diff options
Diffstat (limited to 'clang')
165 files changed, 4735 insertions, 1487 deletions
diff --git a/clang/cmake/caches/Fuchsia-stage2-instrumented.cmake b/clang/cmake/caches/Fuchsia-stage2-instrumented.cmake index ecd478a..b328f3d 100644 --- a/clang/cmake/caches/Fuchsia-stage2-instrumented.cmake +++ b/clang/cmake/caches/Fuchsia-stage2-instrumented.cmake @@ -43,3 +43,6 @@ set(CLANG_BOOTSTRAP_CMAKE_ARGS ${EXTRA_ARGS} -C ${CMAKE_CURRENT_LIST_DIR}/Fuchsia-stage2.cmake CACHE STRING "") + +# Do not use LLVM build for generating PGO data. +set(CLANG_PGO_TRAINING_USE_LLVM_BUILD OFF CACHE BOOL "") diff --git a/clang/docs/ClangFormatStyleOptions.rst b/clang/docs/ClangFormatStyleOptions.rst index bfc8094..d39ee49 100644 --- a/clang/docs/ClangFormatStyleOptions.rst +++ b/clang/docs/ClangFormatStyleOptions.rst @@ -4201,8 +4201,8 @@ the configuration (without a prefix: ``Auto``). * ``""`` means "arbitrary suffix" * ``"$"`` means "no suffix" - For example, if configured to ``"(_test)?$"``, then a header a.h would be seen - as the "main" include in both a.cc and a_test.cc. + For example, if configured to ``"(_test)?$"``, then a header a.h would be + seen as the "main" include in both a.cc and a_test.cc. .. _IncludeIsMainSourceRegex: diff --git a/clang/docs/HIPSupport.rst b/clang/docs/HIPSupport.rst index 406e1c8..b4a671e 100644 --- a/clang/docs/HIPSupport.rst +++ b/clang/docs/HIPSupport.rst @@ -545,37 +545,22 @@ The following restrictions imposed on user code apply to both modes: 1. Pointers to function, and all associated features, such as e.g. dynamic polymorphism, cannot be used (directly or transitively) by the user provided callable passed to an algorithm invocation; -2. Global / namespace scope / ``static`` / ``thread`` storage duration variables - cannot be used (directly or transitively) in name by the user provided - callable; - - - When executing in **HMM Mode** they can be used in address e.g.: - - .. code-block:: C++ - - namespace { int foo = 42; } - - bool never(const std::vector<int>& v) { - return std::any_of(std::execution::par_unseq, std::cbegin(v), std::cend(v), [](auto&& x) { - return x == foo; - }); - } - - bool only_in_hmm_mode(const std::vector<int>& v) { - return std::any_of(std::execution::par_unseq, std::cbegin(v), std::cend(v), - [p = &foo](auto&& x) { return x == *p; }); - } - -3. Only algorithms that are invoked with the ``parallel_unsequenced_policy`` are +2. ``static`` (except for program-wide unique ones) / ``thread`` storage + duration variables cannot be used (directly or transitively) in name by the + user provided callable; +3. User code must be compiled in ``-fgpu-rdc`` mode in order for global / + namespace scope variables / program-wide unique ``static`` storage duration + variables to be usable in name by the user provided callable; +4. Only algorithms that are invoked with the ``parallel_unsequenced_policy`` are candidates for offload; -4. Only algorithms that are invoked with iterator arguments that model +5. Only algorithms that are invoked with iterator arguments that model `random_access_iterator <https://en.cppreference.com/w/cpp/iterator/random_access_iterator>`_ are candidates for offload; -5. `Exceptions <https://en.cppreference.com/w/cpp/language/exceptions>`_ cannot +6. `Exceptions <https://en.cppreference.com/w/cpp/language/exceptions>`_ cannot be used by the user provided callable; -6. Dynamic memory allocation (e.g. ``operator new``) cannot be used by the user +7. Dynamic memory allocation (e.g. ``operator new``) cannot be used by the user provided callable; -7. Selective offload is not possible i.e. it is not possible to indicate that +8. Selective offload is not possible i.e. it is not possible to indicate that only some algorithms invoked with the ``parallel_unsequenced_policy`` are to be executed on the accelerator. @@ -585,15 +570,6 @@ additional restrictions: 1. All code that is expected to interoperate has to be recompiled with the ``--hipstdpar-interpose-alloc`` flag i.e. it is not safe to compose libraries that have been independently compiled; -2. automatic storage duration (i.e. stack allocated) variables cannot be used - (directly or transitively) by the user provided callable e.g. - - .. code-block:: c++ - - bool never(const std::vector<int>& v, int n) { - return std::any_of(std::execution::par_unseq, std::cbegin(v), std::cend(v), - [p = &n](auto&& x) { return x == *p; }); - } Current Support =============== @@ -626,17 +602,12 @@ Linux operating system. Support is synthesised in the following table: The minimum Linux kernel version for running in HMM mode is 6.4. -The forwarding header can be obtained from -`its GitHub repository <https://github.com/ROCm/roc-stdpar>`_. -It will be packaged with a future `ROCm <https://rocm.docs.amd.com/en/latest/>`_ -release. Because accelerated algorithms are provided via -`rocThrust <https://rocm.docs.amd.com/projects/rocThrust/en/latest/>`_, a -transitive dependency on -`rocPrim <https://rocm.docs.amd.com/projects/rocPRIM/en/latest/>`_ exists. Both -can be obtained either by installing their associated components of the -`ROCm <https://rocm.docs.amd.com/en/latest/>`_ stack, or from their respective -repositories. The list algorithms that can be offloaded is available -`here <https://github.com/ROCm/roc-stdpar#algorithm-support-status>`_. +The forwarding header is packaged by +`ROCm <https://rocm.docs.amd.com/en/latest/>`_, and is obtainable by installing +the `hipstdpar` packege. The list algorithms that can be offloaded is available +`here <https://github.com/ROCm/roc-stdpar#algorithm-support-status>`_. More +details are available via the dedicated blog +`<https://rocm.blogs.amd.com/software-tools-optimization/hipstdpar/README.html>`_. HIP Specific Elements --------------------- @@ -690,9 +661,8 @@ HIP Specific Elements Open Questions / Future Developments ==================================== -1. The restriction on the use of global / namespace scope / ``static`` / - ``thread`` storage duration variables in offloaded algorithms will be lifted - in the future, when running in **HMM Mode**; +1. The restriction on the use of ``static`` / ``thread`` storage duration + variables in offloaded algorithms might be lifted; 2. The restriction on the use of dynamic memory allocation in offloaded algorithms will be lifted in the future. 3. The restriction on the use of pointers to function, and associated features diff --git a/clang/docs/LanguageExtensions.rst b/clang/docs/LanguageExtensions.rst index b334e25..34e1bf1 100644 --- a/clang/docs/LanguageExtensions.rst +++ b/clang/docs/LanguageExtensions.rst @@ -848,6 +848,14 @@ of different sizes and signs is forbidden in binary and ternary builtins. semantics, see `LangRef <http://llvm.org/docs/LangRef.html#llvm-min-intrinsics-comparation>`_ for the comparison. + T __builtin_elementwise_maximumnum(T x, T y) return x or y, whichever is larger. Follows IEEE 754-2019 floating point types + semantics, see `LangRef + <http://llvm.org/docs/LangRef.html#llvm-min-intrinsics-comparation>`_ + for the comparison. + T __builtin_elementwise_minimumnum(T x, T y) return x or y, whichever is smaller. Follows IEEE 754-2019 floating point types + semantics, see `LangRef + <http://llvm.org/docs/LangRef.html#llvm-min-intrinsics-comparation>`_ + for the comparison. ============================================== ====================================================================== ========================================= diff --git a/clang/docs/OpenMPSupport.rst b/clang/docs/OpenMPSupport.rst index 986aaab..fc44f4c 100644 --- a/clang/docs/OpenMPSupport.rst +++ b/clang/docs/OpenMPSupport.rst @@ -443,7 +443,7 @@ implementation. +-------------------------------------------------------------+---------------------------+---------------------------+--------------------------------------------------------------------------+ | Traits for default device envirable | :none:`unclaimed` | :none:`unclaimed` | | +-------------------------------------------------------------+---------------------------+---------------------------+--------------------------------------------------------------------------+ -| Optionally omit array length expression | :none:`unclaimed` | :none:`unclaimed` | | +| Optionally omit array length expression | :good:`done` | :none:`unclaimed` | https://github.com/llvm/llvm-project/pull/148048 | +-------------------------------------------------------------+---------------------------+---------------------------+--------------------------------------------------------------------------+ | Canonical loop sequences | :none:`unclaimed` | :none:`unclaimed` | | +-------------------------------------------------------------+---------------------------+---------------------------+--------------------------------------------------------------------------+ @@ -467,6 +467,9 @@ implementation. +-------------------------------------------------------------+---------------------------+---------------------------+--------------------------------------------------------------------------+ | Non-const do_not_sync for nowait/nogroup | :none:`unclaimed` | :none:`unclaimed` | | +-------------------------------------------------------------+---------------------------+---------------------------+--------------------------------------------------------------------------+ +| need_device_addr modifier for adjust_args clause | :part:`partial` | :none:`unclaimed` | Parsing/Sema: https://github.com/llvm/llvm-project/pull/143442 | +| | | | https://github.com/llvm/llvm-project/pull/149586 | ++-------------------------------------------------------------+---------------------------+---------------------------+--------------------------------------------------------------------------+ OpenMP Extensions ================= diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst index 81483c1..dd6b44d 100644 --- a/clang/docs/ReleaseNotes.rst +++ b/clang/docs/ReleaseNotes.rst @@ -344,6 +344,7 @@ Non-comprehensive list of changes in this release - Added `__builtin_elementwise_exp10`. - For AMDPGU targets, added `__builtin_v_cvt_off_f32_i4` that maps to the `v_cvt_off_f32_i4` instruction. - Added `__builtin_elementwise_minnum` and `__builtin_elementwise_maxnum`. +- Added `__builtin_elementwise_minnumnum` and `__builtin_elementwise_maxnumnum`. - No longer crashing on invalid Objective-C categories and extensions when dumping the AST as JSON. (#GH137320) - Clang itself now uses split stacks instead of threads for allocating more @@ -814,6 +815,10 @@ Bug Fixes in This Version - Fixed a failed assertion with an operator call expression which comes from a macro expansion when performing analysis for nullability attributes. (#GH138371) - Fixed a concept equivalent checking crash due to untransformed constraint expressions. (#GH146614) +- Fixed a crash in `clang-scan-deps` when a module with the same name is found + in different locations (#GH134404, #GH146976). +- Fix a crash when marco name is empty in ``#pragma push_macro("")`` or + ``#pragma pop_macro("")``. (GH149762). Bug Fixes to Compiler Builtins ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ @@ -1309,6 +1314,9 @@ OpenMP Support mappers, by using compiler-generated default mappers for the outer structs for such maps. - Deprecation warning has been emitted for deprecated delimited form of ``declare target``. +- Added parsing and semantic analysis support for the 'need_device_addr' + modifier in the 'adjust_args' clause. +- Allow array length to be omitted in array section subscript expression. Improvements ^^^^^^^^^^^^ diff --git a/clang/include/clang-c/Index.h b/clang/include/clang-c/Index.h index b929585..be038d9 100644 --- a/clang/include/clang-c/Index.h +++ b/clang/include/clang-c/Index.h @@ -6953,7 +6953,7 @@ clang_getCursorUnaryOperatorKind(CXCursor cursor); * @} */ -CINDEX_DEPRECATED +/* CINDEX_DEPRECATED - disabled to silence MSVC deprecation warnings */ typedef void *CXRemapping; CINDEX_DEPRECATED CINDEX_LINKAGE CXRemapping clang_getRemappings(const char *); diff --git a/clang/include/clang/Analysis/Analyses/LifetimeSafety.h b/clang/include/clang/Analysis/Analyses/LifetimeSafety.h index 9998702..1c00558 100644 --- a/clang/include/clang/Analysis/Analyses/LifetimeSafety.h +++ b/clang/include/clang/Analysis/Analyses/LifetimeSafety.h @@ -17,14 +17,105 @@ //===----------------------------------------------------------------------===// #ifndef LLVM_CLANG_ANALYSIS_ANALYSES_LIFETIMESAFETY_H #define LLVM_CLANG_ANALYSIS_ANALYSES_LIFETIMESAFETY_H -#include "clang/AST/DeclBase.h" #include "clang/Analysis/AnalysisDeclContext.h" #include "clang/Analysis/CFG.h" -namespace clang { +#include "llvm/ADT/ImmutableSet.h" +#include "llvm/ADT/StringMap.h" +#include <memory> -void runLifetimeSafetyAnalysis(const DeclContext &DC, const CFG &Cfg, - AnalysisDeclContext &AC); +namespace clang::lifetimes { -} // namespace clang +/// The main entry point for the analysis. +void runLifetimeSafetyAnalysis(AnalysisDeclContext &AC); + +namespace internal { +// Forward declarations of internal types. +class Fact; +class FactManager; +class LoanPropagationAnalysis; +class ExpiredLoansAnalysis; +struct LifetimeFactory; + +/// A generic, type-safe wrapper for an ID, distinguished by its `Tag` type. +/// Used for giving ID to loans and origins. +template <typename Tag> struct ID { + uint32_t Value = 0; + + bool operator==(const ID<Tag> &Other) const { return Value == Other.Value; } + bool operator!=(const ID<Tag> &Other) const { return !