLVMContext &Context, EE.addModule(std::move(M)); } -CodeGenOptLevel getOptLevel() { +static CodeGenOptLevel getOptLevel() { if (auto Level = CodeGenOpt::parseLevel(OptLevel)) return *Level; WithColor::error(errs(), "lli") << "invalid optimization level.\n"; @@ -412,10 +409,10 @@ CodeGenOptLevel getOptLevel() { exit(1); } -Error loadDylibs(); -int runOrcJIT(const char *ProgName); -void disallowOrcOptions(); -Expected<std::unique_ptr<orc::ExecutorProcessControl>> launchRemote(); +static Error loadDylibs(); +static int runOrcJIT(const char *ProgName); +static void disallowOrcOptions(); +static Expected<std::unique_ptr<orc::ExecutorProcessControl>> launchRemote(); //===----------------------------------------------------------------------===// // main Driver function @@ -863,7 +860,7 @@ static std::function<void(MemoryBuffer &)> createObjDebugDumper() { llvm_unreachable("Unknown DumpKind"); } -Error loadDylibs() { +static Error loadDylibs() { for (const auto &Dylib : Dylibs) { std::string ErrMsg; if (sys::DynamicLibrary::LoadLibraryPermanently(Dylib.c_str(), &ErrMsg)) @@ -875,7 +872,7 @@ Error loadDylibs() { static void exitOnLazyCallThroughFailure() { exit(1); } -Expected<orc::ThreadSafeModule> +static Expected<orc::ThreadSafeModule> loadModule(StringRef Path, orc::ThreadSafeContext TSCtx) { SMDiagnostic Err; auto M = TSCtx.withContextDo( @@ -895,7 +892,7 @@ loadModule(StringRef Path, orc::ThreadSafeContext TSCtx) { return orc::ThreadSafeModule(std::move(M), std::move(TSCtx)); } -int mingw_noop_main(void) { +static int mingw_noop_main(void) { // Cygwin and MinGW insert calls from the main function to the runtime // function __main. The __main function is responsible for setting up main's // environment (e.g. running static constructors), however this is not needed @@ -912,7 +909,7 @@ int mingw_noop_main(void) { // Try to enable debugger support for the given instance. // This alway returns success, but prints a warning if it's not able to enable // debugger support. -Error tryEnableDebugSupport(orc::LLJIT &J) { +static Error tryEnableDebugSupport(orc::LLJIT &J) { if (auto Err = enableDebuggerSupport(J)) { [[maybe_unused]] std::string ErrMsg = toString(std::move(Err)); LLVM_DEBUG(dbgs() << "lli: " << ErrMsg << "\n"); @@ -920,7 +917,7 @@ Error tryEnableDebugSupport(orc::LLJIT &J) { return Error::success(); } -int runOrcJIT(const char *ProgName) { +static int runOrcJIT(const char *ProgName) { // Start setting up the JIT environment. // Parse the main module. @@ -1187,7 +1184,7 @@ int runOrcJIT(const char *ProgName) { return Result; } -void disallowOrcOptions() { +static void disallowOrcOptions() { // Make sure nobody used an orc-lazy specific option accidentally. if (LazyJITCompileThreads != 0) { @@ -1206,7 +1203,7 @@ void disallowOrcOptions() { } } -Expected<std::unique_ptr<orc::ExecutorProcessControl>> launchRemote() { +static Expected<std::unique_ptr<orc::ExecutorProcessControl>> launchRemote() { #ifndef LLVM_ON_UNIX llvm_unreachable("launchRemote not supported on non-Unix platforms"); #else diff --git a/llvm/tools/llvm-gsymutil/llvm-gsymutil.cpp b/llvm/tools/llvm-gsymutil/llvm-gsymutil.cpp index de83a0d..4c08b57 100644 --- a/llvm/tools/llvm-gsymutil/llvm-gsymutil.cpp +++ b/llvm/tools/llvm-gsymutil/llvm-gsymutil.cpp @@ -386,7 +386,9 @@ static llvm::Error handleObjectFile(ObjectFile &Obj, const std::string &OutFile, // Make a DWARF transformer object and populate the ranges of the code // so we don't end up adding invalid functions to GSYM data. - DwarfTransformer DT(*DICtx, Gsym, LoadDwarfCallSites); + bool IsMachO = dyn_cast<object::MachOObjectFile>(&Obj) != nullptr; + + DwarfTransformer DT(*DICtx, Gsym, LoadDwarfCallSites, IsMachO); if (!TextRanges.empty()) Gsym.SetValidTextRanges(TextRanges); diff --git a/llvm/tools/llvm-mc-assemble-fuzzer/llvm-mc-assemble-fuzzer.cpp b/llvm/tools/llvm-mc-assemble-fuzzer/llvm-mc-assemble-fuzzer.cpp index fa56d0d..fb5c0bf 100644 --- a/llvm/tools/llvm-mc-assemble-fuzzer/llvm-mc-assemble-fuzzer.cpp +++ b/llvm/tools/llvm-mc-assemble-fuzzer/llvm-mc-assemble-fuzzer.cpp @@ -31,6 +31,7 @@ #include "llvm/Support/SourceMgr.h" #include "llvm/Support/TargetSelect.h" #include "llvm/Support/ToolOutputFile.h" +#include "llvm/Support/VirtualFileSystem.h" #include "llvm/Support/raw_ostream.h" #include "llvm/TargetParser/Host.h" #include "llvm/TargetParser/SubtargetFeature.h" @@ -142,6 +143,7 @@ int AssembleOneInput(const uint8_t *Data, size_t Size) { static const