(*this == Other); } + bool operator<(const ID<Tag> &Other) const { return Value < Other.Value; } + ID<Tag> operator++(int) { + ID<Tag> Tmp = *this; + ++Value; + return Tmp; + } + void Profile(llvm::FoldingSetNodeID &IDBuilder) const { + IDBuilder.AddInteger(Value); + } +}; +template <typename Tag> +inline llvm::raw_ostream &operator<<(llvm::raw_ostream &OS, ID<Tag> ID) { + return OS << ID.Value; +} + +using LoanID = ID<struct LoanTag>; +using OriginID = ID<struct OriginTag>; + +// Using LLVM's immutable collections is efficient for dataflow analysis +// as it avoids deep copies during state transitions. +// TODO(opt): Consider using a bitset to represent the set of loans. +using LoanSet = llvm::ImmutableSet<LoanID>; +using OriginSet = llvm::ImmutableSet<OriginID>; + +/// A `ProgramPoint` identifies a location in the CFG by pointing to a specific +/// `Fact`. identified by a lifetime-related event (`Fact`). +/// +/// A `ProgramPoint` has "after" semantics: it represents the location +/// immediately after its corresponding `Fact`. +using ProgramPoint = const Fact *; + +/// Running the lifetime safety analysis and querying its results. It +/// encapsulates the various dataflow analyses. +class LifetimeSafetyAnalysis { +public: + LifetimeSafetyAnalysis(AnalysisDeclContext &AC); + ~LifetimeSafetyAnalysis(); + + void run(); + + /// Returns the set of loans an origin holds at a specific program point. + LoanSet getLoansAtPoint(OriginID OID, ProgramPoint PP) const; + + /// Returns the set of loans that have expired at a specific program point. + LoanSet getExpiredLoansAtPoint(ProgramPoint PP) const; + + /// Finds the OriginID for a given declaration. + /// Returns a null optional if not found. + std::optional<OriginID> getOriginIDForDecl(const ValueDecl *D) const; + + /// Finds the LoanID's for the loan created with the specific variable as + /// their Path. + std::vector<LoanID> getLoanIDForVar(const VarDecl *VD) const; + + /// Retrieves program points that were specially marked in the source code + /// for testing. + /// + /// The analysis recognizes special function calls of the form + /// `void("__lifetime_test_point_<name>")` as test points. This method returns + /// a map from the annotation string (<name>) to the corresponding + /// `ProgramPoint`. This allows test harnesses to query the analysis state at + /// user-defined locations in the code. + /// \note This is intended for testing only. + llvm::StringMap<ProgramPoint> getTestPoints() const; + +private: + AnalysisDeclContext &AC; + std::unique_ptr<LifetimeFactory> Factory; + std::unique_ptr<FactManager> FactMgr; + std::unique_ptr<LoanPropagationAnalysis> LoanPropagation; + std::unique_ptr<ExpiredLoansAnalysis> ExpiredLoans; +}; +} // namespace internal +} // namespace clang::lifetimes #endif // LLVM_CLANG_ANALYSIS_ANALYSES_LIFETIMESAFETY_H diff --git a/clang/include/clang/Basic/Builtins.td b/clang/include/clang/Basic/Builtins.td index 5ebb821..c81714e 100644 --- a/clang/include/clang/Basic/Builtins.td +++ b/clang/include/clang/Basic/Builtins.td @@ -1334,6 +1334,18 @@ def ElementwiseMinimum : Builtin { let Prototype = "void(...)"; } +def ElementwiseMaximumNum : Builtin { + let Spellings = ["__builtin_elementwise_maximumnum"]; + let Attributes = [NoThrow, Const, CustomTypeChecking]; + let Prototype = "void(...)"; +} + +def ElementwiseMinimumNum : Builtin { + let Spellings = ["__builtin_elementwise_minimumnum"]; + let Attributes = [NoThrow, Const, CustomTypeChecking]; + let Prototype = "void(...)"; +} + def ElementwiseCeil : Builtin { let Spellings = ["__builtin_elementwise_ceil"]; let Attributes = [NoThrow, Const, CustomTypeChecking]; diff --git a/clang/include/clang/Basic/CodeGenOptions.def b/clang/include/clang/Basic/CodeGenOptions.def index cfffeb7..e137738 100644 --- a/clang/include/clang/Basic/CodeGenOptions.def +++ b/clang/include/clang/Basic/CodeGenOptions.def @@ -313,9 +313,10 @@ CODEGENOPT(SpeculativeLoadHardening, 1, 0, Benign) ///< Enable speculative load CODEGENOPT(FineGrainedBitfieldAccesses, 1, 0, Benign) ///< Enable fine-grained bitfield accesses. CODEGENOPT(StrictEnums , 1, 0, Benign) ///< Optimize based on strict enum definition. CODEGENOPT(StrictVTablePointers, 1, 0, Benign) ///< Optimize based on the strict vtable pointers -CODEGENOPT(TimePasses , 1, 0, Benign) ///< Set when -ftime-report or -ftime-report= or -ftime-report-json is enabled. +CODEGENOPT(TimePasses , 1, 0, Benign) ///< Set when -ftime-report, -ftime-report=, -ftime-report-json, or -stats-file-timers is enabled. CODEGENOPT(TimePassesPerRun , 1, 0, Benign) ///< Set when -ftime-report=per-pass-run is enabled. CODEGENOPT(TimePassesJson , 1, 0, Benign) ///< Set when -ftime-report-json is enabled. +CODEGENOPT(TimePassesStatsFile , 1, 0, Benign) ///< Set when -stats-file-timers is enabled. CODEGENOPT(TimeTrace , 1, 0, Benign) ///< Set when -ftime-trace is enabled. VALUE_CODEGENOPT(TimeTraceGranularity, 32, 500, Benign) ///< Minimum time granularity (in microseconds), ///< traced by time profiler diff --git a/clang/include/clang/Basic/Diagnostic.h b/clang/include/clang/Basic/Diagnostic.h index c7a6276..cee5bed 100644 --- a/clang/include/clang/Basic/Diagnostic.h +++ b/clang/include/clang/Basic/Diagnostic.h @@ -895,7 +895,10 @@ public: /// \param FormatString A fixed diagnostic format string that will be hashed /// and mapped to a unique DiagID. template <unsigned N> - // TODO: Deprecate this once all uses are removed from Clang. + // FIXME: this API should almost never be used; custom diagnostics do not + // have an associated diagnostic group and thus cannot be controlled by users + // like other diagnostics. The number of times this API is used in Clang + // should only ever be reduced, not increased. // [[deprecated("Use a CustomDiagDesc instead of a Level")]] unsigned getCustomDiagID(Level L, const char (&FormatString)[N]) { return Diags->getCustomDiagID((DiagnosticIDs::Level)L, diff --git a/clang/include/clang/Basic/DiagnosticDriverKinds.td b/clang/include/clang/Basic/DiagnosticDriverKinds.td index 34b6c0d..759ba04 100644 --- a/clang/include/clang/Basic/DiagnosticDriverKinds.td +++ b/clang/include/clang/Basic/DiagnosticDriverKinds.td @@ -116,6 +116,8 @@ def err_drv_cuda_host_arch : Error< "unsupported architecture '%0' for host compilation">; def err_drv_mix_cuda_hip : Error< "mixed CUDA and HIP compilation is not supported">; +def err_drv_mix_offload : Error< + "mixed %0 and %1 offloading compilation is not supported">; def err_drv_bad_target_id : Error< "invalid target ID '%0'; format is a processor name followed by an optional " "colon-delimited list of features followed by an enable/disable sign (e.g., " diff --git a/clang/include/clang/Basic/DiagnosticIDs.h b/clang/include/clang/Basic/DiagnosticIDs.h index 2b095f0..f07a003 100644 --- a/clang/include/clang/Basic/DiagnosticIDs.h +++ b/clang/include/clang/Basic/DiagnosticIDs.h @@ -283,7 +283,10 @@ public: // writing, nearly all callers of this function were invalid. unsigned getCustomDiagID(CustomDiagDesc Diag); - // TODO: Deprecate this once all uses are removed from LLVM + // FIXME: this API should almost never be used; custom diagnostics do not + // have an associated diagnostic group and thus cannot be controlled by users + // like other diagnostics. The number of times this API is used in Clang + // should only ever be reduced, not increased. // [[deprecated("Use a CustomDiagDesc instead of a Level")]] unsigned getCustomDiagID(Level Level, StringRef Message) { return getCustomDiagID([&]() -> CustomDiagDesc { diff --git a/clang/include/clang/Basic/DiagnosticLexKinds.td b/clang/include/clang/Basic/DiagnosticLexKinds.td index 723f5d4..c7fe6e1d 100644 --- a/clang/include/clang/Basic/DiagnosticLexKinds.td +++ b/clang/include/clang/Basic/DiagnosticLexKinds.td @@ -694,6 +694,9 @@ def err_pragma_push_pop_macro_malformed : Error< def warn_pragma_pop_macro_no_push : Warning< "pragma pop_macro could not pop '%0', no matching push_macro">, InGroup<IgnoredPragmas>; +def warn_pargma_push_pop_macro_empty_string : Warning< + "'#pragma %select{push_macro|pop_macro}0' expected a non-empty string">, + InGroup<IgnoredPragmas>; def warn_pragma_message : Warning<"%0">, InGroup<PoundPragmaMessage>, DefaultWarnNoWerror; def err_pragma_message : Error<"%0">; diff --git a/clang/include/clang/Basic/DiagnosticParseKinds.td b/clang/include/clang/Basic/DiagnosticParseKinds.td index 35903af..165f015 100644 --- a/clang/include/clang/Basic/DiagnosticParseKinds.td +++ b/clang/include/clang/Basic/DiagnosticParseKinds.td @@ -1594,6 +1594,9 @@ def err_omp_unknown_adjust_args_op def err_omp_declare_variant_wrong_clause : Error< "expected %select{'match'|'match', 'adjust_args', or 'append_args'}0 clause " "on 'omp declare variant' directive">; +def err_omp_non_by_ref_need_device_addr_modifier_argument + : Error<"expected reference type argument on 'adjust_args' clause with " + "'need_device_addr' modifier">; def err_omp_declare_variant_duplicate_nested_trait : Error< "nested OpenMP context selector contains duplicated trait '%0'" " in selector '%1' and set '%2' with different score">; diff --git a/clang/include/clang/Basic/LangOptions.def b/clang/include/clang/Basic/LangOptions.def index 6ac8d49..08d98a7 100644 --- a/clang/include/clang/Basic/LangOptions.def +++ b/clang/include/clang/Basic/LangOptions.def @@ -491,6 +491,8 @@ LANGOPT(CheckConstexprFunctionBodies, 1, 1, Benign, LANGOPT(BoundsSafety, 1, 0, NotCompatible, "Bounds safety extension for C") +LANGOPT(EnableLifetimeSafety, 1, 0, NotCompatible, "Experimental lifetime safety analysis for C++") + LANGOPT(PreserveVec3Type, 1, 0, NotCompatible, "Preserve 3-component vector type") #undef LANGOPT diff --git a/clang/include/clang/Basic/TargetInfo.h b/clang/include/clang/Basic/TargetInfo.h index 8b66c73..abfbdfa 100644 --- a/clang/include/clang/Basic/TargetInfo.h +++ b/clang/include/clang/Basic/TargetInfo.h @@ -1551,8 +1551,8 @@ public: // Return the target-specific priority for features/cpus/vendors so // that they can be properly sorted for checking. - virtual uint64_t getFMVPriority(ArrayRef<StringRef> Features) const { - return 0; + virtual llvm::APInt getFMVPriority(ArrayRef<StringRef> Features) const { + return llvm::APInt::getZero(32); } // Validate the contents of the __builtin_cpu_is(const char*) diff --git a/clang/include/clang/Basic/arm_neon.td b/clang/include/clang/Basic/arm_neon.td index 0daef4a..ef19610 100644 --- a/clang/include/clang/Basic/arm_neon.td +++ b/clang/include/clang/Basic/arm_neon.td @@ -964,11 +964,11 @@ def SLI_N : WInst<"vsli_n", "...I", "PlQPl", [ImmCheck<2, ImmCheckShiftLeft, 0>] // Right shift narrow high def SHRN_HIGH_N : IOpInst<"vshrn_high_n", "<(<q).I", "HsHiHlHUsHUiHUl", OP_NARROW_HI>; -def QSHRUN_HIGH_N : SOpInst<"vqshrun_high_n", "<(<q).I", +def QSHRUN_HIGH_N : SOpInst<"vqshrun_high_n", "(<U)(<Uq).I", "HsHiHl", OP_NARROW_HI>; def RSHRN_HIGH_N : IOpInst<"vrshrn_high_n", "<(<q).I", "HsHiHlHUsHUiHUl", OP_NARROW_HI>; -def QRSHRUN_HIGH_N : SOpInst<"vqrshrun_high_n", "<(<q).I", +def QRSHRUN_HIGH_N : SOpInst<"vqrshrun_high_n", "(<U)(<Uq).I", "HsHiHl", OP_NARROW_HI>; def QSHRN_HIGH_N : SOpInst<"vqshrn_high_n", "<(<q).I", "HsHiHlHUsHUiHUl", OP_NARROW_HI>; diff --git a/clang/include/clang/CIR/Dialect/IR/CIROps.td b/clang/include/clang/CIR/Dialect/IR/CIROps.td index 694e369..c54075e 100644 --- a/clang/include/clang/CIR/Dialect/IR/CIROps.td +++ b/clang/include/clang/CIR/Dialect/IR/CIROps.td @@ -128,12 +128,12 @@ def CIR_CastKind : CIR_I32EnumAttr<"CastKind", "cast kind", [ // CK_BlockPointerToObjCPointerCast // CK_AnyPointerToBlockPointerCast // CK_ObjCObjectLValueCast - // I32EnumAttrCase<"float_to_complex", 44>, - // I32EnumAttrCase<"float_complex_to_real", 45>, - // I32EnumAttrCase<"float_complex_to_bool", 46>, + I32EnumAttrCase<"float_to_complex", 44>, + I32EnumAttrCase<"float_complex_to_real", 45>, + I32EnumAttrCase<"float_complex_to_bool", 46>, I32EnumAttrCase<"float_complex", 47>, - // I32EnumAttrCase<"float_complex_to_int_complex", 48>, - // I32EnumAttrCase<"int_to_complex", 49>, + I32EnumAttrCase<"float_complex_to_int_complex", 48>, + I32EnumAttrCase<"int_to_complex", 49>, I32EnumAttrCase<"int_complex_to_real", 50>, I32EnumAttrCase<"int_complex_to_bool", 51>, I32EnumAttrCase<"int_complex", 52>, diff --git a/clang/include/clang/Driver/Driver.h b/clang/include/clang/Driver/Driver.h index d9e328f..78a4c57 100644 --- a/clang/include/clang/Driver/Driver.h +++ b/clang/include/clang/Driver/Driver.h @@ -355,6 +355,9 @@ public: phases::ID getFinalPhase(const llvm::opt::DerivedArgList &DAL, llvm::opt::Arg **FinalPhaseArg = nullptr) const; + llvm::Expected<std::unique_ptr<llvm::MemoryBuffer>> + executeProgram(llvm::ArrayRef<llvm::StringRef> Args) const; + private: /// Certain options suppress the 'no input files' warning. LLVM_PREFERRED_TYPE(bool) @@ -367,10 +370,6 @@ private: /// stored in it, and will clean them up when torn down. mutable llvm::StringMap<std::unique_ptr<ToolChain>> ToolChains; - /// The associated offloading architectures with each toolchain. - llvm::DenseMap<const ToolChain *, llvm::SmallVector<llvm::StringRef>> - OffloadArchs; - private: /// TranslateInputArgs - Create a new derived argument list from the input /// arguments, after applying the standard argument translations. @@ -537,8 +536,7 @@ public: /// empty string. llvm::SmallVector<StringRef> getOffloadArchs(Compilation &C, const llvm::opt::DerivedArgList &Args, - Action::OffloadKind Kind, const ToolChain *TC, - bool SpecificToolchain = true) const; + Action::OffloadKind Kind, const ToolChain &TC) const; /// Check that the file referenced by Value exists. If it doesn't, /// issue a diagnostic and return false. diff --git a/clang/include/clang/Driver/Options.td b/clang/include/clang/Driver/Options.td index e30c152..916400e 100644 --- a/clang/include/clang/Driver/Options.td +++ b/clang/include/clang/Driver/Options.td @@ -1156,7 +1156,7 @@ def offload_arch_EQ : CommaJoined<["--"], "offload-arch=">, "If 'native' is used the compiler will detect locally installed architectures. " "For HIP offloading, the device architecture can be followed by target ID features " "delimited by a colon (e.g. gfx908:xnack+:sramecc-). May be specified more than once.">; -def no_offload_arch_EQ : Joined<["--"], "no-offload-arch=">, +def no_offload_arch_EQ : CommaJoined<["--"], "no-offload-arch=">, Visibility<[ClangOption, FlangOption]>, HelpText<"Remove CUDA/HIP offloading device architecture (e.g. sm_35, gfx906) from the list of devices to compile for. " "'all' resets the list to its default value.">; @@ -1624,7 +1624,7 @@ defm auto_import : BoolFOption<"auto-import", // In the future this option will be supported by other offloading // languages and accept other values such as CPU/GPU architectures, // offload kinds and target aliases. -def offload_EQ : CommaJoined<["--"], "offload=">, Flags<[NoXarchOption]>, +def offload_EQ : CommaJoined<["--"], "offload=">, Flags<[NoXarchOption]>, Alias<offload_targets_EQ>, HelpText<"Specify comma-separated list of offloading target triples (CUDA and HIP only)">; // C++ Coroutines @@ -1904,6 +1904,14 @@ defm bounds_safety : BoolFOption< BothFlags<[], [CC1Option], " experimental bounds safety extension for C">>; +defm lifetime_safety : BoolFOption< + "experimental-lifetime-safety", + LangOpts<"EnableLifetimeSafety">, DefaultFalse, + PosFlag<SetTrue, [], [CC1Option], "Enable">, + NegFlag<SetFalse, [], [CC1Option], "Disable">, + BothFlags<[], [CC1Option], + " experimental lifetime safety for C++">>; + defm addrsig : BoolFOption<"addrsig", CodeGenOpts<"Addrsig">, DefaultFalse, PosFlag<SetTrue, [], [ClangOption, CC1Option], "Emit">, @@ -8253,6 +8261,9 @@ def stats_file : Joined<["-"], "stats-file=">, def stats_file_append : Flag<["-"], "stats-file-append">, HelpText<"If stats should be appended to stats-file instead of overwriting it">, MarshallingInfoFlag<FrontendOpts<"AppendStats">>; +def stats_file_timers : Flag<["-"], "stats-file-timers">, + HelpText<"If stats should include timers.">, + MarshallingInfoFlag<CodeGenOpts<"TimePassesStatsFile">>; def fdump_record_layouts_simple : Flag<["-"], "fdump-record-layouts-simple">, HelpText<"Dump record layout information in a simple form used for testing">, MarshallingInfoFlag<LangOpts<"DumpRecordLayoutsSimple">>; diff --git a/clang/include/clang/Driver/ToolChain.h b/clang/include/clang/Driver/ToolChain.h index b8899e7..2430563 100644 --- a/clang/include/clang/Driver/ToolChain.h +++ b/clang/include/clang/Driver/ToolChain.h @@ -202,10 +202,6 @@ protected: ToolChain(const Driver &D, const llvm::Triple &T, const llvm::opt::ArgList &Args); - /// Executes the given \p Executable and returns the stdout. - llvm::Expected<std::unique_ptr<llvm::MemoryBuffer>> - executeToolChainProgram(StringRef Executable) const; - void