std::vector<std::string> NoIncludeDirs; SrcMgr.setIncludeDirs(NoIncludeDirs); + SrcMgr.setVirtualFileSystem(vfs::getRealFileSystem()); static std::string ArchName; std::string Error; diff --git a/llvm/tools/llvm-mc/llvm-mc.cpp b/llvm/tools/llvm-mc/llvm-mc.cpp index 2a89961..3b2d4f8 100644 --- a/llvm/tools/llvm-mc/llvm-mc.cpp +++ b/llvm/tools/llvm-mc/llvm-mc.cpp @@ -40,6 +40,7 @@ #include "llvm/Support/TargetSelect.h" #include "llvm/Support/TimeProfiler.h" #include "llvm/Support/ToolOutputFile.h" +#include "llvm/Support/VirtualFileSystem.h" #include "llvm/Support/WithColor.h" #include "llvm/TargetParser/Host.h" #include <memory> @@ -439,6 +440,7 @@ int main(int argc, char **argv) { // Record the location of the include directories so that the lexer can find // it later. SrcMgr.setIncludeDirs(IncludeDirs); + SrcMgr.setVirtualFileSystem(vfs::getRealFileSystem()); std::unique_ptr<MCRegisterInfo> MRI(TheTarget->createMCRegInfo(TheTriple)); assert(MRI && "Unable to create target register info!"); diff --git a/llvm/tools/llvm-ml/llvm-ml.cpp b/llvm/tools/llvm-ml/llvm-ml.cpp index cda86e7..7b88576 100644 --- a/llvm/tools/llvm-ml/llvm-ml.cpp +++ b/llvm/tools/llvm-ml/llvm-ml.cpp @@ -41,6 +41,7 @@ #include "llvm/Support/SourceMgr.h" #include "llvm/Support/TargetSelect.h" #include "llvm/Support/ToolOutputFile.h" +#include "llvm/Support/VirtualFileSystem.h" #include "llvm/Support/WithColor.h" #include "llvm/TargetParser/Host.h" #include <ctime> @@ -313,6 +314,7 @@ int llvm_ml_main(int Argc, char **Argv, const llvm::ToolContext &) { } } SrcMgr.setIncludeDirs(IncludeDirs); + SrcMgr.setVirtualFileSystem(vfs::getRealFileSystem()); std::unique_ptr<MCRegisterInfo> MRI(TheTarget->createMCRegInfo(TheTriple)); assert(MRI && "Unable to create target register info!"); diff --git a/llvm/tools/llvm-readobj/ELFDumper.cpp b/llvm/tools/llvm-readobj/ELFDumper.cpp index ab93316..9c9b2dd 100644 --- a/llvm/tools/llvm-readobj/ELFDumper.cpp +++ b/llvm/tools/llvm-readobj/ELFDumper.cpp @@ -8155,6 +8155,8 @@ void LLVMELFDumper<ELFT>::printBBAddrMaps(bool PrettyPGOAnalysis) { W.printHex("Offset", BBE.Offset); if (!BBE.CallsiteEndOffsets.empty()) W.printList("Callsite End Offsets", BBE.CallsiteEndOffsets); + if (PAM.FeatEnable.BBHash) + W.printHex("Hash", BBE.Hash); W.printHex("Size", BBE.Size); W.printBoolean("HasReturn", BBE.hasReturn()); W.printBoolean("HasTailCall", BBE.hasTailCall()); diff --git a/llvm/tools/obj2yaml/elf2yaml.cpp b/llvm/tools/obj2yaml/elf2yaml.cpp index ef4552f..68e18f6 100644 --- a/llvm/tools/obj2yaml/elf2yaml.cpp +++ b/llvm/tools/obj2yaml/elf2yaml.cpp @@ -900,7 +900,7 @@ ELFDumper<ELFT>::dumpBBAddrMapSection(const Elf_Shdr *Shdr) { while (Cur && Cur.tell() < Content.size()) { if (Shdr->sh_type == ELF::SHT_LLVM_BB_ADDR_MAP) { Version = Data.getU8(Cur); - if (Cur && Version > 3) + if (Cur && Version > 4) return createStringError( errc::invalid_argument, "invalid SHT_LLVM_BB_ADDR_MAP section version: " + @@ -946,8 +946,11 @@ ELFDumper<ELFT>::dumpBBAddrMapSection(const Elf_Shdr *Shdr) { } uint64_t Size = Data.getULEB128(Cur); uint64_t Metadata = Data.getULEB128(Cur); + std::optional<llvm::yaml::Hex64> Hash; + if (FeatureOrErr->BBHash) + Hash = Data.getU64(Cur); BBEntries.push_back( - {ID, Offset, Size, Metadata, std::move(CallsiteEndOffsets)}); + {ID, Offset, Size, Metadata, std::move(CallsiteEndOffsets), Hash}); } TotalNumBlocks += BBEntries.size(); BBRanges.push_back({BaseAddress, /*NumBlocks=*/{}, BBEntries}); diff --git a/llvm/unittests/ADT/SmallVectorTest.cpp b/llvm/unittests/ADT/SmallVectorTest.cpp index 137dd43..e2e778f 100644 --- a/llvm/unittests/ADT/SmallVectorTest.cpp +++ b/llvm/unittests/ADT/SmallVectorTest.cpp @@ -127,24 +127,24 @@ public: return c0.getValue() == c1.getValue(); } - friend bool LLVM_ATTRIBUTE_UNUSED operator!=(const Constructable &c0, - const Constructable &c1) { + [[maybe_unused]] friend bool operator!=(const Constructable &c0, + const Constructable &c1) { return c0.getValue() != c1.getValue(); } friend bool operator<(const Constructable &c0, const Constructable &c1) { return c0.getValue() < c1.getValue(); } - friend bool LLVM_ATTRIBUTE_UNUSED operator<=(const Constructable &c0, - const Constructable &c1) { + [[maybe_unused]] friend bool operator<=(const Constructable &c0, + const Constructable &c1) { return c0.getValue() <= c1.getValue(); } - friend bool LLVM_ATTRIBUTE_UNUSED operator>(const Constructable &c0, - const Constructable &c1) { + [[maybe_unused]] friend bool operator>(const Constructable &c0, + const Constructable &c1) { return c0.getValue() > c1.getValue(); } - friend bool LLVM_ATTRIBUTE_UNUSED operator>=(const Constructable &c0, - const Constructable &c1) { + [[maybe_unused]] friend bool operator>=(const Constructable &c0, + const Constructable &c1) { return c0.getValue() >= c1.getValue(); } }; diff --git a/llvm/unittests/ADT/TypeTraitsTest.cpp b/llvm/unittests/ADT/TypeTraitsTest.cpp index a56aa7e..f9b8d6d 100644 --- a/llvm/unittests/ADT/TypeTraitsTest.cpp +++ b/llvm/unittests/ADT/TypeTraitsTest.cpp @@ -40,9 +40,7 @@ struct Foo { struct CheckMethodPointer : CheckFunctionTraits<decltype(&Foo::func)> {}; /// Test lambda references. -LLVM_ATTRIBUTE_UNUSED auto lambdaFunc = [](const int &v) -> bool { - return true; -}; +[[maybe_unused]] auto lambdaFunc = [](const int &v) -> bool { return true; }; struct CheckLambda : CheckFunctionTraits<decltype(lambdaFunc)> {}; } // end anonymous namespace diff --git a/llvm/unittests/DebugInfo/GSYM/GSYMTest.cpp b/llvm/unittests/DebugInfo/GSYM/GSYMTest.cpp index 33f53de..d560073 100644 --- a/llvm/unittests/DebugInfo/GSYM/GSYMTest.cpp +++ b/llvm/unittests/DebugInfo/GSYM/GSYMTest.cpp @@ -4899,3 +4899,189 @@ TEST(GSYMTest, TestLookupsOfOverlappingAndUnequalRanges) { for (const auto &Line : ExpectedDumpLines) EXPECT_TRUE(DumpStr.find(Line) != std::string::npos); } + +TEST(GSYMTest, TestUnableToLocateDWO) { + // Test that llvm-gsymutil will not produce "uanble to locate DWO file" for + // Apple binaries. Apple uses DW_AT_GNU_dwo_id for non split DWARF purposes + // and this makes llvm-gsymutil create warnings and errors. + // + // 0x0000000b: DW_TAG_compile_unit + // DW_AT_name ("main.cpp") + // DW_AT_language (DW_LANG_C) + // DW_AT_GNU_dwo_id (0xfffffffe) + StringRef yamldata = R"( + debug_str: + - '' + - main.cpp + debug_abbrev: + - ID: 0 + Table: + - Code: 0x1 + Tag: DW_TAG_compile_unit + Children: DW_CHILDREN_no + Attributes: + - Attribute: DW_AT_name + Form: DW_FORM_strp + - Attribute: DW_AT_language + Form: DW_FORM_udata + - Attribute: DW_AT_GNU_dwo_id + Form: DW_FORM_data4 + debug_info: + - Length: 0x11 + Version: 4 + AbbrevTableID: 0 + AbbrOffset: 0x0 + AddrSize: 8 + Entries: + - AbbrCode: 0x1 + Values: + - Value: 0x1 + - Value: 0x2 + - Value: 0xFFFFFFFE + )"; + auto ErrOrSections = DWARFYAML::emitDebugSections(yamldata); + ASSERT_THAT_EXPECTED(ErrOrSections, Succeeded()); + std::unique_ptr<DWARFContext> DwarfContext = + DWARFContext::create(*ErrOrSections, 8); + ASSERT_TRUE(DwarfContext.get() != nullptr); + std::string errors; + raw_string_ostream OS(errors); + OutputAggregator OSAgg(&OS); + GsymCreator GC; + // Make a DWARF transformer that is MachO (Apple) to avoid warnings about + // not finding DWO files. + DwarfTransformer DT(*DwarfContext, GC, /*LDCS=*/false, /*MachO*/ true); + const uint32_t ThreadCount = 1; + ASSERT_THAT_ERROR(DT.convert(ThreadCount, OSAgg), Succeeded()); + ASSERT_THAT_ERROR(GC.finalize(OSAgg), Succeeded()); + + // Make sure this warning is not in the binary + std::string warn("warning: Unable to retrieve DWO .debug_info section for"); + EXPECT_TRUE(errors.find(warn) == std::string::npos); +} + +TEST(GSYMTest, TestDWARFTransformNoErrorForMissingFileDecl) { + // Test that if llvm-gsymutil finds a line table for a compile unit and if + // there are no matching entries for a function in that compile unit, that + // it doesn't print out a error saying that a DIE has an invalid file index + // if there is no DW_AT_decl_file attribute. + // + // 0x0000000b: DW_TAG_compile_unit + // DW_AT_name ("main.cpp") + // DW_AT_language (DW_LANG_C) + // DW_AT_stmt_list (0x00000000) + // + // 0x00000015: DW_TAG_subprogram + // DW_AT_name ("foo") + // DW_AT_low_pc (0x0000000000001000) + // DW_AT_high_pc (0x0000000000001050) + // + // 0x0000002a: NULL + // + // Line table that has entries, but none that match "foo": + // + // Address Line Column File ISA Discriminator OpIndex Flags + // ------------------ ------ ------ ------ --- ------------- ------- ----- + // 0x0000000000002000 10 0 1 0 0 0 is_stmt + // 0x0000000000002050 13 0 1 0 0 0 is_stmt + + StringRef yamldata = R"( + debug_str: + - '' + - main.cpp + debug_abbrev: + - ID: 0 + Table: + - Code: 0x1 + Tag: DW_TAG_compile_unit + Children: DW_CHILDREN_yes + Attributes: + - Attribute: DW_AT_name + Form: DW_FORM_strp + - Attribute: DW_AT_language + Form: DW_FORM_udata + - Attribute: DW_AT_stmt_list + Form: DW_FORM_sec_offset + - Code: 0x2 + Tag: DW_TAG_subprogram + Children: DW_CHILDREN_no + Attributes: + - Attribute: DW_AT_name + Form: DW_FORM_string + - Attribute: DW_AT_low_pc + Form: DW_FORM_addr + - Attribute: DW_AT_high_pc + Form: DW_FORM_addr + debug_info: + - Length: 0x27 + Version: 4 + AbbrevTableID: 0 + AbbrOffset: 0x0 + AddrSize: 8 + Entries: + - AbbrCode: 0x1 + Values: + - Value: 0x1 + - Value: 0x2 + - Value: 0x0 + - AbbrCode: 0x2 + Values: + - Value: 0xDEADBEEFDEADBEEF + CStr: foo + - Value: 0x1000 + - Value: 0x1050 + - AbbrCode: 0x0 + debug_line: + - Length: 58 + Version: 2 + PrologueLength: 31 + MinInstLength: 1 + DefaultIsStmt: 1 + LineBase: 251 + LineRange: 14 + OpcodeBase: 13 + StandardOpcodeLengths: [ 0, 1, 1, 1, 1, 0, 0, 0, 1, 0, 0, 1 ] + Files: + - Name: main.cpp + DirIdx: 0 + ModTime: 0 + Length: 0 + Opcodes: + - Opcode: DW_LNS_extended_op + ExtLen: 9 + SubOpcode: DW_LNE_set_address + Data: 8192 + - Opcode: DW_LNS_advance_line + SData: 9 + Data: 0 + - Opcode: DW_LNS_copy + Data: 0 + - Opcode: DW_LNS_advance_pc + Data: 80 + - Opcode: DW_LNS_advance_line + SData: 3 + Data: 0 + - Opcode: DW_LNS_extended_op + ExtLen: 1 + SubOpcode: DW_LNE_end_sequence + Data: 0 + )"; + auto ErrOrSections = DWARFYAML::emitDebugSections(yamldata); + ASSERT_THAT_EXPECTED(ErrOrSections, Succeeded()); + std::unique_ptr<DWARFContext> DwarfContext = + DWARFContext::create(*ErrOrSections, 8); + ASSERT_TRUE(DwarfContext.get() != nullptr); + std::string errors; + raw_string_ostream OS(errors); + OutputAggregator OSAgg(&OS); + GsymCreator GC; + DwarfTransformer DT(*DwarfContext, GC); + const uint32_t ThreadCount = 1; + ASSERT_THAT_ERROR(DT.convert(ThreadCount, OSAgg), Succeeded()); + ASSERT_THAT_ERROR(GC.finalize(OSAgg), Succeeded()); + + // Make sure this warning is not in the binary + std::string error_str("error: function DIE at 0x00000015 has an invalid file " + "index 4294967295 in its DW_AT_decl_file attribute"); + EXPECT_TRUE(errors.find(error_str) == std::string::npos); +} diff --git a/llvm/unittests/ExecutionEngine/Orc/ReOptimizeLayerTest.cpp b/llvm/unittests/ExecutionEngine/Orc/ReOptimizeLayerTest.cpp index cd10ffe..f35a378 100644 --- a/llvm/unittests/ExecutionEngine/Orc/ReOptimizeLayerTest.cpp +++ b/llvm/unittests/ExecutionEngine/Orc/ReOptimizeLayerTest.cpp @@ -9,6 +9,7 @@ #include "llvm/ExecutionEngine/Orc/IRTransformLayer.h" #include "llvm/ExecutionEngine/Orc/JITLinkRedirectableSymbolManager.h" #include "llvm/ExecutionEngine/Orc/JITTargetMachineBuilder.h" +#include "llvm/ExecutionEngine/Orc/MapperJITLinkMemoryManager.h" #include "llvm/ExecutionEngine/Orc/ObjectLinkingLayer.h" #include "llvm/ExecutionEngine/Orc/ObjectTransformLayer.h" #include "llvm/ExecutionEngine/Orc/SelfExecutorProcessControl.h" @@ -84,8 +85,11 @@ protected: ES = std::make_unique<ExecutionSession>(std::move(*EPC)); JD = &ES->createBareJITDylib("main"); + ObjLinkingLayer = std::make_unique<ObjectLinkingLayer>( - *ES, std::make_unique<InProcessMemoryManager>(*PageSize)); + *ES, std::make_unique<MapperJITLinkMemoryManager>( + 10 * 1024 * 1024, + std::make_unique<InProcessMemoryMapper>(*PageSize))); DL = std::make_unique<DataLayout>(std::move(*DLOrErr)); auto TM = JTMB->createTargetMachine(); diff --git a/llvm/unittests/Object/ELFObjectFileTest.cpp b/llvm/unittests/Object/ELFObjectFileTest.cpp index 17d9f50..d6a3ca5 100644 --- a/llvm/unittests/Object/ELFObjectFileTest.cpp +++ b/llvm/unittests/Object/ELFObjectFileTest.cpp @@ -531,7 +531,7 @@ Sections: // Check that we can detect unsupported versions. SmallString<128> UnsupportedVersionYamlString(CommonYamlString); UnsupportedVersionYamlString += R"( - - Version: 4 + - Version: 5 BBRanges: - BaseAddress: 0x11111 BBEntries: @@ -543,7 +543,7 @@ Sections: { SCOPED_TRACE("unsupported version"); DoCheck(UnsupportedVersionYamlString, - "unsupported SHT_LLVM_BB_ADDR_MAP version: 4"); + "unsupported SHT_LLVM_BB_ADDR_MAP version: 5"); } SmallString<128> ZeroBBRangesYamlString(CommonYamlString); @@ -761,14 +761,14 @@ Sections: BBAddrMap E1 = { {{0x11111, - {{1, 0x0, 0x3, {false, true, false, false, false}, {0x1, 0x2}}}}}}; + {{1, 0x0, 0x3, {false, true, false, false, false}, {0x1, 0x2}, 0}}}}}; BBAddrMap E2 = { - {{0x22222, {{2, 0x0, 0x2, {false, false, true, false, false}, {}}}}, - {0xFFFFF, {{15, 0xF0, 0xF1, {true, true, true, true, true}, {}}}}}}; + {{0x22222, {{2, 0x0, 0x2, {false, false, true, false, false}, {}, 0}}}, + {0xFFFFF, {{15, 0xF0, 0xF1, {true, true, true, true, true}, {}, 0}}}}}; BBAddrMap E3 = { - {{0x33333, {{0, 0x0, 