setTripleEnvironment(llvm::Triple::EnvironmentType Env); virtual Tool *buildAssembler() const; @@ -806,7 +802,8 @@ public: /// Get paths for device libraries. virtual llvm::SmallVector<BitCodeLibraryInfo, 12> - getDeviceLibs(const llvm::opt::ArgList &Args) const; + getDeviceLibs(const llvm::opt::ArgList &Args, + const Action::OffloadKind DeviceOffloadingKind) const; /// Add the system specific linker arguments to use /// for the given HIP runtime library type. diff --git a/clang/include/clang/Sema/Sema.h b/clang/include/clang/Sema/Sema.h index b331acb..73eb730 100644 --- a/clang/include/clang/Sema/Sema.h +++ b/clang/include/clang/Sema/Sema.h @@ -834,6 +834,13 @@ enum class CCEKind { ///< message. }; +/// Enums for the diagnostics of target, target_version and target_clones. +namespace DiagAttrParams { +enum DiagType { Unsupported, Duplicate, Unknown }; +enum Specifier { None, CPU, Tune }; +enum AttrName { Target, TargetClones, TargetVersion }; +} // end namespace DiagAttrParams + void inferNoReturnAttr(Sema &S, const Decl *D); /// Sema - This implements semantic analysis and AST building for C. @@ -4922,13 +4929,6 @@ public: // handled later in the process, once we know how many exist. bool checkTargetAttr(SourceLocation LiteralLoc, StringRef Str); - /// Check Target Version attrs - bool checkTargetVersionAttr(SourceLocation Loc, Decl *D, StringRef Str); - bool checkTargetClonesAttrString( - SourceLocation LiteralLoc, StringRef Str, const StringLiteral *Literal, - Decl *D, bool &HasDefault, bool &HasCommas, bool &HasNotDefault, - SmallVectorImpl<SmallString<64>> &StringsBuffer); - ErrorAttr *mergeErrorAttr(Decl *D, const AttributeCommonInfo &CI, StringRef NewUserDiagnostic); FormatAttr *mergeFormatAttr(Decl *D, const AttributeCommonInfo &CI, diff --git a/clang/include/clang/Sema/SemaARM.h b/clang/include/clang/Sema/SemaARM.h index 788a7ab..e77d65f 100644 --- a/clang/include/clang/Sema/SemaARM.h +++ b/clang/include/clang/Sema/SemaARM.h @@ -91,6 +91,11 @@ public: /// Return true if the given vector types are lax-compatible SVE vector types, /// false otherwise. bool areLaxCompatibleSveTypes(QualType FirstType, QualType SecondType); + + bool checkTargetVersionAttr(const StringRef Str, const SourceLocation Loc); + bool checkTargetClonesAttr(SmallVectorImpl<StringRef> &Params, + SmallVectorImpl<SourceLocation> &Locs, + SmallVectorImpl<SmallString<64>> &NewParams); }; SemaARM::ArmStreamingType getArmStreamingFnType(const FunctionDecl *FD); diff --git a/clang/include/clang/Sema/SemaRISCV.h b/clang/include/clang/Sema/SemaRISCV.h index 8d2e1c6..844cc3c 100644 --- a/clang/include/clang/Sema/SemaRISCV.h +++ b/clang/include/clang/Sema/SemaRISCV.h @@ -55,6 +55,11 @@ public: bool DeclareAndesVectorBuiltins = false; std::unique_ptr<sema::RISCVIntrinsicManager> IntrinsicManager; + + bool checkTargetVersionAttr(const StringRef Param, const SourceLocation Loc); + bool checkTargetClonesAttr(SmallVectorImpl<StringRef> &Params, + SmallVectorImpl<SourceLocation> &Locs, + SmallVectorImpl<SmallString<64>> &NewParams); }; std::unique_ptr<sema::RISCVIntrinsicManager> diff --git a/clang/include/clang/Sema/SemaX86.h b/clang/include/clang/Sema/SemaX86.h index b5a23f1..20783e3 100644 --- a/clang/include/clang/Sema/SemaX86.h +++ b/clang/include/clang/Sema/SemaX86.h @@ -37,6 +37,10 @@ public: void handleAnyInterruptAttr(Decl *D, const ParsedAttr &AL); void handleForceAlignArgPointerAttr(Decl *D, const ParsedAttr &AL); + + bool checkTargetClonesAttr(SmallVectorImpl<StringRef> &Params, + SmallVectorImpl<SourceLocation> &Locs, + SmallVectorImpl<SmallString<64>> &NewParams); }; } // namespace clang diff --git a/clang/include/clang/StaticAnalyzer/Frontend/CheckerRegistry.h b/clang/include/clang/StaticAnalyzer/Frontend/CheckerRegistry.h index 43dbfb1..da3efd7 100644 --- a/clang/include/clang/StaticAnalyzer/Frontend/CheckerRegistry.h +++ b/clang/include/clang/StaticAnalyzer/Frontend/CheckerRegistry.h @@ -38,29 +38,29 @@ // function clang_registerCheckers. For example: // // extern "C" -// void clang_registerCheckers (CheckerRegistry ®istry) { -// registry.addChecker<MainCallChecker>("example.MainCallChecker", -// "Disallows calls to functions called main"); +// void clang_registerCheckers(CheckerRegistry &Registry) { +// Registry.addChecker<MainCallChecker>( +// "example.MainCallChecker", +// "Disallows calls to functions called main"); // } // -// The first method argument is the full name of the checker, including its -// enclosing package. By convention, the registered name of a checker is the -// name of the associated class (the template argument). -// The second method argument is a short human-readable description of the -// checker. +// The first argument of this templated method is the full name of the checker +// (including its package), while the second argument is a short description +// that is printed by `-analyzer-checker-help`. // -// The clang_registerCheckers function may add any number of checkers to the -// registry. If any checkers require additional initialization, use the three- -// argument form of CheckerRegistry::addChecker. +// A plugin may register several separate checkers by calling `addChecker()` +// multiple times. If a checker requires custom registration functions (e.g. +// checker option handling) use the non-templated overload of `addChecker` that +// takes two callback functions as the first two parameters. // // To load a checker plugin, specify the full path to the dynamic library as // the argument to the -load option in the cc1 frontend. You can then enable // your custom checker using the -analyzer-checker: // -// clang -cc1 -load </path/to/plugin.dylib> -analyze -// -analyzer-checker=<example.MainCallChecker> +// clang -cc1 -load /path/to/plugin.dylib -analyze +// -analyzer-checker=example.MainCallChecker // -// For a complete working example, see examples/analyzer-plugin. +// For complete examples, see clang/lib/Analysis/plugins/SampleAnalyzer #ifndef CLANG_ANALYZER_API_VERSION_STRING // FIXME: The Clang version string is not particularly granular; @@ -108,30 +108,25 @@ private: mgr.template registerChecker<T>(); } - template <typename T> static bool returnTrue(const CheckerManager &mgr) { - return true; - } + static bool returnTrue(const CheckerManager &) { return true; } public: - /// Adds a checker to the registry. Use this non-templated overload when your - /// checker requires custom initialization. - void addChecker(RegisterCheckerFn Fn, ShouldRegisterFunction sfn, - StringRef FullName, StringRef Desc, StringRef DocsUri, - bool IsHidden); - - /// Adds a checker to the registry. Use this templated overload when your - /// checker does not require any custom initialization. - /// This function isn't really needed and probably causes more headaches than - /// the tiny convenience that it provides, but external plugins might use it, - /// and there isn't a strong incentive to remove it. + /// Adds a checker to the registry. + /// Use this for a checker defined in a plugin if it requires custom + /// registration functions (e.g. for handling checker options). + /// NOTE: As of now `DocsUri` is never queried from the checker registry. + void addChecker(RegisterCheckerFn Fn, ShouldRegisterFunction Sfn, + StringRef FullName, StringRef Desc, + StringRef DocsUri = "NoDocsUri", bool IsHidden = false); + + /// Adds a checker to the registry. + /// Use this for a checker defined in a plugin if it doesn't require custom + /// registration functions. template <class T> - void addChecker(StringRef FullName, StringRef Desc, StringRef DocsUri, - bool IsHidden = false) { - // Avoid MSVC's Compiler Error C2276: - // http://msdn.microsoft.com/en-us/library/850cstw1(v=VS.80).aspx + void addChecker(StringRef FullName, StringRef Desc, + StringRef DocsUri = "NoDocsUri", bool IsHidden = false) { addChecker(&CheckerRegistry::initializeManager<CheckerManager, T>, - &CheckerRegistry::returnTrue<T>, FullName, Desc, DocsUri, - IsHidden); + &CheckerRegistry::returnTrue, FullName, Desc, DocsUri, IsHidden); } /// Makes the checker with the full name \p fullName depend on the checker diff --git a/clang/include/clang/Tooling/Inclusions/IncludeStyle.h b/clang/include/clang/Tooling/Inclusions/IncludeStyle.h index fba90d8..bf06061 100644 --- a/clang/include/clang/Tooling/Inclusions/IncludeStyle.h +++ b/clang/include/clang/Tooling/Inclusions/IncludeStyle.h @@ -126,8 +126,8 @@ struct IncludeStyle { /// * ``""`` means "arbitrary suffix" /// * ``"$"`` means "no suffix" /// - /// For example, if configured to ``"(_test)?$"``, then a header a.h would be seen - /// as the "main" include in both a.cc and a_test.cc. + /// For example, if configured to ``"(_test)?$"``, then a header a.h would be + /// seen as the "main" include in both a.cc and a_test.cc. /// \version 3.9 std::string IncludeIsMainRegex; diff --git a/clang/lib/AST/ByteCode/Compiler.cpp b/clang/lib/AST/ByteCode/Compiler.cpp index 07efd6f8..63ac536 100644 --- a/clang/lib/AST/ByteCode/Compiler.cpp +++ b/clang/lib/AST/ByteCode/Compiler.cpp @@ -25,34 +25,6 @@ using APSInt = llvm::APSInt; namespace clang { namespace interp { -static bool refersToUnion(const Expr *E) { - for (;;) { - if (const auto *ME = dyn_cast<MemberExpr>(E)) { - if (const auto *FD = dyn_cast<FieldDecl>(ME->getMemberDecl()); - FD && FD->getParent()->isUnion()) - return true; - E = ME->getBase(); - continue; - } - - if (const auto *ASE = dyn_cast<ArraySubscriptExpr>(E)) { - E = ASE->getBase()->IgnoreImplicit(); - continue; - } - - if (const auto *ICE = dyn_cast<ImplicitCastExpr>(E); - ICE && (ICE->getCastKind() == CK_NoOp || - ICE->getCastKind() == CK_DerivedToBase || - ICE->getCastKind() == CK_UncheckedDerivedToBase)) { - E = ICE->getSubExpr(); - continue; - } - - break; - } - return false; -} - static std::optional<bool> getBoolValue(const Expr *E) { if (const auto *CE = dyn_cast_if_present<ConstantExpr>(E); CE && CE->hasAPValueResult() && @@ -5401,6 +5373,53 @@ bool Compiler<Emitter>::maybeEmitDeferredVarInit(const VarDecl *VD) { return true; } +static bool hasTrivialDefaultCtorParent(const FieldDecl *FD) { + assert(FD); + assert(FD->getParent()->isUnion()); + const auto *CXXRD = dyn_cast<CXXRecordDecl>(FD->getParent()); + return !CXXRD || CXXRD->hasTrivialDefaultConstructor(); +} + +template <class Emitter> bool Compiler<Emitter>::refersToUnion(const Expr *E) { + for (;;) { + if (const auto *ME = dyn_cast<MemberExpr>(E)) { + if (const auto *FD = dyn_cast<FieldDecl>(ME->getMemberDecl()); + FD && FD->getParent()->isUnion() && hasTrivialDefaultCtorParent(FD)) + return true; + E = ME->getBase(); + continue; + } + + if (const auto *ASE = dyn_cast<ArraySubscriptExpr>(E)) { + E = ASE->getBase()->IgnoreImplicit(); + continue; + } + + if (const auto *ICE = dyn_cast<ImplicitCastExpr>(E); + ICE && (ICE->getCastKind() == CK_NoOp || + ICE->getCastKind() == CK_DerivedToBase || + ICE->getCastKind() == CK_UncheckedDerivedToBase)) { + E = ICE->getSubExpr(); + continue; + } + + if (const auto *This = dyn_cast<CXXThisExpr>(E)) { + const auto *ThisRecord = + This->getType()->getPointeeType()->getAsRecordDecl(); + if (!ThisRecord->isUnion()) + return false; + // Otherwise, always activate if we're in the ctor. + if (const auto *Ctor = + dyn_cast_if_present<CXXConstructorDecl>(CompilingFunction)) + return Ctor->getParent() == ThisRecord; + return false; + } + + break; + } + return false; +} + template <class Emitter> bool Compiler<Emitter>::visitDeclStmt(const DeclStmt *DS, bool EvaluateConditionDecl) { @@ -5933,16 +5952,15 @@ bool Compiler<Emitter>::compileConstructor(const CXXConstructorDecl *Ctor) { return false; if (OptPrimType T = this->classify(InitExpr)) { + if (Activate && !this->emitActivateThisField(FieldOffset, InitExpr)) + return false; + if (!this->visit(InitExpr)) return false; bool BitField = F->isBitField(); - if (BitField && Activate) - return this->emitInitThisBitFieldActivate(*T, F, FieldOffset, InitExpr); if (BitField) return this->emitInitThisBitField(*T, F, FieldOffset, InitExpr); - if (Activate) - return this->emitInitThisFieldActivate(*T, FieldOffset, InitExpr); return this->emitInitThisField(*T, FieldOffset, InitExpr); } // Non-primitive case. Get a pointer to the field-to-initialize diff --git a/clang/lib/AST/ByteCode/Compiler.h b/clang/lib/AST/ByteCode/Compiler.h index 5032693..3a26342 100644 --- a/clang/lib/AST/ByteCode/Compiler.h +++ b/clang/lib/AST/ByteCode/Compiler.h @@ -401,6 +401,8 @@ private: bool checkLiteralType(const Expr *E); bool maybeEmitDeferredVarInit(const VarDecl *VD); + bool refersToUnion(const Expr *E); + protected: /// Variable to storage mapping. llvm::DenseMap<const ValueDecl *, Scope::Local> Locals; diff --git a/clang/lib/AST/ByteCode/Interp.h b/clang/lib/AST/ByteCode/Interp.h index 7f29200..9012442 100644 --- a/clang/lib/AST/ByteCode/Interp.h +++ b/clang/lib/AST/ByteCode/Interp.h @@ -1983,6 +1983,16 @@ static inline bool Activate(InterpState &S, CodePtr OpPC) { return true; } +static inline bool ActivateThisField(InterpState &S, CodePtr OpPC, uint32_t I) { + if (S.checkingPotentialConstantExpression()) + return false; + + const Pointer &Ptr = S.Current->getThis(); + assert(Ptr.atField(I).canBeInitialized()); + Ptr.atField(I).activate(); + return true; +} + template <PrimType Name, class T = typename PrimConv<Name>::T> bool StoreActivate(InterpState &S, CodePtr OpPC) { const T &Value = S.Stk.pop<T>(); @@ -3557,12 +3567,22 @@ inline bool BitCastPrim(InterpState &S, CodePtr OpPC, bool TargetIsUCharOrByte, Floating Result = S.allocFloat(*Sem); Floating::bitcastFromMemory(Buff.data(), *Sem, &Result); S.Stk.push<Floating>(Result); - - // S.Stk.push<Floating>(T::bitcastFromMemory(Buff.data(), *Sem)); } else if constexpr (needsAlloc<T>()) { T Result = S.allocAP<T>(ResultBitWidth); T::bitcastFromMemory(Buff.data(), ResultBitWidth, &Result); S.Stk.push<T>(Result); + } else if constexpr (std::is_same_v<T, Boolean>) { + // Only allow to cast single-byte integers to bool if they are either 0 + // or 1. + assert(FullBitWidth.getQuantity() == 8); + auto Val = static_cast<unsigned int>(Buff[0]); + if (Val > 1) { + S.FFDiag(S.Current->getSource(OpPC), + diag::note_constexpr_bit_cast_unrepresentable_value) + << S.getASTContext().BoolTy << Val; + return false; + } + S.Stk.push<T>(T::bitcastFromMemory(Buff.data(), ResultBitWidth)); } else { assert(!Sem); S.Stk.push<T>(T::bitcastFromMemory(Buff.data(), ResultBitWidth)); diff --git a/clang/lib/AST/ByteCode/Opcodes.td b/clang/lib/AST/ByteCode/Opcodes.td index 80703ad..abfed77 100644 --- a/clang/lib/AST/ByteCode/Opcodes.td +++ b/clang/lib/AST/ByteCode/Opcodes.td @@ -510,6 +510,7 @@ def StoreBitFieldActivate : StoreBitFieldOpcode {} def StoreBitFieldActivatePop : StoreBitFieldOpcode {} def Activate : Opcode {} +def ActivateThisField : Opcode { let Args = [ArgUint32]; } // [Pointer, Value] -> [] def Init : StoreOpcode {} diff --git a/clang/lib/AST/TextNodeDumper.cpp b/clang/lib/AST/TextNodeDumper.cpp index 3d9397f..6b524cf 100644 --- a/clang/lib/AST/TextNodeDumper.cpp +++ b/clang/lib/AST/TextNodeDumper.cpp @@ -843,7 +843,10 @@ void TextNodeDumper::Visit(const APValue &Value, QualType Ty) { } ColorScope Color(OS, ShowColors, DeclNameColor); - OS << Value.getMemberPointerDecl()->getDeclName(); + if (const ValueDecl *MemDecl = Value.getMemberPointerDecl()) + OS << MemDecl->getDeclName(); + else + OS << "null"; return; } case APValue::AddrLabelDiff: diff --git a/clang/lib/Analysis/LifetimeSafety.cpp b/clang/lib/Analysis/LifetimeSafety.cpp index a95db6d..94b8197 100644 --- a/clang/lib/Analysis/LifetimeSafety.cpp +++ b/clang/lib/Analysis/LifetimeSafety.cpp @@ -23,9 +23,10 @@ #include "llvm/Support/Debug.h" #include "llvm/Support/TimeProfiler.h" #include <cstdint> +#include <memory> -namespace clang { -namespace { +namespace clang::lifetimes { +namespace internal { /// Represents the storage location being borrowed, e.g., a specific stack /// variable. @@ -36,32 +37,6 @@ struct AccessPath { AccessPath(const clang::ValueDecl *D) : D(D) {} }; -/// A generic, type-safe wrapper for an ID, distinguished by its `Tag` type. -/// Used for giving ID to loans and origins. -template <typename Tag> struct ID { - uint32_t Value = 0; - - bool operator==(const ID<Tag> &Other) const { return Value == Other.Value; } - bool operator!=(const ID<Tag> &Other) const { return !(*this == Other); } - bool operator<(const ID<Tag> &Other) const { return Value < Other.Value; } - ID<Tag> operator++(int) { - ID<Tag> Tmp = *this; - ++Value; - return Tmp; - } - void Profile(llvm::FoldingSetNodeID &IDBuilder) const { - IDBuilder.AddInteger(Value); - } -}; - -template <typename Tag> -inline llvm::raw_ostream &operator<<(llvm::raw_ostream &OS, ID<Tag> ID) { - return OS << ID.Value; -} - -using LoanID = ID<struct LoanTag>; -using OriginID = ID<struct OriginTag>; - /// Information about a single borrow, or "Loan". A loan is created when a /// reference or pointer is created. struct Loan { @@ -223,7 +198,9 @@ public: /// An origin is propagated from a source to a destination (e.g., p = q). AssignOrigin, /// An origin escapes the function by flowing into the return value. - ReturnOfOrigin + ReturnOfOrigin, + /// A marker for a specific point in the code, for testing. + TestPoint, }; private: @@ -310,6 +287,24 @@ public: } }; +/// A dummy-fact used to mark a specific point in the code for testing. +/// It is generated by recognizing a `void("__lifetime_test_point_...")` cast. +class TestPointFact : public Fact { + StringRef Annotation; + +public: + static bool classof(const Fact *F) { return F->getKind() == Kind::TestPoint; } + + explicit TestPointFact(StringRef Annotation) + : Fact(Kind::TestPoint), Annotation(Annotation) {} + + StringRef getAnnotation() const { return Annotation; } + + void dump(llvm::raw_ostream &OS) const override { + OS << "TestPoint (Annotation: \"" << getAnnotation() << "\")\n"; + } +}; + class FactManager { public: llvm::ArrayRef<const Fact *> getFacts(const CFGBlock *B) const { @@ -363,6 +358,7 @@ private: }; class FactGenerator : public ConstStmtVisitor<FactGenerator> { + using Base = ConstStmtVisitor<FactGenerator>; public: FactGenerator(FactManager &FactMgr, AnalysisDeclContext &AC) @@ -458,6 +454,15 @@ public: } } + void VisitCXXFunctionalCastExpr(const CXXFunctionalCastExpr *FCE) { + // Check if this is a test point marker. If so, we are done with this + // expression. + if (VisitTestPoint(FCE)) + return; + // Visit as normal otherwise. + Base::VisitCXXFunctionalCastExpr(FCE); + } + private: // Check if a type has an origin. bool hasOrigin(QualType QT) { return QT->isPointerOrReferenceType(); } @@ -491,6 +496,27 @@ private: } } + /// Checks if the expression is a `void("__lifetime_test_point_...")` cast. + /// If so, creates a `TestPointFact` and returns true. + bool VisitTestPoint(const CXXFunctionalCastExpr *FCE) { + if (!FCE->getType()->isVoidType()) + return false; + + const auto *SubExpr = FCE->getSubExpr()->IgnoreParenImpCasts(); + if (const auto *SL = dyn_cast<StringLiteral>(SubExpr)) { + llvm::StringRef LiteralValue = SL->getString(); + const std::string Prefix = "__lifetime_test_point_"; + + if (LiteralValue.starts_with(Prefix)) { + StringRef Annotation = LiteralValue.drop_front(Prefix.length()); + CurrentBlockFacts.push_back( + FactMgr.createFact<TestPointFact>(Annotation)); + return true; + } + } + return false; + } + FactManager &FactMgr; AnalysisDeclContext &AC; llvm::SmallVector<Fact *> CurrentBlockFacts; @@ -637,6 +663,8 @@ private: return D->transfer(In, *F->getAs<AssignOriginFact>()); case Fact::Kind::ReturnOfOrigin: return D->transfer(In, *F->getAs<ReturnOfOriginFact>()); + case Fact::Kind::TestPoint: + return D->transfer(In, *F->getAs<TestPointFact>()); } llvm_unreachable("Unknown fact kind"); } @@ -646,14 +674,16 @@ public: Lattice transfer(Lattice In, const ExpireFact &) { return In; } Lattice transfer(Lattice In, const AssignOriginFact &) { return In; } Lattice transfer(Lattice In, const ReturnOfOriginFact &) { return In; } + Lattice transfer(Lattice In, const TestPointFact &) { return In; } }; namespace utils { /// Computes the union of two ImmutableSets. template <typename T> -llvm::ImmutableSet<T> join(llvm::ImmutableSet<T> A, llvm::ImmutableSet<T> B, - typename llvm::ImmutableSet<T>::Factory &F) { +static llvm::ImmutableSet<T> join(llvm::ImmutableSet<T> A, + llvm::ImmutableSet<T> B, + typename llvm::ImmutableSet<T>::Factory &F) { if (A.getHeight() < B.getHeight()) std::swap(A, B); for (const T &E : B) @@ -666,7 +696,7 @@ llvm::ImmutableSet<T> join(llvm::ImmutableSet<T> A, llvm::ImmutableSet<T> B, // efficient merge could be implemented using a Patricia Trie or HAMT // instead of the current AVL-tree-based ImmutableMap. template <typename K, typename V, typename Joiner> -llvm::ImmutableMap<K, V> +static llvm::ImmutableMap<K, V> join(llvm::ImmutableMap<K, V> A, llvm::ImmutableMap<K, V> B, typename llvm::ImmutableMap<K, V>::Factory &F, Joiner joinValues) { if (A.getHeight() < B.getHeight()) @@ -690,10 +720,6 @@ join(llvm::ImmutableMap<K, V> A, llvm::ImmutableMap<K, V> B, // Loan Propagation Analysis // ========================================================================= // -// Using LLVM's immutable collections is efficient for dataflow analysis -// as it avoids deep copies during state transitions. -// TODO(opt): Consider using a bitset to represent the set of loans. -using LoanSet = llvm::ImmutableSet<LoanID>; using OriginLoanMap = llvm::ImmutableMap<OriginID, LoanSet>; /// An object to hold the factories for immutable collections, ensuring @@ -802,22 +828,118 @@ private: }; // ========================================================================= // +// Expired Loans Analysis +// ========================================================================= // + +/// The dataflow lattice for tracking the set of expired loans. +struct ExpiredLattice { + LoanSet Expired; + + ExpiredLattice() : Expired(nullptr) {}; + explicit ExpiredLattice(LoanSet S) : Expired(S) {} + + bool operator==(const ExpiredLattice &Other) const { + return Expired == Other.Expired; + } + bool operator!=(const ExpiredLattice &Other) const { + return !(*this == Other); + } + + void dump(llvm::raw_ostream &OS) const { + OS << "ExpiredLattice State:\n"; + if (Expired.isEmpty()) + OS << " <empty>\n"; + for (const LoanID &LID : Expired) + OS << " Loan " << LID << " is expired\n"; + } +}; + +/// The analysis that tracks which loans have expired. +class ExpiredLoansAnalysis + : public DataflowAnalysis<ExpiredLoansAnalysis, ExpiredLattice, + Direction::Forward> { + + LoanSet::Factory &Factory; + +public: + ExpiredLoansAnalysis(const CFG &C, AnalysisDeclContext &AC, FactManager &F, + LifetimeFactory &Factory) + : DataflowAnalysis(C, AC, F), Factory(Factory.LoanSetFactory) {} + + using Base::transfer; + + StringRef getAnalysisName() const { return "ExpiredLoans"; } + + Lattice getInitialState() { return Lattice(Factory.getEmptySet()); } + + /// Merges two lattices by taking the union of the expired loan sets. + Lattice join(Lattice L1, Lattice L2) const { + return Lattice(utils::join(L1.Expired, L2.Expired, Factory)); + } + + Lattice transfer(Lattice In, const ExpireFact &F) { + return Lattice(Factory.add(In.Expired, F.getLoanID())); + } + + // Removes the loan from the set of expired loans. + // + // When a loan is re-issued (e.g., in a loop), it is no longer considered + // expired. A loan can be in the expired set at the point of issue due to + // the dataflow state from a previous loop iteration being propagated along + // a backedge in the CFG. + // + // Note: This has a subtle false-negative though where a loan from previous + // iteration is not overwritten by a reissue. This needs careful tracking + // of loans "across iterations" which can be considered for future + // enhancements. + // + // void foo(int safe) { + // int* p = &safe; + // int* q = &safe; + // while (condition()) { + // int x = 1; + // p = &x; // A loan to 'x' is issued to 'p' in every iteration. + // if (condition()) { + // q = p; + // } + // (void)*p; // OK — 'p' points to 'x' from new iteration. + // (void)*q; // UaF - 'q' still points to 'x' from previous iteration + // // which is now destroyed. + // } + // } + Lattice transfer(Lattice In, const IssueFact &F) { + return Lattice(Factory.remove(In.Expired, F.getLoanID())); + } +}; + +// ========================================================================= // // TODO: // - Modify loan expiry analysis to answer `bool isExpired(Loan L, Point P)` // - Modify origin liveness analysis to answer `bool isLive(Origin O, Point P)` // - Using the above three to perform the final error reporting. // ========================================================================= // -} // anonymous namespace -void runLifetimeSafetyAnalysis(const DeclContext &DC, const CFG &Cfg, - AnalysisDeclContext &AC) { +// ========================================================================= // +// LifetimeSafetyAnalysis Class Implementation +// ========================================================================= // + +// We need this here for unique_ptr with forward declared class. +LifetimeSafetyAnalysis::~LifetimeSafetyAnalysis() = default; + +LifetimeSafetyAnalysis::LifetimeSafetyAnalysis(AnalysisDeclContext &AC) + : AC(AC), Factory(std::make_unique<LifetimeFactory>()), + FactMgr(std::make_unique<FactManager>()) {} + +void LifetimeSafetyAnalysis::run() { llvm::TimeTraceScope TimeProfile("LifetimeSafetyAnalysis"); + + const CFG &Cfg = *AC.getCFG(); DEBUG_WITH_TYPE("PrintCFG", Cfg.dump(AC.getASTContext().getLangOpts(), /*ShowColors=*/true)); - FactManager FactMgr; - FactGenerator FactGen(FactMgr, AC); + + FactGenerator FactGen(*FactMgr, AC); FactGen.run(); - DEBUG_WITH_TYPE("LifetimeFacts", FactMgr.dump(Cfg, AC)); + DEBUG_WITH_TYPE("LifetimeFacts", FactMgr->dump(Cfg, AC)); /// TODO(opt): Consider optimizing individual blocks before running the /// dataflow analysis. @@ -828,9 +950,65 @@ void runLifetimeSafetyAnalysis(const DeclContext &DC, const CFG &Cfg, /// blocks; only Decls are visible. Therefore, loans in a block that /// never reach an Origin associated with a Decl can be safely dropped by /// the analysis. - LifetimeFactory Factory; - LoanPropagationAnalysis LoanPropagation(Cfg, AC, FactMgr, Factory); - LoanPropagation.run(); - DEBUG_WITH_TYPE("LifetimeLoanPropagation", LoanPropagation.dump()); + LoanPropagation = + std::make_unique<LoanPropagationAnalysis>(Cfg, AC, *FactMgr, *Factory); + LoanPropagation->run(); + + ExpiredLoans = + std::make_unique<ExpiredLoansAnalysis>(Cfg, AC, *FactMgr, *Factory); + ExpiredLoans->run(); +} + +LoanSet LifetimeSafetyAnalysis::getLoansAtPoint(OriginID OID, + ProgramPoint PP) const { + assert(LoanPropagation && "Analysis has not been run."); + return LoanPropagation->getLoans(OID, PP); +} + +LoanSet LifetimeSafetyAnalysis::getExpiredLoansAtPoint(ProgramPoint PP) const { + assert(ExpiredLoans && "ExpiredLoansAnalysis has not been run."); + return ExpiredLoans->getState(PP).Expired; +} + +std::optional<OriginID> +LifetimeSafetyAnalysis::getOriginIDForDecl(const ValueDecl *D) const { + assert(FactMgr && "FactManager not initialized"); + // This assumes the OriginManager's `get` can find an existing origin. + // We might need a `find` method on OriginManager to avoid `getOrCreate` logic + // in a const-query context if that becomes an issue. + return FactMgr->getOriginMgr().get(*D); +} + +std::vector<LoanID> +LifetimeSafetyAnalysis::getLoanIDForVar(const VarDecl *VD) const { + assert(FactMgr && "FactManager not initialized"); + std::vector<LoanID> Result; + for (const Loan &L : FactMgr->getLoanMgr().getLoans()) + if (L.Path.D == VD) + Result.push_back(L.ID); + return Result; +} + +llvm::StringMap<ProgramPoint> LifetimeSafetyAnalysis::getTestPoints() const { + assert(FactMgr && "FactManager not initialized"); + llvm::StringMap<ProgramPoint> AnnotationToPointMap; + for (const CFGBlock *Block : *AC.getCFG()) { + for (const Fact *F : FactMgr->getFacts(Block)) { + if (const auto *TPF = F->getAs<TestPointFact>()) { + StringRef PointName = TPF->getAnnotation(); + assert(AnnotationToPointMap.find(PointName) == + AnnotationToPointMap.end() && + "more than one test points with the same name"); + AnnotationToPointMap[PointName] = F; + } + } + } + return AnnotationToPointMap; +} +} // namespace internal + +void runLifetimeSafetyAnalysis(AnalysisDeclContext &AC) { + internal::LifetimeSafetyAnalysis Analysis(AC); + Analysis.run(); } -} // namespace clang +} // namespace clang::lifetimes diff --git a/clang/lib/Analysis/plugins/CheckerDependencyHandling/CheckerDependencyHandling.cpp b/clang/lib/Analysis/plugins/CheckerDependencyHandling/CheckerDependencyHandling.cpp index aacb886..518f9e7 100644 --- a/clang/lib/Analysis/plugins/CheckerDependencyHandling/CheckerDependencyHandling.cpp +++ b/clang/lib/Analysis/plugins/CheckerDependencyHandling/CheckerDependencyHandling.cpp @@ -2,6 +2,9 @@ #include "clang/StaticAnalyzer/Core/PathSensitive/CheckerContext.h" #include "clang/StaticAnalyzer/Frontend/CheckerRegistry.h" +// This barebones plugin is used by clang/test/Analysis/checker-plugins.c +// to test dependency handling among checkers loaded from plugins. + using namespace clang; using namespace ento; @@ -15,12 +18,12 @@ struct DependendentChecker : public Checker<check::BeginFunction> { } // end anonymous namespace // Register plugin! -extern "C" void clang_registerCheckers(CheckerRegistry ®istry) { - registry.addChecker<Dependency>("example.Dependency", "", ""); - registry.addChecker<DependendentChecker>("example.DependendentChecker", "", - ""); +extern "C" void clang_registerCheckers(CheckerRegistry &Registry) { + Registry.addChecker<Dependency>("example.Dependency", "MockDescription"); + Registry.addChecker<DependendentChecker>("example.DependendentChecker", + "MockDescription"); - registry.addDependency("example.DependendentChecker", "example.Dependency"); + Registry.addDependency("example.DependendentChecker", "example.Dependency"); } extern "C" const char clang_analyzerAPIVersionString[] = diff --git a/clang/lib/Analysis/plugins/CheckerOptionHandling/CheckerOptionHandling.cpp b/clang/lib/Analysis/plugins/CheckerOptionHandling/CheckerOptionHandling.cpp index 82c1058..2adb934 100644 --- a/clang/lib/Analysis/plugins/CheckerOptionHandling/CheckerOptionHandling.cpp +++ b/clang/lib/Analysis/plugins/CheckerOptionHandling/CheckerOptionHandling.cpp @@ -5,6 +5,9 @@ using namespace clang; using namespace ento; +// This barebones plugin is used by clang/test/Analysis/checker-plugins.c +// to test option handling on checkers loaded from plugins. + namespace { struct MyChecker : public Checker<check::BeginFunction> { void checkBeginFunction(CheckerContext &Ctx) const {} @@ -25,13 +28,11 @@ bool shouldRegisterMyChecker(const CheckerManager &mgr) { return true; } } // end anonymous namespace // Register plugin! -extern "C" void clang_registerCheckers(CheckerRegistry ®istry) { - registry.addChecker(registerMyChecker, shouldRegisterMyChecker, - "example.MyChecker", "Example Description", - "example.mychecker.documentation.nonexistent.html", - /*isHidden*/false); +extern "C" void clang_registerCheckers(CheckerRegistry &Registry) { + Registry.addChecker(registerMyChecker, shouldRegisterMyChecker, + "example.MyChecker", "Example Description"); - registry.addCheckerOption(/*OptionType*/ "bool", + Registry.addCheckerOption(/*OptionType*/ "bool", /*CheckerFullName*/ "example.MyChecker", /*OptionName*/ "ExampleOption", /*DefaultValStr*/ "false", diff --git a/clang/lib/Analysis/plugins/SampleAnalyzer/MainCallChecker.cpp b/clang/lib/Analysis/plugins/SampleAnalyzer/MainCallChecker.cpp index fd210d7..53a01d2 100644 --- a/clang/lib/Analysis/plugins/SampleAnalyzer/MainCallChecker.cpp +++ b/clang/lib/Analysis/plugins/SampleAnalyzer/MainCallChecker.cpp @@ -3,12 +3,16 @@ #include "clang/StaticAnalyzer/Core/PathSensitive/CheckerContext.h" #include "clang/StaticAnalyzer/Frontend/CheckerRegistry.h" +// This simple plugin is used by clang/test/Analysis/checker-plugins.c +// to test the use of a checker that is defined in a plugin. + using namespace clang; using namespace ento; namespace { class MainCallChecker : public Checker<check::PreStmt<CallExpr>> { - mutable std::unique_ptr<BugType> BT; + + const BugType BT{this, "call to main", "example analyzer plugin"}; public: void checkPreStmt(const CallExpr *CE, CheckerContext &C) const; @@ -33,21 +37,17 @@ void MainCallChecker::checkPreStmt(const CallExpr *CE, if (!N) return; - if (!BT) - BT.reset(new BugType(this, "call to main", "example analyzer plugin")); - auto report = - std::make_unique<PathSensitiveBugReport>(*BT, BT->getDescription(), N); + std::make_unique<PathSensitiveBugReport>(BT, BT.getDescription(), N); report->addRange(Callee->getSourceRange()); C.emitReport(std::move(report)); } } // Register plugin! -extern "C" void clang_registerCheckers(CheckerRegistry ®istry) { - registry.addChecker<MainCallChecker>( - "example.MainCallChecker", "Disallows calls to functions called main", - ""); +extern "C" void clang_registerCheckers(CheckerRegistry &Registry) { + Registry.addChecker<MainCallChecker>("example.MainCallChecker", + "Example Description"); } extern "C" const char clang_analyzerAPIVersionString[] = diff --git a/clang/lib/Basic/TargetInfo.cpp b/clang/lib/Basic/TargetInfo.cpp index 09b6a1f..21fc084 100644 --- a/clang/lib/Basic/TargetInfo.cpp +++ b/clang/lib/Basic/TargetInfo.cpp @@ -172,7 +172,7 @@ TargetInfo::TargetInfo(const llvm::Triple &T) : Triple(T) { ComplexLongDoubleUsesFP2Ret = false; // Set the C++ ABI based on the triple. - TheCXXABI.set(Triple.isKnownWindowsMSVCEnvironment() + TheCXXABI.set(Triple.isKnownWindowsMSVCEnvironment() || Triple.isUEFI() ? TargetCXXABI::Microsoft : TargetCXXABI::GenericItanium); diff --git a/clang/lib/Basic/Targets.cpp b/clang/lib/Basic/Targets.cpp index 5c2af9b..e3f9760 100644 --- a/clang/lib/Basic/Targets.cpp +++ b/clang/lib/Basic/Targets.cpp @@ -757,6 +757,9 @@ std::unique_ptr<TargetInfo> AllocateTarget(const llvm::Triple &Triple, case llvm::Triple::FreeBSD: return std::make_unique<FreeBSDTargetInfo<LoongArch64TargetInfo>>(Triple, Opts); + case llvm::Triple::OpenBSD: + return std::make_unique<OpenBSDTargetInfo<LoongArch64TargetInfo>>(Triple, + Opts); default: return std::make_unique<LoongArch64TargetInfo>(Triple, Opts); } diff --git a/clang/lib/Basic/Targets/AArch64.cpp b/clang/lib/Basic/Targets/AArch64.cpp index 72d2e5f..2b023e5 100644 --- a/clang/lib/Basic/Targets/AArch64.cpp +++ b/clang/lib/Basic/Targets/AArch64.cpp @@ -786,7 +786,8 @@ AArch64TargetInfo::getVScaleRange(const LangOptions &LangOpts, return std::nullopt; } -uint64_t AArch64TargetInfo::getFMVPriority(ArrayRef<StringRef> Features) const { +llvm::APInt +AArch64TargetInfo::getFMVPriority(ArrayRef<StringRef> Features) const { return llvm::AArch64::getFMVPriority(Features); } diff --git a/clang/lib/Basic/Targets/AArch64.h b/clang/lib/Basic/Targets/AArch64.h index f4277e9..dfd89be 100644 --- a/clang/lib/Basic/Targets/AArch64.h +++ b/clang/lib/Basic/Targets/AArch64.h @@ -152,7 +152,7 @@ public: void fillValidCPUList(SmallVectorImpl<StringRef> &Values) const override; bool setCPU(const std::string &Name) override; - uint64_t getFMVPriority(ArrayRef<StringRef> Features) const override; + llvm::APInt getFMVPriority(ArrayRef<StringRef> Features) const override; bool useFP16ConversionIntrinsics() const override { return false; diff --git a/clang/lib/Basic/Targets/ARM.cpp b/clang/lib/Basic/Targets/ARM.cpp index 7ff8e51..29de34bb 100644 --- a/clang/lib/Basic/Targets/ARM.cpp +++ b/clang/lib/Basic/Targets/ARM.cpp @@ -623,13 +623,15 @@ bool ARMTargetInfo::handleTargetFeatures(std::vector<std::string> &Features, LDREX = LDREX_W; break; case 7: + case 8: if (ArchProfile == llvm::ARM::ProfileKind::M) LDREX = LDREX_W | LDREX_H | LDREX_B; else LDREX = LDREX_D | LDREX_W | LDREX_H | LDREX_B; break; - case 8: case 9: + assert(ArchProfile != llvm::ARM::ProfileKind::M && + "No Armv9-M architectures defined"); LDREX = LDREX_D | LDREX_W | LDREX_H | LDREX_B; } diff --git a/clang/lib/Basic/Targets/OSTargets.h b/clang/lib/Basic/Targets/OSTargets.h index 42cff65..94b018a 100644 --- a/clang/lib/Basic/Targets/OSTargets.h +++ b/clang/lib/Basic/Targets/OSTargets.h @@ -496,6 +496,7 @@ public: case llvm::Triple::sparcv9: this->MCountName = "_mcount"; break; + case llvm::Triple::loongarch64: case llvm::Triple::riscv64: break; } diff --git a/clang/lib/Basic/Targets/RISCV.cpp b/clang/lib/Basic/Targets/RISCV.cpp index 8a28c07..a6a5ec4 100644 --- a/clang/lib/Basic/Targets/RISCV.cpp +++ b/clang/lib/Basic/Targets/RISCV.cpp @@ -568,7 +568,8 @@ ParsedTargetAttr RISCVTargetInfo::parseTargetAttr(StringRef Features) const { return Ret; } -uint64_t RISCVTargetInfo::getFMVPriority(ArrayRef<StringRef> Features) const { +llvm::APInt +RISCVTargetInfo::getFMVPriority(ArrayRef<StringRef> Features) const { // Priority is explicitly specified on RISC-V unlike on other targets, where // it is derived by all the features of a specific version. Therefore if a // feature contains the priority string, then return it immediately. @@ -580,12 +581,12 @@ uint64_t RISCVTargetInfo::getFMVPriority(ArrayRef<StringRef> Features) const { Feature = RHS; else continue; - uint64_t Priority; + unsigned Priority; if (!Feature.getAsInteger(0, Priority)) - return Priority; + return llvm::APInt(32, Priority); } // Default Priority is zero. - return 0; + return llvm::APInt::getZero(32); } TargetInfo::CallingConvCheckResult diff --git a/clang/lib/Basic/Targets/RISCV.h b/clang/lib/Basic/Targets/RISCV.h index 8d629ab..58bfad1 100644 --- a/clang/lib/Basic/Targets/RISCV.h +++ b/clang/lib/Basic/Targets/RISCV.h @@ -123,7 +123,7 @@ public: void fillValidTuneCPUList(SmallVectorImpl<StringRef> &Values) const override; bool supportsTargetAttributeTune() const override { return true; } ParsedTargetAttr parseTargetAttr(StringRef Str) const override; - uint64_t getFMVPriority(ArrayRef<StringRef> Features) const override; + llvm::APInt getFMVPriority(ArrayRef<StringRef> Features) const override; std::pair<unsigned, unsigned> hardwareInterferenceSizes() const override { return std::make_pair(32, 32); diff --git a/clang/lib/Basic/Targets/X86.cpp b/clang/lib/Basic/Targets/X86.cpp index a1f5aa2..24ecec2 100644 --- a/clang/lib/Basic/Targets/X86.cpp +++ b/clang/lib/Basic/Targets/X86.cpp @@ -1390,8 +1390,8 @@ static llvm::X86::ProcessorFeatures getFeature(StringRef Name) { // correct, so it asserts if the value is out of range. } -uint64_t X86TargetInfo::getFMVPriority(ArrayRef<StringRef> Features) const { - auto getPriority = [](StringRef Feature) -> uint64_t { +llvm::APInt X86TargetInfo::getFMVPriority(ArrayRef<StringRef> Features) const { + auto getPriority = [](StringRef Feature) -> unsigned { // Valid CPUs have a 'key feature' that compares just better than its key // feature. using namespace llvm::X86; @@ -1405,11 +1405,11 @@ uint64_t X86TargetInfo::getFMVPriority(ArrayRef<StringRef> Features) const { return getFeaturePriority(getFeature(Feature)) << 1; }; - uint64_t Priority = 0; + unsigned Priority = 0; for (StringRef Feature : Features) if (!Feature.empty()) Priority = std::max(Priority, getPriority(Feature)); - return Priority; + return llvm::APInt(32, Priority); } bool X86TargetInfo::validateCPUSpecificCPUDispatch(StringRef Name) const { diff --git a/clang/lib/Basic/Targets/X86.h b/clang/lib/Basic/Targets/X86.h index ebc59c9..eb15103 100644 --- a/clang/lib/Basic/Targets/X86.h +++ b/clang/lib/Basic/Targets/X86.h @@ -388,7 +388,7 @@ public: return CPU != llvm::X86::CK_None; } - uint64_t getFMVPriority(ArrayRef<StringRef> Features) const override; + llvm::APInt getFMVPriority(ArrayRef<StringRef> Features) const override; bool setFPMath(StringRef Name) override; diff --git a/clang/lib/CIR/CodeGen/CIRGenExprComplex.cpp b/clang/lib/CIR/CodeGen/CIRGenExprComplex.cpp index 6756a7c..d52da0c 100644 --- a/clang/lib/CIR/CodeGen/CIRGenExprComplex.cpp +++ b/clang/lib/CIR/CodeGen/CIRGenExprComplex.cpp @@ -34,11 +34,20 @@ public: } mlir::Value emitLoadOfLValue(LValue lv, SourceLocation loc); + /// Store the specified real/imag parts into the /// specified value pointer. void emitStoreOfComplex(mlir::Location loc, mlir::Value val, LValue lv, bool isInit); + /// Emit a cast from complex value Val to DestType. + mlir::Value emitComplexToComplexCast(mlir::Value value, QualType srcType, + QualType destType, SourceLocation loc); + + /// Emit a cast from scalar value Val to DestType. + mlir::Value emitScalarToComplexCast(mlir::Value value, QualType srcType, + QualType destType, SourceLocation loc); + mlir::Value VisitAbstractConditionalOperator(const AbstractConditionalOperator *e); mlir::Value VisitArraySubscriptExpr(Expr *e); @@ -164,14 +173,106 @@ LValue ComplexExprEmitter::emitBinAssignLValue(const BinaryOperator *e, mlir::Value ComplexExprEmitter::emitCast(CastKind ck, Expr *op, QualType destTy) { switch (ck) { + case CK_Dependent: + llvm_unreachable("dependent type must be resolved before the CIR codegen"); + case CK_NoOp: case CK_LValueToRValue: return Visit(op); - default: - break; + + case CK_AtomicToNonAtomic: + case CK_NonAtomicToAtomic: + case CK_UserDefinedConversion: { + cgf.cgm.errorNYI( + "ComplexExprEmitter::emitCast Atmoic & UserDefinedConversion"); + return {}; } - cgf.cgm.errorNYI("ComplexType Cast"); - return {}; + + case CK_LValueBitCast: { + cgf.cgm.errorNYI("ComplexExprEmitter::emitCast CK_LValueBitCast"); + return {}; + } + + case CK_LValueToRValueBitCast: { + cgf.cgm.errorNYI("ComplexExprEmitter::emitCast CK_LValueToRValueBitCast"); + return {}; + } + + case CK_BitCast: + case CK_BaseToDerived: + case CK_DerivedToBase: + case CK_UncheckedDerivedToBase: + case CK_Dynamic: + case CK_ToUnion: + case CK_ArrayToPointerDecay: + case CK_FunctionToPointerDecay: + case CK_NullToPointer: + case CK_NullToMemberPointer: + case CK_BaseToDerivedMemberPointer: + case CK_DerivedToBaseMemberPointer: + case CK_MemberPointerToBoolean: + case CK_ReinterpretMemberPointer: + case CK_ConstructorConversion: + case CK_IntegralToPointer: + case CK_PointerToIntegral: + case CK_PointerToBoolean: + case CK_ToVoid: + case CK_VectorSplat: + case CK_IntegralCast: + case CK_BooleanToSignedIntegral: + case CK_IntegralToBoolean: + case CK_IntegralToFloating: + case CK_FloatingToIntegral: + case CK_FloatingToBoolean: + case CK_FloatingCast: + case CK_CPointerToObjCPointerCast: + case CK_BlockPointerToObjCPointerCast: + case CK_AnyPointerToBlockPointerCast: + case CK_ObjCObjectLValueCast: + case CK_FloatingComplexToReal: + case CK_FloatingComplexToBoolean: + case CK_IntegralComplexToReal: + case CK_IntegralComplexToBoolean: + case CK_ARCProduceObject: + case CK_ARCConsumeObject: + case CK_ARCReclaimReturnedObject: + case CK_ARCExtendBlockObject: + case CK_CopyAndAutoreleaseBlockObject: + case CK_BuiltinFnToFnPtr: + case CK_ZeroToOCLOpaqueType: + case CK_AddressSpaceConversion: + case CK_IntToOCLSampler: + case CK_FloatingToFixedPoint: + case CK_FixedPointToFloating: + case CK_FixedPointCast: + case CK_FixedPointToBoolean: + case CK_FixedPointToIntegral: + case CK_IntegralToFixedPoint: + case CK_MatrixCast: + case CK_HLSLVectorTruncation: + case CK_HLSLArrayRValue: + case CK_HLSLElementwiseCast: + case CK_HLSLAggregateSplatCast: + llvm_unreachable("invalid cast kind for complex value"); + + case CK_FloatingRealToComplex: + case CK_IntegralRealToComplex: { + assert(!cir::MissingFeatures::cgFPOptionsRAII()); + return emitScalarToComplexCast(cgf.emitScalarExpr(op), op->getType(), + destTy, op->getExprLoc()); + } + + case CK_FloatingComplexCast: + case CK_FloatingComplexToIntegralComplex: + case CK_IntegralComplexCast: + case CK_IntegralComplexToFloatingComplex: { + assert(!cir::MissingFeatures::cgFPOptionsRAII()); + return emitComplexToComplexCast(Visit(op), op->getType(), destTy, + op->getExprLoc()); + } + } + + llvm_unreachable("unknown cast resulting in complex value"); } mlir::Value ComplexExprEmitter::emitConstant( @@ -207,6 +308,49 @@ void ComplexExprEmitter::emitStoreOfComplex(mlir::Location loc, mlir::Value val, builder.createStore(loc, val, destAddr); } +mlir::Value ComplexExprEmitter::emitComplexToComplexCast(mlir::Value val, + QualType srcType, + QualType destType, + SourceLocation loc) { + if (srcType == destType) + return val; + + // Get the src/dest element type. + QualType srcElemTy = srcType->castAs<ComplexType>()->getElementType(); + QualType destElemTy = destType->castAs<ComplexType>()->getElementType(); + + cir::CastKind castOpKind; + if (srcElemTy->isFloatingType() && destElemTy->isFloatingType()) + castOpKind = cir::CastKind::float_complex; + else if (srcElemTy->isFloatingType() && destElemTy->isIntegerType()) + castOpKind = cir::CastKind::float_complex_to_int_complex; + else if (srcElemTy->isIntegerType() && destElemTy->isFloatingType()) + castOpKind = cir::CastKind::int_complex_to_float_complex; + else if (srcElemTy->isIntegerType() && destElemTy->isIntegerType()) + castOpKind = cir::CastKind::int_complex; + else + llvm_unreachable("unexpected src type or dest type"); + + return builder.createCast(cgf.getLoc(loc), castOpKind, val, + cgf.convertType(destType)); +} + +mlir::Value ComplexExprEmitter::emitScalarToComplexCast(mlir::Value val, + QualType srcType, + QualType destType, + SourceLocation loc) { + cir::CastKind castOpKind; + if (srcType->isFloatingType()) + castOpKind = cir::CastKind::float_to_complex; + else if (srcType->isIntegerType()) + castOpKind = cir::CastKind::int_to_complex; + else + llvm_unreachable("unexpected src type"); + + return builder.createCast(cgf.getLoc(loc), castOpKind, val, + cgf.convertType(destType)); +} + mlir::Value ComplexExprEmitter::VisitAbstractConditionalOperator( const AbstractConditionalOperator *e) { mlir::Value condValue = Visit(e->getCond()); diff --git a/clang/lib/CIR/CodeGen/CIRGenExprScalar.cpp b/clang/lib/CIR/CodeGen/CIRGenExprScalar.cpp index eba6bff..2523b0f 100644 --- a/clang/lib/CIR/CodeGen/CIRGenExprScalar.cpp +++ b/clang/lib/CIR/CodeGen/CIRGenExprScalar.cpp @@ -88,6 +88,10 @@ public: // Utilities //===--------------------------------------------------------------------===// + mlir::Value emitComplexToScalarConversion(mlir::Location loc, + mlir::Value value, CastKind kind, + QualType destTy); + mlir::Value emitPromotedValue(mlir::Value result, QualType promotionType) { return builder.createFloatingCast(result, cgf.convertType(promotionType)); } @@ -1125,7 +1129,7 @@ LValue ScalarExprEmitter::emitCompoundAssignLValue( // 'An assignment expression has the value of the left operand after the // assignment...'