0x3, {false, true, true, false, false}, {}}}}}}; + {{0x33333, {{0, 0x0, 0x3, {false, true, true, false, false}, {}, 0}}}}}; BBAddrMap E4 = { - {{0x44444, {{0, 0x0, 0x4, {false, false, false, true, true}, {}}}}}}; + {{0x44444, {{0, 0x0, 0x4, {false, false, false, true, true}, {}, 0}}}}}; std::vector<BBAddrMap> Section0BBAddrMaps = {E4}; std::vector<BBAddrMap> Section1BBAddrMaps = {E3}; @@ -988,6 +988,123 @@ Sections: } } +// Test for the ELFObjectFile::readBBAddrMap API with BBHash. +TEST(ELFObjectFileTest, ReadBBHash) { + StringRef CommonYamlString(R"( +--- !ELF +FileHeader: + Class: ELFCLASS64 + Data: ELFDATA2LSB + Type: ET_EXEC +Sections: + - Name: .llvm_bb_addr_map_1 + Type: SHT_LLVM_BB_ADDR_MAP + Link: 1 + Entries: + - Version: 4 + Feature: 0x60 + BBRanges: + - BaseAddress: 0x11111 + BBEntries: + - ID: 1 + AddressOffset: 0x0 + Size: 0x1 + Metadata: 0x2 + CallsiteEndOffsets: [ 0x1 , 0x1 ] + Hash: 0x1 + - Name: .llvm_bb_addr_map_2 + Type: SHT_LLVM_BB_ADDR_MAP + Link: 1 + Entries: + - Version: 4 + Feature: 0x48 + BBRanges: + - BaseAddress: 0x22222 + BBEntries: + - ID: 2 + AddressOffset: 0x0 + Size: 0x2 + Metadata: 0x4 + Hash: 0x2 + - BaseAddress: 0xFFFFF + BBEntries: + - ID: 15 + AddressOffset: 0xF0 + Size: 0xF1 + Metadata: 0x1F + Hash: 0xF + - Name: .llvm_bb_addr_map_3 + Type: SHT_LLVM_BB_ADDR_MAP + Link: 2 + Entries: + - Version: 4 + Feature: 0x40 + BBRanges: + - BaseAddress: 0x33333 + BBEntries: + - ID: 0 + AddressOffset: 0x0 + Size: 0x3 + Metadata: 0x6 + Hash: 0x3 + - Name: .llvm_bb_addr_map_4 + Type: SHT_LLVM_BB_ADDR_MAP + # Link: 0 (by default, can be overriden) + Entries: + - Version: 4 + Feature: 0x40 + BBRanges: + - BaseAddress: 0x44444 + BBEntries: + - ID: 0 + AddressOffset: 0x0 + Size: 0x4 + Metadata: 0x18 + Hash: 0x4 +)"); + + BBAddrMap E1 = { + {{0x11111, + {{1, 0x0, 0x3, {false, true, false, false, false}, {0x1, 0x2}, 0x1}}}}}; + BBAddrMap E2 = { + {{0x22222, {{2, 0x0, 0x2, {false, false, true, false, false}, {}, 0x2}}}, + {0xFFFFF, {{15, 0xF0, 0xF1, {true, true, true, true, true}, {}, 0xF}}}}}; + BBAddrMap E3 = { + {{0x33333, {{0, 0x0, 0x3, {false, true, true, false, false}, {}, 0x3}}}}}; + BBAddrMap E4 = { + {{0x44444, {{0, 0x0, 0x4, {false, false, false, true, true}, {}, 0x4}}}}}; + + std::vector<BBAddrMap> Section0BBAddrMaps = {E4}; + std::vector<BBAddrMap> Section1BBAddrMaps = {E3}; + std::vector<BBAddrMap> Section2BBAddrMaps = {E1, E2}; + std::vector<BBAddrMap> AllBBAddrMaps = {E1, E2, E3, E4}; + + auto DoCheckSucceeds = [&](StringRef YamlString, + std::optional<unsigned> TextSectionIndex, + std::vector<BBAddrMap> ExpectedResult) { + SCOPED_TRACE("for TextSectionIndex: " + + (TextSectionIndex ? llvm::Twine(*TextSectionIndex) : "{}") + + " and object yaml:\n" + YamlString); + SmallString<0> Storage; + Expected<ELFObjectFile<ELF64LE>> ElfOrErr = + toBinary<ELF64LE>(Storage, YamlString); + ASSERT_THAT_EXPECTED(ElfOrErr, Succeeded()); + + Expected<const typename ELF64LE::Shdr *> BBAddrMapSecOrErr = + ElfOrErr->getELFFile().getSection(1); + ASSERT_THAT_EXPECTED(BBAddrMapSecOrErr, Succeeded()); + auto BBAddrMaps = ElfOrErr->readBBAddrMap(TextSectionIndex); + ASSERT_THAT_EXPECTED(BBAddrMaps, Succeeded()); + EXPECT_EQ(*BBAddrMaps, ExpectedResult); + }; + + DoCheckSucceeds(CommonYamlString, /*TextSectionIndex=*/std::nullopt, + AllBBAddrMaps); + DoCheckSucceeds(CommonYamlString, /*TextSectionIndex=*/0, Section0BBAddrMaps); + DoCheckSucceeds(CommonYamlString, /*TextSectionIndex=*/2, Section1BBAddrMaps); + DoCheckSucceeds(CommonYamlString, /*TextSectionIndex=*/1, Section2BBAddrMaps); +} + // Test for the ELFObjectFile::readBBAddrMap API with PGOAnalysisMap. TEST(ELFObjectFileTest, ReadPGOAnalysisMap) { StringRef CommonYamlString(R"( @@ -1159,29 +1276,32 @@ Sections: )"); BBAddrMap E1 = { - {{0x11111, {{1, 0x0, 0x1, {false, true, false, false, false}, {}}}}}}; - PGOAnalysisMap P1 = {892, {}, {true, false, false, false, false, false}}; + {{0x11111, {{1, 0x0, 0x1, {false, true, false, false, false}, {}, 0}}}}}; + PGOAnalysisMap P1 = { + 892, {}, {true, false, false, false, false, false, false}}; BBAddrMap E2 = { - {{0x22222, {{2, 0x0, 0x2, {false, false, true, false, false}, {}}}}}}; + {{0x22222, {{2, 0x0, 0x2, {false, false, true, false, false}, {}, 0}}}}}; PGOAnalysisMap P2 = {{}, {{BlockFrequency(343), {}}}, - {false, true, false, false, false, false}}; - BBAddrMap E3 = {{{0x33333, - {{0, 0x0, 0x3, {false, true, true, false, false}, {}}, - {1, 0x3, 0x3, {false, false, true, false, false}, {}}, - {2, 