. if (lhsLV.isBitField()) - cgf.cgm.errorNYI(e->getSourceRange(), "store through bitfield lvalue"); + cgf.emitStoreThroughBitfieldLValue(RValue::get(result), lhsLV); else cgf.emitStoreThroughLValue(RValue::get(result), lhsLV); @@ -1135,6 +1139,31 @@ LValue ScalarExprEmitter::emitCompoundAssignLValue( return lhsLV; } +mlir::Value ScalarExprEmitter::emitComplexToScalarConversion(mlir::Location lov, + mlir::Value value, + CastKind kind, + QualType destTy) { + cir::CastKind castOpKind; + switch (kind) { + case CK_FloatingComplexToReal: + castOpKind = cir::CastKind::float_complex_to_real; + break; + case CK_IntegralComplexToReal: + castOpKind = cir::CastKind::int_complex_to_real; + break; + case CK_FloatingComplexToBoolean: + castOpKind = cir::CastKind::float_complex_to_bool; + break; + case CK_IntegralComplexToBoolean: + castOpKind = cir::CastKind::int_complex_to_bool; + break; + default: + llvm_unreachable("invalid complex-to-scalar cast kind"); + } + + return builder.createCast(lov, castOpKind, value, cgf.convertType(destTy)); +} + mlir::Value ScalarExprEmitter::emitPromoted(const Expr *e, QualType promotionType) { e = e->IgnoreParens(); @@ -1758,6 +1787,15 @@ mlir::Value ScalarExprEmitter::VisitCastExpr(CastExpr *ce) { ce->getExprLoc(), opts); } + case CK_FloatingComplexToReal: + case CK_IntegralComplexToReal: + case CK_FloatingComplexToBoolean: + case CK_IntegralComplexToBoolean: { + mlir::Value value = cgf.emitComplexExpr(subExpr); + return emitComplexToScalarConversion(cgf.getLoc(ce->getExprLoc()), value, + kind, destTy); + } + case CK_FloatingRealToComplex: case CK_FloatingComplexCast: case CK_IntegralRealToComplex: diff --git a/clang/lib/CIR/Dialect/IR/CIRDialect.cpp b/clang/lib/CIR/Dialect/IR/CIRDialect.cpp index f0416b6..cd77166 100644 --- a/clang/lib/CIR/Dialect/IR/CIRDialect.cpp +++ b/clang/lib/CIR/Dialect/IR/CIRDialect.cpp @@ -489,6 +489,104 @@ LogicalResult cir::CastOp::verify() { return emitOpError() << "requires two types differ in addrspace only"; return success(); } + case cir::CastKind::float_to_complex: { + if (!mlir::isa<cir::FPTypeInterface>(srcType)) + return emitOpError() << "requires !cir.float type for source"; + auto resComplexTy = mlir::dyn_cast<cir::ComplexType>(resType); + if (!resComplexTy) + return emitOpError() << "requires !cir.complex type for result"; + if (srcType != resComplexTy.getElementType()) + return emitOpError() << "requires source type match result element type"; + return success(); + } + case cir::CastKind::int_to_complex: { + if (!mlir::isa<cir::IntType>(srcType)) + return emitOpError() << "requires !cir.int type for source"; + auto resComplexTy = mlir::dyn_cast<cir::ComplexType>(resType); + if (!resComplexTy) + return emitOpError() << "requires !cir.complex type for result"; + if (srcType != resComplexTy.getElementType()) + return emitOpError() << "requires source type match result element type"; + return success(); + } + case cir::CastKind::float_complex_to_real: { + auto srcComplexTy = mlir::dyn_cast<cir::ComplexType>(srcType); + if (!srcComplexTy) + return emitOpError() << "requires !cir.complex type for source"; + if (!mlir::isa<cir::FPTypeInterface>(resType)) + return emitOpError() << "requires !cir.float type for result"; + if (srcComplexTy.getElementType() != resType) + return emitOpError() << "requires source element type match result type"; + return success(); + } + case cir::CastKind::int_complex_to_real: { + auto srcComplexTy = mlir::dyn_cast<cir::ComplexType>(srcType); + if (!srcComplexTy) + return emitOpError() << "requires !cir.complex type for source"; + if (!mlir::isa<cir::IntType>(resType)) + return emitOpError() << "requires !cir.int type for result"; + if (srcComplexTy.getElementType() != resType) + return emitOpError() << "requires source element type match result type"; + return success(); + } + case cir::CastKind::float_complex_to_bool: { + auto srcComplexTy = mlir::dyn_cast<cir::ComplexType>(srcType); + if (!srcComplexTy || !srcComplexTy.isFloatingPointComplex()) + return emitOpError() + << "requires floating point !cir.complex type for source"; + if (!mlir::isa<cir::BoolType>(resType)) + return emitOpError() << "requires !cir.bool type for result"; + return success(); + } + case cir::CastKind::int_complex_to_bool: { + auto srcComplexTy = mlir::dyn_cast<cir::ComplexType>(srcType); + if (!srcComplexTy || !srcComplexTy.isIntegerComplex()) + return emitOpError() + << "requires floating point !cir.complex type for source"; + if (!mlir::isa<cir::BoolType>(resType)) + return emitOpError() << "requires !cir.bool type for result"; + return success(); + } + case cir::CastKind::float_complex: { + auto srcComplexTy = mlir::dyn_cast<cir::ComplexType>(srcType); + if (!srcComplexTy || !srcComplexTy.isFloatingPointComplex()) + return emitOpError() + << "requires floating point !cir.complex type for source"; + auto resComplexTy = mlir::dyn_cast<cir::ComplexType>(resType); + if (!resComplexTy || !resComplexTy.isFloatingPointComplex()) + return emitOpError() + << "requires floating point !cir.complex type for result"; + return success(); + } + case cir::CastKind::float_complex_to_int_complex: { + auto srcComplexTy = mlir::dyn_cast<cir::ComplexType>(srcType); + if (!srcComplexTy || !srcComplexTy.isFloatingPointComplex()) + return emitOpError() + << "requires floating point !cir.complex type for source"; + auto resComplexTy = mlir::dyn_cast<cir::ComplexType>(resType); + if (!resComplexTy || !resComplexTy.isIntegerComplex()) + return emitOpError() << "requires integer !cir.complex type for result"; + return success(); + } + case cir::CastKind::int_complex: { + auto srcComplexTy = mlir::dyn_cast<cir::ComplexType>(srcType); + if (!srcComplexTy || !srcComplexTy.isIntegerComplex()) + return emitOpError() << "requires integer !cir.complex type for source"; + auto resComplexTy = mlir::dyn_cast<cir::ComplexType>(resType); + if (!resComplexTy || !resComplexTy.isIntegerComplex()) + return emitOpError() << "requires integer !cir.complex type for result"; + return success(); + } + case cir::CastKind::int_complex_to_float_complex: { + auto srcComplexTy = mlir::dyn_cast<cir::ComplexType>(srcType); + if (!srcComplexTy || !srcComplexTy.isIntegerComplex()) + return emitOpError() << "requires integer !cir.complex type for source"; + auto resComplexTy = mlir::dyn_cast<cir::ComplexType>(resType); + if (!resComplexTy || !resComplexTy.isFloatingPointComplex()) + return emitOpError() + << "requires floating point !cir.complex type for result"; + return success(); + } default: llvm_unreachable("Unknown CastOp kind?"); } diff --git a/clang/lib/CIR/Dialect/Transforms/LoweringPrepare.cpp b/clang/lib/CIR/Dialect/Transforms/LoweringPrepare.cpp index 8f848c7..0f13417 100644 --- a/clang/lib/CIR/Dialect/Transforms/LoweringPrepare.cpp +++ b/clang/lib/CIR/Dialect/Transforms/LoweringPrepare.cpp @@ -13,6 +13,7 @@ #include "clang/CIR/Dialect/IR/CIROpsEnums.h" #include "clang/CIR/Dialect/Passes.h" +#include <iostream> #include <memory> using namespace mlir; @@ -24,11 +25,97 @@ struct LoweringPreparePass : public LoweringPrepareBase<LoweringPreparePass> { void runOnOperation() override; void runOnOp(mlir::Operation *op); + void lowerCastOp(cir::CastOp op); void lowerUnaryOp(cir::UnaryOp op); }; } // namespace +static mlir::Value lowerScalarToComplexCast(mlir::MLIRContext &ctx, + cir::CastOp op) { + cir::CIRBaseBuilderTy builder(ctx); + builder.setInsertionPoint(op); + + mlir::Value src = op.getSrc(); + mlir::Value imag = builder.getNullValue(src.getType(), op.getLoc()); + return builder.createComplexCreate(op.getLoc(), src, imag); +} + +static mlir::Value lowerComplexToScalarCast(mlir::MLIRContext &ctx, + cir::CastOp op, + cir::CastKind elemToBoolKind) { + cir::CIRBaseBuilderTy builder(ctx); + builder.setInsertionPoint(op); + + mlir::Value src = op.getSrc(); + if (!mlir::isa<cir::BoolType>(op.getType())) + return builder.createComplexReal(op.getLoc(), src); + + // Complex cast to bool: (bool)(a+bi) => (bool)a || (bool)b + mlir::Value srcReal = builder.createComplexReal(op.getLoc(), src); + mlir::Value srcImag = builder.createComplexImag(op.getLoc(), src); + + cir::BoolType boolTy = builder.getBoolTy(); + mlir::Value srcRealToBool = + builder.createCast(op.getLoc(), elemToBoolKind, srcReal, boolTy); + mlir::Value srcImagToBool = + builder.createCast(op.getLoc(), elemToBoolKind, srcImag, boolTy); + return builder.createLogicalOr(op.getLoc(), srcRealToBool, srcImagToBool); +} + +static mlir::Value lowerComplexToComplexCast(mlir::MLIRContext &ctx, + cir::CastOp op, + cir::CastKind scalarCastKind) { + CIRBaseBuilderTy builder(ctx); + builder.setInsertionPoint(op); + + mlir::Value src = op.getSrc(); + auto dstComplexElemTy = + mlir::cast<cir::ComplexType>(op.getType()).getElementType(); + + mlir::Value srcReal = builder.createComplexReal(op.getLoc(), src); + mlir::Value srcImag = builder.createComplexImag(op.getLoc(), src); + + mlir::Value dstReal = builder.createCast(op.getLoc(), scalarCastKind, srcReal, + dstComplexElemTy); + mlir::Value dstImag = builder.createCast(op.getLoc(), scalarCastKind, srcImag, + dstComplexElemTy); + return builder.createComplexCreate(op.getLoc(), dstReal, dstImag); +} + +void LoweringPreparePass::lowerCastOp(cir::CastOp op) { + mlir::MLIRContext &ctx = getContext(); + mlir::Value loweredValue = [&]() -> mlir::Value { + switch (op.getKind()) { + case cir::CastKind::float_to_complex: + case cir::CastKind::int_to_complex: + return lowerScalarToComplexCast(ctx, op); + case cir::CastKind::float_complex_to_real: + case cir::CastKind::int_complex_to_real: + return lowerComplexToScalarCast(ctx, op, op.getKind()); + case cir::CastKind::float_complex_to_bool: + return lowerComplexToScalarCast(ctx, op, cir::CastKind::float_to_bool); + case cir::CastKind::int_complex_to_bool: + return lowerComplexToScalarCast(ctx, op, cir::CastKind::int_to_bool); + case cir::CastKind::float_complex: + return lowerComplexToComplexCast(ctx, op, cir::CastKind::floating); + case cir::CastKind::float_complex_to_int_complex: + return lowerComplexToComplexCast(ctx, op, cir::CastKind::float_to_int); + case cir::CastKind::int_complex: + return lowerComplexToComplexCast(ctx, op, cir::CastKind::integral); + case cir::CastKind::int_complex_to_float_complex: + return lowerComplexToComplexCast(ctx, op, cir::CastKind::int_to_float); + default: + return nullptr; + } + }(); + + if (loweredValue) { + op.replaceAllUsesWith(loweredValue); + op.erase(); + } +} + void LoweringPreparePass::lowerUnaryOp(cir::UnaryOp op) { mlir::Type ty = op.getType(); if (!mlir::isa<cir::ComplexType>(ty)) @@ -72,7 +159,9 @@ void LoweringPreparePass::lowerUnaryOp(cir::UnaryOp op) { } void LoweringPreparePass::runOnOp(mlir::Operation *op) { - if (auto unary = dyn_cast<cir::UnaryOp>(op)) + if (auto cast = mlir::dyn_cast<cir::CastOp>(op)) + lowerCastOp(cast); + else if (auto unary = mlir::dyn_cast<cir::UnaryOp>(op)) lowerUnaryOp(unary); } @@ -82,7 +171,7 @@ void LoweringPreparePass::runOnOperation() { llvm::SmallVector<mlir::Operation *> opsToTransform; op->walk([&](mlir::Operation *op) { - if (mlir::isa<cir::UnaryOp>(op)) + if (mlir::isa<cir::CastOp, cir::UnaryOp>(op)) opsToTransform.push_back(op); }); diff --git a/clang/lib/CodeGen/ABIInfo.cpp b/clang/lib/CodeGen/ABIInfo.cpp index d981d69..3ef430e1 100644 --- a/clang/lib/CodeGen/ABIInfo.cpp +++ b/clang/lib/CodeGen/ABIInfo.cpp @@ -218,8 +218,8 @@ void ABIInfo::appendAttributeMangling(StringRef AttrStr, // only have "+" prefixes here. assert(LHS.starts_with("+") && RHS.starts_with("+") && "Features should always have a prefix."); - return TI.getFMVPriority({LHS.substr(1)}) > - TI.getFMVPriority({RHS.substr(1)}); + return TI.getFMVPriority({LHS.substr(1)}) + .ugt(TI.getFMVPriority({RHS.substr(1)})); }); bool IsFirst = true; diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp index 5f2eb76..3f784fc 100644 --- a/clang/lib/CodeGen/CGBuiltin.cpp +++ b/clang/lib/CodeGen/CGBuiltin.cpp @@ -4108,6 +4108,22 @@ RValue CodeGenFunction::EmitBuiltinExpr(const GlobalDecl GD, unsigned BuiltinID, return RValue::get(Result); } + case Builtin::BI__builtin_elementwise_maximumnum: { + Value *Op0 = EmitScalarExpr(E->getArg(0)); + Value *Op1 = EmitScalarExpr(E->getArg(1)); + Value *Result = Builder.CreateBinaryIntrinsic( + Intrinsic::maximumnum, Op0, Op1, nullptr, "elt.maximumnum"); + return RValue::get(Result); + } + + case Builtin::BI__builtin_elementwise_minimumnum: { + Value *Op0 = EmitScalarExpr(E->getArg(0)); + Value *Op1 = EmitScalarExpr(E->getArg(1)); + Value *Result = Builder.CreateBinaryIntrinsic( + Intrinsic::minimumnum, Op0, Op1, nullptr, "elt.minimumnum"); + return RValue::get(Result); + } + case Builtin::BI__builtin_reduce_max: { auto GetIntrinsicID = [this](QualType QT) { if (auto *VecTy = QT->getAs<VectorType>()) diff --git a/clang/lib/CodeGen/CGStmt.cpp b/clang/lib/CodeGen/CGStmt.cpp index 26259a8..1a8c6f0 100644 --- a/clang/lib/CodeGen/CGStmt.cpp +++ b/clang/lib/CodeGen/CGStmt.cpp @@ -2676,6 +2676,9 @@ static void UpdateAsmCallInst(llvm::CallBase &Result, bool HasSideEffect, llvm::ConstantAsMetadata::get(Loc))); } + // Make inline-asm calls Key for the debug info feature Key Instructions. + CGF.addInstToNewSourceAtom(&Result, nullptr); + if (!NoConvergent && CGF.getLangOpts().assumeFunctionsAreConvergent()) // Conservatively, mark all inline asm blocks in CUDA or OpenCL as // convergent (meaning, they may call an intrinsically convergent op, such @@ -2754,6 +2757,7 @@ EmitAsmStores(CodeGenFunction &CGF, const AsmStmt &S, } } + ApplyAtomGroup Grp(CGF.getDebugInfo()); LValue Dest = ResultRegDests[i]; // ResultTypeRequiresCast elements correspond to the first // ResultTypeRequiresCast.size() elements of RegResults. @@ -2761,7 +2765,8 @@ EmitAsmStores(CodeGenFunction &CGF, const AsmStmt &S, unsigned Size = CGF.getContext().getTypeSize(ResultRegQualTys[i]); Address A = Dest.getAddress().withElementType(ResultRegTypes[i]); if (CGF.getTargetHooks().isScalarizableAsmOperand(CGF, TruncTy)) { - Builder.CreateStore(Tmp, A); + llvm::StoreInst *S = Builder.CreateStore(Tmp, A); + CGF.addInstToCurrentSourceAtom(S, S->getValueOperand()); continue; } diff --git a/clang/lib/CodeGen/CodeGenModule.cpp b/clang/lib/CodeGen/CodeGenModule.cpp index 236cc3d..834b1c0 100644 --- a/clang/lib/CodeGen/CodeGenModule.cpp +++ b/clang/lib/CodeGen/CodeGenModule.cpp @@ -4418,8 +4418,9 @@ void CodeGenModule::EmitGlobalDefinition(GlobalDecl GD, llvm::GlobalValue *GV) { static void ReplaceUsesOfNonProtoTypeWithRealFunction(llvm::GlobalValue *Old, llvm::Function *NewFn); -static uint64_t getFMVPriority(const TargetInfo &TI, - const CodeGenFunction::FMVResolverOption &RO) { +static llvm::APInt +getFMVPriority(const TargetInfo &TI, + const CodeGenFunction::FMVResolverOption &RO) { llvm::SmallVector<StringRef, 8> Features{RO.Features}; if (RO.Architecture) Features.push_back(*RO.Architecture); @@ -4544,7 +4545,7 @@ void CodeGenModule::emitMultiVersionFunctions() { llvm::stable_sort( Options, [&TI](const CodeGenFunction::FMVResolverOption &LHS, const CodeGenFunction::FMVResolverOption &RHS) { - return getFMVPriority(TI, LHS) > getFMVPriority(TI, RHS); + return getFMVPriority(TI, LHS).ugt(getFMVPriority(TI, RHS)); }); CodeGenFunction CGF(*this); CGF.EmitMultiVersionResolver(ResolverFunc, Options); diff --git a/clang/lib/CodeGen/TargetBuiltins/ARM.cpp b/clang/lib/CodeGen/TargetBuiltins/ARM.cpp index 7e6a47f..2e6b4b3 100644 --- a/clang/lib/CodeGen/TargetBuiltins/ARM.cpp +++ b/clang/lib/CodeGen/TargetBuiltins/ARM.cpp @@ -8112,7 +8112,7 @@ Value *CodeGenFunction::EmitAArch64CpuSupports(const CallExpr *E) { llvm::Value * CodeGenFunction::EmitAArch64CpuSupports(ArrayRef<StringRef> FeaturesStrs) { - uint64_t FeaturesMask = llvm::AArch64::getCpuSupportsMask(FeaturesStrs); + llvm::APInt FeaturesMask = llvm::AArch64::getCpuSupportsMask(FeaturesStrs); Value *Result = Builder.getTrue(); if (FeaturesMask != 0) { // Get features from structure in runtime library @@ -8128,7 +8128,7 @@ CodeGenFunction::EmitAArch64CpuSupports(ArrayRef<StringRef> FeaturesStrs) { {ConstantInt::get(Int32Ty, 0), ConstantInt::get(Int32Ty, 0)}); Value *Features = Builder.CreateAlignedLoad(Int64Ty, CpuFeatures, CharUnits::fromQuantity(8)); - Value *Mask = Builder.getInt64(FeaturesMask); + Value *Mask = Builder.getInt(FeaturesMask.trunc(64)); Value *Bitset = Builder.CreateAnd(Features, Mask); Value *Cmp = Builder.CreateICmpEQ(Bitset, Mask); Result = Builder.CreateAnd(Result, Cmp); diff --git a/clang/lib/Driver/Driver.cpp b/clang/lib/Driver/Driver.cpp index eeb4823..892049e 100644 --- a/clang/lib/Driver/Driver.cpp +++ b/clang/lib/Driver/Driver.cpp @@ -68,6 +68,7 @@ #include "clang/Driver/Types.h" #include "llvm/ADT/ArrayRef.h" #include "llvm/ADT/STLExtras.h" +#include "llvm/ADT/SmallSet.h" #include "llvm/ADT/StringExtras.h" #include "llvm/ADT/StringRef.h" #include "llvm/ADT/StringSet.h" @@ -83,6 +84,7 @@ #include "llvm/Support/ErrorHandling.h" #include "llvm/Support/ExitCodes.h" #include "llvm/Support/FileSystem.h" +#include "llvm/Support/FileUtilities.h" #include "llvm/Support/FormatVariadic.h" #include "llvm/Support/MD5.h" #include "llvm/Support/Path.h" @@ -109,65 +111,6 @@ using namespace clang::driver; using namespace clang; using namespace llvm::opt; -static std::optional<llvm::Triple> getOffloadTargetTriple(const Driver &D, - const ArgList &Args) { - auto OffloadTargets = Args.getAllArgValues(options::OPT_offload_EQ); - // Offload compilation flow does not support multiple targets for now. We - // need the HIPActionBuilder (and possibly the CudaActionBuilder{,Base}too) - // to support multiple tool chains first. - switch (OffloadTargets.size()) { - default: - D.Diag(diag::err_drv_only_one_offload_target_supported); - return std::nullopt; - case 0: - D.Diag(diag::err_drv_invalid_or_unsupported_offload_target) << ""; - return std::nullopt; - case 1: - break; - } - return llvm::Triple(OffloadTargets[0]); -} - -static std::optional<llvm::Triple> -getNVIDIAOffloadTargetTriple(const Driver &D, const ArgList &Args, - const llvm::Triple &HostTriple) { - if (!Args.hasArg(options::OPT_offload_EQ)) { - return llvm::Triple(HostTriple.isArch64Bit() ? "nvptx64-nvidia-cuda" - : "nvptx-nvidia-cuda"); - } - auto TT = getOffloadTargetTriple(D, Args); - if (TT && (TT->getArch() == llvm::Triple::spirv32 || - TT->getArch() == llvm::Triple::spirv64)) { - if (Args.hasArg(options::OPT_emit_llvm)) - return TT; - D.Diag(diag::err_drv_cuda_offload_only_emit_bc); - return std::nullopt; - } - D.Diag(diag::err_drv_invalid_or_unsupported_offload_target) << TT->str(); - return std::nullopt; -} - -static std::optional<llvm::Triple> -getHIPOffloadTargetTriple(const Driver &D, const ArgList &Args) { - if (!Args.hasArg(options::OPT_offload_EQ)) { - auto OffloadArchs = Args.getAllArgValues(options::OPT_offload_arch_EQ); - if (llvm::is_contained(OffloadArchs, "amdgcnspirv") && - OffloadArchs.size() == 1) - return llvm::Triple("spirv64-amd-amdhsa"); - return llvm::Triple("amdgcn-amd-amdhsa"); // Default HIP triple. - } - auto TT = getOffloadTargetTriple(D, Args); - if (!TT) - return std::nullopt; - if (TT->isAMDGCN() && TT->getVendor() == llvm::Triple::AMD && - TT->getOS() == llvm::Triple::AMDHSA) - return TT; - if (TT->getArch() == llvm::Triple::spirv64) - return TT; - D.Diag(diag::err_drv_invalid_or_unsupported_offload_target) << TT->str(); - return std::nullopt; -} - template <typename F> static bool usesInput(const ArgList &Args, F &&Fn) { return llvm::any_of(Args, [&](Arg *A) { return (A->getOption().matches(options::OPT_x) && @@ -458,6 +401,44 @@ phases::ID Driver::getFinalPhase(const DerivedArgList &DAL, return FinalPhase; } +llvm::Expected<std::unique_ptr<llvm::MemoryBuffer>> +Driver::executeProgram(llvm::ArrayRef<llvm::StringRef> Args) const { + llvm::SmallString<64> OutputFile; + llvm::sys::fs::createTemporaryFile("driver-program", "txt", OutputFile, + llvm::sys::fs::OF_Text); + llvm::FileRemover OutputRemover(OutputFile.c_str()); + std::optional<llvm::StringRef> Redirects[] = { + {""}, + OutputFile.str(), + {""}, + }; + + std::string ErrorMessage; + int SecondsToWait = 60; + if (std::optional<std::string> Str = + llvm::sys::Process::GetEnv("CLANG_TOOLCHAIN_PROGRAM_TIMEOUT")) { + if (!llvm::to_integer(*Str, SecondsToWait)) + return llvm::createStringError(std::error_code(), + "CLANG_TOOLCHAIN_PROGRAM_TIMEOUT expected " + "an integer, got '" + + *Str + "'"); + SecondsToWait = std::max(SecondsToWait, 0); // infinite + } + StringRef Executable = Args[0]; + if (llvm::sys::ExecuteAndWait(Executable, Args, {}, Redirects, SecondsToWait, + /*MemoryLimit=*/0, &ErrorMessage)) + return llvm::createStringError(std::error_code(), + Executable + ": " + ErrorMessage); + + llvm::ErrorOr<std::unique_ptr<llvm::MemoryBuffer>> OutputBuf = + llvm::MemoryBuffer::getFile(OutputFile.c_str()); + if (!OutputBuf) + return llvm::createStringError(OutputBuf.getError(), + "Failed to read stdout of " + Executable + + ": " + OutputBuf.getError().message()); + return std::move(*OutputBuf); +} + static Arg *MakeInputArg(DerivedArgList &Args, const OptTable &Opts, StringRef Value, bool Claim = true) { Arg *A = new Arg(Opts.getOption(options::OPT_INPUT), Value, @@ -921,250 +902,266 @@ Driver::OpenMPRuntimeKind Driver::getOpenMPRuntime(const ArgList &Args) const { return RT; } -static llvm::Triple getSYCLDeviceTriple(StringRef TargetArch) { - SmallVector<StringRef, 5> SYCLAlias = {"spir", "spir64", "spirv", "spirv32", - "spirv64"}; - if (llvm::is_contained(SYCLAlias, TargetArch)) { - llvm::Triple TargetTriple; - TargetTriple.setArchName(TargetArch); - TargetTriple.setVendor(llvm::Triple::UnknownVendor); - TargetTriple.setOS(llvm::Triple::UnknownOS); - return TargetTriple; - } - return llvm::Triple(TargetArch); +// Handles `native` offload architectures by using the 'offload-arch' utility. +static llvm::SmallVector<std::string> +getSystemOffloadArchs(Compilation &C, Action::OffloadKind Kind) { + StringRef Program = C.getArgs().getLastArgValue( + options::OPT_offload_arch_tool_EQ, "offload-arch"); + + SmallVector<std::string> GPUArchs; + if (llvm::ErrorOr<std::string> Executable = + llvm::sys::findProgramByName(Program)) { + llvm::SmallVector<StringRef> Args{*Executable}; + if (Kind == Action::OFK_HIP) + Args.push_back("--only=amdgpu"); + else if (Kind == Action::OFK_Cuda) + Args.push_back("--only=nvptx"); + auto StdoutOrErr = C.getDriver().executeProgram(Args); + + if (!StdoutOrErr) { + C.getDriver().Diag(diag::err_drv_undetermined_gpu_arch) + << Action::GetOffloadKindName(Kind) << StdoutOrErr.takeError() + << "--offload-arch"; + return GPUArchs; + } + if ((*StdoutOrErr)->getBuffer().empty()) { + C.getDriver().Diag(diag::err_drv_undetermined_gpu_arch) + << Action::GetOffloadKindName(Kind) << "No GPU detected in the system" + << "--offload-arch"; + return GPUArchs; + } + + for (StringRef Arch : llvm::split((*StdoutOrErr)->getBuffer(), "\n")) + if (!Arch.empty()) + GPUArchs.push_back(Arch.str()); + } else { + C.getDriver().Diag(diag::err_drv_command_failure) << "offload-arch"; + } + return GPUArchs; } -static bool addSYCLDefaultTriple(Compilation &C, - SmallVectorImpl<llvm::Triple> &SYCLTriples) { - // Check current set of triples to see if the default has already been set. - for (const auto &SYCLTriple : SYCLTriples) { - if (SYCLTriple.getSubArch() == llvm::Triple::NoSubArch && - SYCLTriple.isSPIROrSPIRV()) - return false; +// Attempts to infer the correct offloading toolchain triple by looking at the +// requested offloading kind and architectures. +static llvm::DenseSet<llvm::StringRef> +inferOffloadToolchains(Compilation &C, Action::OffloadKind Kind) { + std::set<std::string> Archs; + for (Arg *A : C.getInputArgs()) { + for (StringRef Arch : A->getValues()) { + if (A->getOption().matches(options::OPT_offload_arch_EQ)) { + if (Arch == "native") { + for (StringRef Str : getSystemOffloadArchs(C, Kind)) + Archs.insert(Str.str()); + } else { + Archs.insert(Arch.str()); + } + } else if (A->getOption().matches(options::OPT_no_offload_arch_EQ)) { + if (Arch == "all") + Archs.clear(); + else + Archs.erase(Arch.str()); + } + } } - // Add the default triple as it was not found. - llvm::Triple DefaultTriple = getSYCLDeviceTriple( - C.getDefaultToolChain().getTriple().isArch32Bit() ? "spirv32" - : "spirv64"); - SYCLTriples.insert(SYCLTriples.begin(), DefaultTriple); - return true; + + llvm::DenseSet<llvm::StringRef> Triples; + for (llvm::StringRef Arch : Archs) { + OffloadArch ID = StringToOffloadArch(Arch); + if (ID == OffloadArch::UNKNOWN) + ID = StringToOffloadArch( + getProcessorFromTargetID(llvm::Triple("amdgcn-amd-amdhsa"), Arch)); + + if (Kind == Action::OFK_HIP && !IsAMDOffloadArch(ID)) { + C.getDriver().Diag(clang::diag::err_drv_offload_bad_gpu_arch) + << "HIP" << Arch; + return llvm::DenseSet<llvm::StringRef>(); + } + if (Kind == Action::OFK_Cuda && !IsNVIDIAOffloadArch(ID)) { + C.getDriver().Diag(clang::diag::err_drv_offload_bad_gpu_arch) + << "CUDA" << Arch; + return llvm::DenseSet<llvm::StringRef>(); + } + if (Kind == Action::OFK_OpenMP && + (ID == OffloadArch::UNKNOWN || ID == OffloadArch::UNUSED)) { + C.getDriver().Diag(clang::diag::err_drv_failed_to_deduce_target_from_arch) + << Arch; + return llvm::DenseSet<llvm::StringRef>(); + } + if (ID == OffloadArch::UNKNOWN || ID == OffloadArch::UNUSED) { + C.getDriver().Diag(clang::diag::err_drv_offload_bad_gpu_arch) + << "offload" << Arch; + return llvm::DenseSet<llvm::StringRef>(); + } + + StringRef Triple; + if (ID == OffloadArch::AMDGCNSPIRV) + Triple = "spirv64-amd-amdhsa"; + else if (IsNVIDIAOffloadArch(ID)) + Triple = C.getDefaultToolChain().getTriple().isArch64Bit() + ? "nvptx64-nvidia-cuda" + : "nvptx-nvidia-cuda"; + else if (IsAMDOffloadArch(ID)) + Triple = "amdgcn-amd-amdhsa"; + else + continue; + + // Make a new argument that dispatches this argument to the appropriate + // toolchain. This is required when we infer it and create potentially + // incompatible toolchains from the global option. + Option Opt = C.getDriver().getOpts().getOption(options::OPT_Xarch__); + unsigned Index = C.getArgs().getBaseArgs().MakeIndex("-Xarch_"); + Arg *A = new Arg(Opt, C.getArgs().getArgString(Index), Index, + C.getArgs().MakeArgString(Triple.split("-").first), + C.getArgs().MakeArgString("--offload-arch=" + Arch)); + C.getArgs().append(A); + C.getArgs().AddSynthesizedArg(A); + Triples.insert(Triple); + } + + // Infer the default target triple if no specific architectures are given. + if (Archs.empty() && Kind == Action::OFK_HIP) + Triples.insert("amdgcn-amd-amdhsa"); + else if (Archs.empty() && Kind == Action::OFK_Cuda) + Triples.insert(C.getDefaultToolChain().getTriple().isArch64Bit() + ? "nvptx64-nvidia-cuda" + : "nvptx-nvidia-cuda"); + else if (Archs.empty() && Kind == Action::OFK_SYCL) + Triples.insert(C.getDefaultToolChain().getTriple().isArch64Bit() + ? "spirv64-unknown-unknown" + : "spirv32-unknown-unknown"); + + // We need to dispatch these to the appropriate toolchain now. + C.getArgs().eraseArg(options::OPT_offload_arch_EQ); + C.getArgs().eraseArg(options::OPT_no_offload_arch_EQ); + + return Triples; } void Driver::CreateOffloadingDeviceToolChains(Compilation &C, InputList &Inputs) { - - // - // CUDA/HIP - // - // We need to generate a CUDA/HIP toolchain if any of the inputs has a CUDA - // or HIP type. However, mixed CUDA/HIP compilation is not supported. + bool UseLLVMOffload = C.getInputArgs().hasArg( + options::OPT_foffload_via_llvm, options::OPT_fno_offload_via_llvm, false); bool IsCuda = - llvm::any_of(Inputs, [](std::pair<types::ID, const llvm::opt::Arg *> &I) { - return types::isCuda(I.first); - }); - bool IsHIP = llvm::any_of(Inputs, [](std::pair<types::ID, const llvm::opt::Arg *> &I) { - return types::isHIP(I.first); - }) || - C.getInputArgs().hasArg(options::OPT_hip_link) || - C.getInputArgs().hasArg(options::OPT_hipstdpar); - bool UseLLVMOffload = C.getInputArgs().hasArg( - options::OPT_foffload_via_llvm, options::OPT_fno_offload_via_llvm, false); - if (IsCuda && IsHIP) { - Diag(clang::diag::err_drv_mix_cuda_hip); + return types::isCuda(I.first); + }) && + !UseLLVMOffload; + bool IsHIP = + (llvm::any_of(Inputs, + [](std::pair<types::ID, const llvm::opt::Arg *> &I) { + return types::isHIP(I.first); + }) || + C.getInputArgs().hasArg(options::OPT_hip_link) || + C.getInputArgs().hasArg(options::OPT_hipstdpar)) && + !UseLLVMOffload; + bool IsSYCL = C.getInputArgs().hasFlag(options::OPT_fsycl, + options::OPT_fno_sycl, false); + bool IsOpenMPOffloading = + UseLLVMOffload || + (C.getInputArgs().hasFlag(options::OPT_fopenmp, options::OPT_fopenmp_EQ, + options::OPT_fno_openmp, false) && + (C.getInputArgs().hasArg(options::OPT_offload_targets_EQ) || + (C.getInputArgs().hasArg(options::OPT_offload_arch_EQ) && + !