0x6, 0x3, {false, false, false, false, false}, {}}}}}}; + {false, true, false, false, false, false, false}}; + BBAddrMap E3 = { + {{0x33333, + {{0, 0x0, 0x3, {false, true, true, false, false}, {}, 0}, + {1, 0x3, 0x3, {false, false, true, false, false}, {}, 0}, + {2, 0x6, 0x3, {false, false, false, false, false}, {}, 0}}}}}; PGOAnalysisMap P3 = {{}, {{{}, {{1, BranchProbability::getRaw(0x1111'1111)}, {2, BranchProbability::getRaw(0xeeee'eeee)}}}, {{}, {{2, BranchProbability::getRaw(0xffff'ffff)}}}, {{}, {}}}, - {false, false, true, false, false, false}}; - BBAddrMap E4 = {{{0x44444, - {{0, 0x0, 0x4, {false, false, false, true, true}, {}}, - {1, 0x4, 0x4, {false, false, false, false, false}, {}}, - {2, 0x8, 0x4, {false, false, false, false, false}, {}}, - {3, 0xc, 0x4, {false, false, false, false, false}, {}}}}}}; + {false, false, true, false, false, false, false}}; + BBAddrMap E4 = { + {{0x44444, + {{0, 0x0, 0x4, {false, false, false, true, true}, {}, 0}, + {1, 0x4, 0x4, {false, false, false, false, false}, {}, 0}, + {2, 0x8, 0x4, {false, false, false, false, false}, {}, 0}, + {3, 0xc, 0x4, {false, false, false, false, false}, {}, 0}}}}}; PGOAnalysisMap P4 = { 1000, {{BlockFrequency(1000), @@ -1193,22 +1313,24 @@ Sections: {3, BranchProbability::getRaw(0xeeee'eeee)}}}, {BlockFrequency(18), {{3, BranchProbability::getRaw(0xffff'ffff)}}}, {BlockFrequency(1000), {}}}, - {true, true, true, false, false, false}}; + {true, true, true, false, false, false, false}}; BBAddrMap E5 = { - {{0x55555, {{2, 0x0, 0x2, {false, false, true, false, false}, {}}}}}}; - PGOAnalysisMap P5 = {{}, {}, {false, false, false, false, false, false}}; + {{0x55555, {{2, 0x0, 0x2, {false, false, true, false, false}, {}, 0}}}}}; + PGOAnalysisMap P5 = { + {}, {}, {false, false, false, false, false, false, false}}; BBAddrMap E6 = { {{0x66666, - {{0, 0x0, 0x6, {false, true, true, false, false}, {}}, - {1, 0x6, 0x6, {false, false, true, false, false}, {}}}}, - {0x666661, {{2, 0x0, 0x6, {false, false, false, false, false}, {}}}}}}; + {{0, 0x0, 0x6, {false, true, true, false, false}, {}, 0}, + {1, 0x6, 0x6, {false, false, true, false, false}, {}, 0}}}, + {0x666661, + {{2, 0x0, 0x6, {false, false, false, false, false}, {}, 0}}}}}; PGOAnalysisMap P6 = {{}, {{{}, {{1, BranchProbability::getRaw(0x2222'2222)}, {2, BranchProbability::getRaw(0xcccc'cccc)}}}, {{}, {{2, BranchProbability::getRaw(0x8888'8888)}}}, {{}, {}}}, - {false, false, true, true, false, false}}; + {false, false, true, true, false, false, false}}; std::vector<BBAddrMap> Section0BBAddrMaps = {E4, E5, E6}; std::vector<BBAddrMap> Section1BBAddrMaps = {E3}; diff --git a/llvm/unittests/Object/ELFTypesTest.cpp b/llvm/unittests/Object/ELFTypesTest.cpp index f88931b5f..1765e15 100644 --- a/llvm/unittests/Object/ELFTypesTest.cpp +++ b/llvm/unittests/Object/ELFTypesTest.cpp @@ -101,21 +101,22 @@ static_assert( "PGOAnalysisMap should use the same type for basic block ID as BBAddrMap"); TEST(ELFTypesTest, BBAddrMapFeaturesEncodingTest) { - const std::array<BBAddrMap::Features, 11> Decoded = { - {{false, false, false, false, false, false}, - {true, false, false, false, false, false}, - {false, true, false, false, false, false}, - {false, false, true, false, false, false}, - {false, false, false, true, false, false}, - {true, true, false, false, false, false}, - {false, true, true, false, false, false}, - {false, true, true, true, false, false}, - {true, true, true, true, false, false}, - {false, false, false, false, true, false}, - {false, false, false, false, false, true}}}; - const std::array<uint8_t, 11> Encoded = {{0b0000, 0b0001, 0b0010, 0b0100, - 0b1000, 0b0011, 0b0110, 0b1110, - 0b1111, 0b1'0000, 0b10'0000}}; + const std::array<BBAddrMap::Features, 12> Decoded = { + {{false, false, false, false, false, false, false}, + {true, false, false, false, false, false, false}, + {false, true, false, false, false, false, false}, + {false, false, true, false, false, false, false}, + {false, false, false, true, false, false, false}, + {true, true, false, false, false, false, false}, + {false, true, true, false, false, false, false}, + {false, true, true, true, false, false, false}, + {true, true, true, true, false, false, false}, + {false, false, false, false, true, false, false}, + {false, false, false, false, false, true, false}, + {false, false, false, false, false, false, true}}}; + const std::array<uint8_t, 12> Encoded = { + {0b0000, 0b0001, 0b0010, 0b0100, 0b1000, 0b0011, 0b0110, 0b1110, 0b1111, + 0b1'0000, 0b10'0000, 0b100'0000}}; for (const auto &[Feat, EncodedVal] : llvm::zip(Decoded, Encoded)) EXPECT_EQ(Feat.encode(), EncodedVal); for (const auto &[Feat, EncodedVal] : llvm::zip(Decoded, Encoded)) { @@ -128,9 +129,9 @@ TEST(ELFTypesTest, BBAddrMapFeaturesEncodingTest) { TEST(ELFTypesTest, BBAddrMapFeaturesInvalidEncodingTest) { const std::array<std::string, 2> Errors = { - "invalid encoding for BBAddrMap::Features: 0x40", + "invalid encoding for BBAddrMap::Features: 0x80", "invalid encoding for BBAddrMap::Features: 0xf0"}; - const std::array<uint8_t, 2> Values = {{0b100'0000, 0b1111'0000}}; + const std::array<uint8_t, 2> Values = {{0b1000'0000, 0b1111'0000}}; for (const auto &[Val, Error] : llvm::zip(Values, Errors)) { EXPECT_THAT_ERROR(BBAddrMap::Features::decode(Val).takeError(), FailedWithMessage(Error)); diff --git a/llvm/unittests/Support/SourceMgrTest.cpp b/llvm/unittests/Support/SourceMgrTest.cpp index 301b64f..c65f001 100644 --- a/llvm/unittests/Support/SourceMgrTest.cpp +++ b/llvm/unittests/Support/SourceMgrTest.cpp @@ -8,6 +8,7 @@ #include "llvm/Support/SourceMgr.h" #include "llvm/Support/MemoryBuffer.h" +#include "llvm/Support/VirtualFileSystem.h" #include "llvm/Support/raw_ostream.h" #include "gtest/gtest.h" @@ -506,3 +507,13 @@ TEST_F(SourceMgrTest, PrintWithoutLoc) { Diag.print(nullptr, OS, false, false, false); EXPECT_EQ("message\n", Output); } + +TEST_F(SourceMgrTest, IncludeDirs) { + auto VFS = makeIntrusiveRefCnt<vfs::InMemoryFileSystem>(); + VFS->addFile("include/file", 0, MemoryBuffer::getMemBuffer("contents")); + SM.setVirtualFileSystem(std::move(VFS)); + SM.setIncludeDirs({"include"}); + std::string ResolvedPath; + unsigned NumBuffers = SM.AddIncludeFile("file", SMLoc(), ResolvedPath); + EXPECT_EQ(NumBuffers, 1u); +} diff --git a/llvm/utils/TableGen/Basic/DirectiveEmitter.cpp b/llvm/utils/TableGen/Basic/DirectiveEmitter.cpp index b4d816e..3c6ff11 100644 --- a/llvm/utils/TableGen/Basic/DirectiveEmitter.cpp +++ b/llvm/utils/TableGen/Basic/DirectiveEmitter.cpp @@ -266,10 +266,9 @@ static void emitDirectivesDecl(const RecordKeeper &Records, raw_ostream &OS) { return; StringRef Lang = DirLang.getName(); + IncludeGuardEmitter IncGuard(OS, (Twine("LLVM_") + Lang + "_INC").str()); - OS << "#ifndef LLVM_" << Lang << "_INC\n"; - OS << "#define LLVM_" << Lang << "_INC\n"; - OS << "\n#include \"llvm/ADT/ArrayRef.h\"\n"; + OS << "#include \"llvm/ADT/ArrayRef.h\"\n"; if (DirLang.hasEnableBitmaskEnumInNamespace()) OS << "#include \"llvm/ADT/BitmaskEnum.h\"\n"; @@ -370,7 +369,6 @@ static void emitDirectivesDecl(const RecordKeeper &Records, raw_ostream &OS) { OS << "};\n"; } LlvmNS.close(); - OS << "#endif // LLVM_" << Lang << "_INC\n"; } // Given a list of spellings (for a given clause/directive), order them diff --git a/llvm/utils/TableGen/Basic/RISCVTargetDefEmitter.cpp b/llvm/utils/TableGen/Basic/RISCVTargetDefEmitter.cpp index df14c77..f795937 100644 --- a/llvm/utils/TableGen/Basic/RISCVTargetDefEmitter.cpp +++ b/llvm/utils/TableGen/Basic/RISCVTargetDefEmitter.cpp @@ -68,13 +68,14 @@ static void emitRISCVExtensions(const RecordKeeper &Records, raw_ostream &OS) { if (!Extensions.empty()) { OS << "\nstatic constexpr ImpliedExtsEntry ImpliedExts[] = {\n"; for (const Record *Ext : Extensions) { - auto ImpliesList = Ext->getValueAsListOfDefs("Implies"); + std::vector<const Record *> ImpliesList = + Ext->getValueAsListOfDefs("Implies"); if (ImpliesList.empty()) continue; StringRef Name = getExtensionName(Ext); - for (auto *ImpliedExt : ImpliesList) { + for (const Record *ImpliedExt : ImpliesList) { if (!ImpliedExt->isSubClassOf("RISCVExtension")) continue; @@ -150,11 +151,12 @@ static void emitRISCVProfiles(const RecordKeeper &Records, raw_ostream &OS) { OS << "#ifdef GET_SUPPORTED_PROFILES\n"; OS << "#undef GET_SUPPORTED_PROFILES\n\n"; - auto Profiles = Records.getAllDerivedDefinitionsIfDefined("RISCVProfile"); + ArrayRef<const Record *> Profiles = + Records.getAllDerivedDefinitionsIfDefined("RISCVProfile"); if (!Profiles.empty()) { printProfileTable(OS, Profiles, /*Experimental=*/false); - bool HasExperimentalProfiles = any_of(Profiles, [&](auto &Rec) { + bool HasExperimentalProfiles = any_of(Profiles, [&](const Record *Rec) { return Rec->getValueAsBit("Experimental"); }); if (HasExperimentalProfiles) @@ -173,15 +175,17 @@ static void emitRISCVProcs(const RecordKeeper &RK, raw_ostream &OS) { // Iterate on all definition records. for (const Record *Rec : RK.getAllDerivedDefinitionsIfDefined("RISCVProcessorModel")) { - const std::vector<const Record *> &Features = + std::vector<const Record *> Features = Rec->getValueAsListOfDefs("Features"); - bool FastScalarUnalignedAccess = any_of(Features, [&](auto &Feature) { - return Feature->getValueAsString("Name") == "unaligned-scalar-mem"; - }); - - bool FastVectorUnalignedAccess = any_of(Features, [&](auto &Feature) { - return Feature->getValueAsString("Name") == "unaligned-vector-mem"; - }); + bool FastScalarUnalignedAccess = + any_of(Features, [&](const Record *Feature) { + return Feature->getValueAsString("Name") == "unaligned-scalar-mem"; + }); + + bool FastVectorUnalignedAccess = + any_of(Features, [&](const Record *Feature) { + return Feature->getValueAsString("Name") == "unaligned-vector-mem"; + }); OS << "PROC(" << Rec->getName() << ", {\"" << Rec->getValueAsString("Name") << "\"}, {\""; diff --git a/llvm/utils/TableGen/Common/Types.cpp b/llvm/utils/TableGen/Common/Types.cpp index 35b79b3..8e8d6f6 100644 --- a/llvm/utils/TableGen/Common/Types.cpp +++ b/llvm/utils/TableGen/Common/Types.cpp @@ -8,16 +8,12 @@ #include "Types.h" -// For LLVM_ATTRIBUTE_UNUSED -#include "llvm/Support/Compiler.h" - #include <cassert> using namespace llvm; -const char * -llvm::getMinimalTypeForRange(uint64_t Range, - unsigned MaxSize LLVM_ATTRIBUTE_UNUSED) { +const char *llvm::getMinimalTypeForRange(uint64_t Range, + [[maybe_unused]] unsigned MaxSize) { // TODO: The original callers only used 32 and 64 so these are the only // values permitted. Rather than widen the supported values we should // allow 64 for the callers that currently use 32 and remove the diff --git a/llvm/utils/TableGen/FastISelEmitter.cpp b/llvm/utils/TableGen/FastISelEmitter.cpp index dba8bde..e0be104 100644 --- a/llvm/utils/TableGen/FastISelEmitter.cpp +++ b/llvm/utils/TableGen/FastISelEmitter.cpp @@ -555,7 +555,7 @@ void FastISelMap::collectPatterns(const CodeGenDAGPatterns &CGP) { raw_string_ostream SuffixOS(ManglingSuffix); Operands.PrintManglingSuffix(SuffixOS, ImmediatePredicates, true); if (!StringSwitch<bool>(ManglingSuffix) - .Cases("", "r", "rr", "ri", "i", "f", true) + .Cases({"", "r", "rr", "ri", "i", "f"}, true) .Default(false)) continue; diff --git a/llvm/utils/gn/secondary/clang/unittests/Basic/BUILD.gn b/llvm/utils/gn/secondary/clang/unittests/Basic/BUILD.gn index 1449dc7..954de88 100644 --- a/llvm/utils/gn/secondary/clang/unittests/Basic/BUILD.gn +++ b/llvm/utils/gn/secondary/clang/unittests/Basic/BUILD.gn @@ -14,6 +14,7 @@ unittest("BasicTests") { "DiagnosticTest.cpp", "FileEntryTest.cpp", "FileManagerTest.cpp", + "LangOptionsTest.cpp", "LineOffsetMappingTest.cpp", "OffloadArchTest.cpp", "SanitizersTest.cpp", diff --git a/llvm/utils/profcheck-xfail.txt b/llvm/utils/profcheck-xfail.txt index 343c2bb71..3f8be5e 100644 --- a/llvm/utils/profcheck-xfail.txt +++ b/llvm/utils/profcheck-xfail.txt @@ -906,7 +906,6 @@ Transforms/InstCombine/select_frexp.ll Transforms/InstCombine/select.ll Transforms/InstCombine/select-min-max.ll Transforms/InstCombine/select-of-symmetric-selects.ll -Transforms/InstCombine/select-safe-impliedcond-transforms.ll Transforms/InstCombine/select-safe-transforms.ll Transforms/InstCombine/select-select.ll Transforms/InstCombine/select-with-extreme-eq-cond.ll @@ -1237,7 +1236,6 @@ Transforms/PartiallyInlineLibCalls/X86/good-prototype.ll Transforms/PGOProfile/chr-dead-pred.ll Transforms/PGOProfile/chr-dup-threshold.ll Transforms/PGOProfile/chr-lifetimes.ll -Transforms/PGOProfile/chr.ll Transforms/PGOProfile/chr-poison.ll Transforms/PGOProfile/comdat.ll Transforms/PGOProfile/memop_profile_funclet_wasm.ll @@ -1312,17 +1310,6 @@ Transforms/SimpleLoopUnswitch/pr60736.ll Transforms/SimpleLoopUnswitch/trivial-unswitch-freeze-individual-conditions.ll Transforms/SimpleLoopUnswitch/trivial-unswitch.ll Transforms/SimpleLoopUnswitch/trivial-unswitch-logical-and-or.ll -Transforms/SROA/addrspacecast.ll -Transforms/SROA/phi-and-select.ll -Transforms/SROA/phi-gep.ll -Transforms/SROA/scalable-vectors-with-known-vscale.ll -Transforms/SROA/select-gep.ll -Transforms/SROA/select-load.ll -Transforms/SROA/slice-width.ll -Transforms/SROA/std-clamp.ll -Transforms/SROA/vector-conversion.ll -Transforms/SROA/vector-promotion-cannot-tree-structure-merge.ll -Transforms/SROA/vector-promotion.ll Transforms/StackProtector/cross-dso-cfi-stack-chk-fail.ll Transforms/StructurizeCFG/AMDGPU/uniform-regions.ll Transforms/StructurizeCFG/hoist-zerocost.ll |