(IsCuda || IsHIP)))); + + llvm::SmallSet<Action::OffloadKind, 4> Kinds; + const std::pair<bool, Action::OffloadKind> ActiveKinds[] = { + {IsCuda, Action::OFK_Cuda}, + {IsHIP, Action::OFK_HIP}, + {IsOpenMPOffloading, Action::OFK_OpenMP}, + {IsSYCL, Action::OFK_SYCL}}; + for (const auto &[Active, Kind] : ActiveKinds) + if (Active) + Kinds.insert(Kind); + + // We currently don't support any kind of mixed offloading. + if (Kinds.size() > 1) { + Diag(clang::diag::err_drv_mix_offload) + << Action::GetOffloadKindName(*Kinds.begin()).upper() + << Action::GetOffloadKindName(*(++Kinds.begin())).upper(); return; } - if (IsCuda && !UseLLVMOffload) { - auto CudaTriple = getNVIDIAOffloadTargetTriple( - *this, C.getInputArgs(), C.getDefaultToolChain().getTriple()); - if (!CudaTriple) - return; - - auto &TC = - getOffloadToolChain(C.getInputArgs(), Action::OFK_Cuda, *CudaTriple, - C.getDefaultToolChain().getTriple()); - - // Emit a warning if the detected CUDA version is too new. - const CudaInstallationDetector &CudaInstallation = - static_cast<const toolchains::CudaToolChain &>(TC).CudaInstallation; - if (CudaInstallation.isValid()) - CudaInstallation.WarnIfUnsupportedVersion(); - C.addOffloadDeviceToolChain(&TC, Action::OFK_Cuda); - OffloadArchs[&TC] = getOffloadArchs(C, C.getArgs(), Action::OFK_Cuda, &TC, - /*SpecificToolchain=*/true); - } else if (IsHIP && !UseLLVMOffload) { - if (auto *OMPTargetArg = - C.getInputArgs().getLastArg(options::OPT_offload_targets_EQ)) { - Diag(clang::diag::err_drv_unsupported_opt_for_language_mode) - << OMPTargetArg->getSpelling() << "HIP"; - return; - } - - auto HIPTriple = getHIPOffloadTargetTriple(*this, C.getInputArgs()); - if (!HIPTriple) - return; - - auto &TC = - getOffloadToolChain(C.getInputArgs(), Action::OFK_HIP, *HIPTriple, - C.getDefaultToolChain().getTriple()); - C.addOffloadDeviceToolChain(&TC, Action::OFK_HIP); - - // TODO: Fix 'amdgcnspirv' handling with the new driver. - if (C.getInputArgs().hasFlag(options::OPT_offload_new_driver, - options::OPT_no_offload_new_driver, false)) - OffloadArchs[&TC] = getOffloadArchs(C, C.getArgs(), Action::OFK_HIP, &TC, - /*SpecificToolchain=*/true); - } + // Initialize the compilation identifier used for unique CUDA / HIP names. if (IsCuda || IsHIP) CUIDOpts = CUIDOptions(C.getArgs(), *this); - // - // OpenMP - // - // We need to generate an OpenMP toolchain if the user specified targets with - // the -fopenmp-targets option or used --offload-arch with OpenMP enabled. - bool IsOpenMPOffloading = - ((IsCuda || IsHIP) && UseLLVMOffload) || - (C.getInputArgs().hasFlag(options::OPT_fopenmp, options::OPT_fopenmp_EQ, - options::OPT_fno_openmp, false) && - (C.getInputArgs().hasArg(options::OPT_offload_targets_EQ) || - C.getInputArgs().hasArg(options::OPT_offload_arch_EQ))); - if (IsOpenMPOffloading) { - // We expect that -fopenmp-targets is always used in conjunction with the - // option -fopenmp specifying a valid runtime with offloading support, i.e. - // libomp or libiomp. - OpenMPRuntimeKind RuntimeKind = getOpenMPRuntime(C.getInputArgs()); - if (RuntimeKind != OMPRT_OMP && RuntimeKind != OMPRT_IOMP5) { - Diag(clang::diag::err_drv_expecting_fopenmp_with_fopenmp_targets); - return; - } - - // If the user specified -fopenmp-targets= we create a toolchain for each - // valid triple. Otherwise, if only --offload-arch= was specified we instead - // attempt to derive the appropriate toolchains from the arguments. - if (Arg *OpenMPTargets = - C.getInputArgs().getLastArg(options::OPT_offload_targets_EQ)) { - if (OpenMPTargets && !OpenMPTargets->getNumValues()) { - Diag(clang::diag::warn_drv_empty_joined_argument) - << OpenMPTargets->getAsString(C.getInputArgs()); + // Get the list of requested offloading toolchains. If they were not + // explicitly specified we will infer them based on the offloading language + // and requested architectures. + std::multiset<llvm::StringRef> Triples; + if (C.getInputArgs().hasArg(options::OPT_offload_targets_EQ)) { + std::vector<std::string> ArgValues = + C.getInputArgs().getAllArgValues(options::OPT_offload_targets_EQ); + for (llvm::StringRef Target : ArgValues) + Triples.insert(C.getInputArgs().MakeArgString(Target)); + + if (ArgValues.empty()) + Diag(clang::diag::warn_drv_empty_joined_argument) + << C.getInputArgs() + .getLastArg(options::OPT_offload_targets_EQ) + ->getAsString(C.getInputArgs()); + } else if (Kinds.size() > 0) { + for (Action::OffloadKind Kind : Kinds) { + llvm::DenseSet<llvm::StringRef> Derived = inferOffloadToolchains(C, Kind); + Triples.insert(Derived.begin(), Derived.end()); + } + } + + // Build an offloading toolchain for every requested target and kind. + llvm::StringMap<StringRef> FoundNormalizedTriples; + for (StringRef Target : Triples) { + // OpenMP offloading requires a compatible libomp. + if (Kinds.contains(Action::OFK_OpenMP)) { + OpenMPRuntimeKind RuntimeKind = getOpenMPRuntime(C.getInputArgs()); + if (RuntimeKind != OMPRT_OMP && RuntimeKind != OMPRT_IOMP5) { + Diag(clang::diag::err_drv_expecting_fopenmp_with_fopenmp_targets); return; } + } - // Make sure these show up in a deterministic order. - std::multiset<StringRef> OpenMPTriples; - for (StringRef T : OpenMPTargets->getValues()) - OpenMPTriples.insert(T); - - llvm::StringMap<StringRef> FoundNormalizedTriples; - for (StringRef T : OpenMPTriples) { - llvm::Triple TT(ToolChain::getOpenMPTriple(T)); - std::string NormalizedName = TT.normalize(); - - // Make sure we don't have a duplicate triple. - auto [TripleIt, Inserted] = - FoundNormalizedTriples.try_emplace(NormalizedName, T); - if (!Inserted) { - Diag(clang::diag::warn_drv_omp_offload_target_duplicate) - << T << TripleIt->second; - continue; - } - - // If the specified target is invalid, emit a diagnostic. - if (TT.getArch() == llvm::Triple::UnknownArch) { - Diag(clang::diag::err_drv_invalid_omp_target) << T; - continue; - } + // Certain options are not allowed when combined with SYCL compilation. + if (Kinds.contains(Action::OFK_SYCL)) { + for (auto ID : + {options::OPT_static_libstdcxx, options::OPT_ffreestanding}) + if (Arg *IncompatArg = C.getInputArgs().getLastArg(ID)) + Diag(clang::diag::err_drv_argument_not_allowed_with) + << IncompatArg->getSpelling() << "-fsycl"; + } - auto &TC = getOffloadToolChain(C.getInputArgs(), Action::OFK_OpenMP, TT, - C.getDefaultToolChain().getTriple()); - C.addOffloadDeviceToolChain(&TC, Action::OFK_OpenMP); - OffloadArchs[&TC] = - getOffloadArchs(C, C.getArgs(), Action::OFK_OpenMP, &TC, - /*SpecificToolchain=*/true); - } - } else if (C.getInputArgs().hasArg(options::OPT_offload_arch_EQ) && - ((!IsHIP && !IsCuda) || UseLLVMOffload)) { - llvm::Triple AMDTriple("amdgcn-amd-amdhsa"); - llvm::Triple NVPTXTriple("nvptx64-nvidia-cuda"); - - for (StringRef Arch : - C.getInputArgs().getAllArgValues(options::OPT_offload_arch_EQ)) { - bool IsNVPTX = IsNVIDIAOffloadArch( - StringToOffloadArch(getProcessorFromTargetID(NVPTXTriple, Arch))); - bool IsAMDGPU = IsAMDOffloadArch( - StringToOffloadArch(getProcessorFromTargetID(AMDTriple, Arch))); - if (!IsNVPTX && !IsAMDGPU && !Arch.empty() && - !Arch.equals_insensitive("native")) { - Diag(clang::diag::err_drv_failed_to_deduce_target_from_arch) << Arch; - return; - } + // Create a device toolchain for every specified kind and triple. + for (Action::OffloadKind Kind : Kinds) { + llvm::Triple TT = Kind == Action::OFK_OpenMP + ? ToolChain::getOpenMPTriple(Target) + : llvm::Triple(Target); + if (TT.getArch() == llvm::Triple::ArchType::UnknownArch) { + Diag(diag::err_drv_invalid_or_unsupported_offload_target) << TT.str(); + continue; } - // Attempt to deduce the offloading triple from the set of architectures. - // We can only correctly deduce NVPTX / AMDGPU triples currently. - for (const llvm::Triple &TT : {AMDTriple, NVPTXTriple}) { - auto &TC = getOffloadToolChain(C.getInputArgs(), Action::OFK_OpenMP, TT, - C.getDefaultToolChain().getTriple()); - - llvm::SmallVector<StringRef> Archs = - getOffloadArchs(C, C.getArgs(), Action::OFK_OpenMP, &TC, - /*SpecificToolchain=*/false); - if (!Archs.empty()) { - C.addOffloadDeviceToolChain(&TC, Action::OFK_OpenMP); - OffloadArchs[&TC] = Archs; - } + std::string NormalizedName = TT.normalize(); + auto [TripleIt, Inserted] = + FoundNormalizedTriples.try_emplace(NormalizedName, Target); + if (!Inserted) { + Diag(clang::diag::warn_drv_omp_offload_target_duplicate) + << Target << TripleIt->second; + continue; } - // If the set is empty then we failed to find a native architecture. - auto TCRange = C.getOffloadToolChains(Action::OFK_OpenMP); - if (TCRange.first == TCRange.second) - Diag(clang::diag::err_drv_failed_to_deduce_target_from_arch) - << "native"; - } - } else if (C.getInputArgs().hasArg(options::OPT_offload_targets_EQ)) { - Diag(clang::diag::err_drv_expecting_fopenmp_with_fopenmp_targets); - return; - } + auto &TC = getOffloadToolChain(C.getInputArgs(), Kind, TT, + C.getDefaultToolChain().getTriple()); - // We need to generate a SYCL toolchain if the user specified -fsycl. - bool IsSYCL = C.getInputArgs().hasFlag(options::OPT_fsycl, - options::OPT_fno_sycl, false); - - auto argSYCLIncompatible = [&](OptSpecifier OptId) { - if (!IsSYCL) - return; - if (Arg *IncompatArg = C.getInputArgs().getLastArg(OptId)) - Diag(clang::diag::err_drv_argument_not_allowed_with) - << IncompatArg->getSpelling() << "-fsycl"; - }; - // -static-libstdc++ is not compatible with -fsycl. - argSYCLIncompatible(options::OPT_static_libstdcxx); - // -ffreestanding cannot be used with -fsycl - argSYCLIncompatible(options::OPT_ffreestanding); - - llvm::SmallVector<llvm::Triple, 4> UniqueSYCLTriplesVec; - - if (IsSYCL) { - addSYCLDefaultTriple(C, UniqueSYCLTriplesVec); + // Emit a warning if the detected CUDA version is too new. + if (Kind == Action::OFK_Cuda) { + auto &CudaInstallation = + static_cast<const toolchains::CudaToolChain &>(TC).CudaInstallation; + if (CudaInstallation.isValid()) + CudaInstallation.WarnIfUnsupportedVersion(); + } - // We'll need to use the SYCL and host triples as the key into - // getOffloadingDeviceToolChain, because the device toolchains we're - // going to create will depend on both. - const ToolChain *HostTC = C.getSingleOffloadToolChain<Action::OFK_Host>(); - for (const auto &TT : UniqueSYCLTriplesVec) { - auto &TC = getOffloadToolChain(C.getInputArgs(), Action::OFK_SYCL, TT, - HostTC->getTriple()); - C.addOffloadDeviceToolChain(&TC, Action::OFK_SYCL); - OffloadArchs[&TC] = getOffloadArchs(C, C.getArgs(), Action::OFK_SYCL, &TC, - /*SpecificToolchain=*/true); + C.addOffloadDeviceToolChain(&TC, Kind); } } - - // - // TODO: Add support for other offloading programming models here. - // } bool Driver::loadZOSCustomizationFile(llvm::cl::ExpansionContext &ExpCtx) { @@ -3306,9 +3303,6 @@ class OffloadingActionBuilder final { // architecture. If we are in host-only mode we return 'success' so that // the host uses the CUDA offload kind. if (auto *IA = dyn_cast<InputAction>(HostAction)) { - assert(!GpuArchList.empty() && - "We should have at least one GPU architecture."); - // If the host input is not CUDA or HIP, we don't need to bother about // this input. if (!(IA->getType() == types::TY_CUDA || @@ -3408,10 +3402,6 @@ class OffloadingActionBuilder final { CudaDeviceActions.clear(); } - /// Get canonicalized offload arch option. \returns empty StringRef if the - /// option is invalid. - virtual StringRef getCanonicalOffloadArch(StringRef Arch) = 0; - virtual std::optional<std::pair<llvm::StringRef, llvm::StringRef>> getConflictOffloadArchCombination(const std::set<StringRef> &GpuArchs) = 0; @@ -3440,91 +3430,25 @@ class OffloadingActionBuilder final { return true; } - ToolChains.push_back( - AssociatedOffloadKind == Action::OFK_Cuda - ? C.getSingleOffloadToolChain<Action::OFK_Cuda>() - : C.getSingleOffloadToolChain<Action::OFK_HIP>()); - - CompileHostOnly = C.getDriver().offloadHostOnly(); - EmitLLVM = Args.getLastArg(options::OPT_emit_llvm); - EmitAsm = Args.getLastArg(options::OPT_S); - - // --offload and --offload-arch options are mutually exclusive. - if (Args.hasArgNoClaim(options::OPT_offload_EQ) && - Args.hasArgNoClaim(options::OPT_offload_arch_EQ, - options::OPT_no_offload_arch_EQ)) { - C.getDriver().Diag(diag::err_opt_not_valid_with_opt) << "--offload-arch" - << "--offload"; - } - - // Collect all offload arch parameters, removing duplicates. std::set<StringRef> GpuArchs; - bool Error = false; - const ToolChain &TC = *ToolChains.front(); - for (Arg *A : C.getArgsForToolChain(&TC, /*BoundArch=*/"", - AssociatedOffloadKind)) { - if (!(A->getOption().matches(options::OPT_offload_arch_EQ) || - A->getOption().matches(options::OPT_no_offload_arch_EQ))) - continue; - A->claim(); - - for (StringRef ArchStr : llvm::split(A->getValue(), ",")) { - if (A->getOption().matches(options::OPT_no_offload_arch_EQ) && - ArchStr == "all") { - GpuArchs.clear(); - } else if (ArchStr == "native") { - auto GPUsOrErr = ToolChains.front()->getSystemGPUArchs(Args); - if (!GPUsOrErr) { - TC.getDriver().Diag(diag::err_drv_undetermined_gpu_arch) - << llvm::Triple::getArchTypeName(TC.getArch()) - << llvm::toString(GPUsOrErr.takeError()) << "--offload-arch"; - continue; - } + for (Action::OffloadKind Kind : {Action::OFK_Cuda, Action::OFK_HIP}) { + for (auto &I : llvm::make_range(C.getOffloadToolChains(Kind))) { + ToolChains.push_back(I.second); - for (auto GPU : *GPUsOrErr) { - GpuArchs.insert(Args.MakeArgString(GPU)); - } - } else { - ArchStr = getCanonicalOffloadArch(ArchStr); - if (ArchStr.empty()) { - Error = true; - } else if (A->getOption().matches(options::OPT_offload_arch_EQ)) - GpuArchs.insert(ArchStr); - else if (A->getOption().matches(options::OPT_no_offload_arch_EQ)) - GpuArchs.erase(ArchStr); - else - llvm_unreachable("Unexpected option."); - } + for (auto Arch : + C.getDriver().getOffloadArchs(C, C.getArgs(), Kind, *I.second)) + GpuArchs.insert(Arch); } } - auto &&ConflictingArchs = getConflictOffloadArchCombination(GpuArchs); - if (ConflictingArchs) { - C.getDriver().Diag(clang::diag::err_drv_bad_offload_arch_combo) - << ConflictingArchs->first << ConflictingArchs->second; - C.setContainsError(); - return true; - } - - // Collect list of GPUs remaining in the set. for (auto Arch : GpuArchs) GpuArchList.push_back(Arch.data()); - // Default to sm_20 which is the lowest common denominator for - // supported GPUs. sm_20 code should work correctly, if - // suboptimally, on all newer GPUs. - if (GpuArchList.empty()) { - if (ToolChains.front()->getTriple().isSPIRV()) { - if (ToolChains.front()->getTriple().getVendor() == llvm::Triple::AMD) - GpuArchList.push_back(OffloadArch::AMDGCNSPIRV); - else - GpuArchList.push_back(OffloadArch::Generic); - } else { - GpuArchList.push_back(DefaultOffloadArch); - } - } + CompileHostOnly = C.getDriver().offloadHostOnly(); + EmitLLVM = Args.getLastArg(options::OPT_emit_llvm); + EmitAsm = Args.getLastArg(options::OPT_S); - return Error; + return false; } }; @@ -3538,15 +3462,6 @@ class OffloadingActionBuilder final { DefaultOffloadArch = OffloadArch::CudaDefault; } - StringRef getCanonicalOffloadArch(StringRef ArchStr) override { - OffloadArch Arch = StringToOffloadArch(ArchStr); - if (Arch == OffloadArch::UNKNOWN || !IsNVIDIAOffloadArch(Arch)) { - C.getDriver().Diag(clang::diag::err_drv_cuda_bad_gpu_arch) << ArchStr; - return StringRef(); - } - return OffloadArchToString(Arch); - } - std::optional<std::pair<llvm::StringRef, llvm::StringRef>> getConflictOffloadArchCombination( const std::set<StringRef> &GpuArchs) override { @@ -3705,24 +3620,6 @@ class OffloadingActionBuilder final { bool canUseBundlerUnbundler() const override { return true; } - StringRef getCanonicalOffloadArch(StringRef IdStr) override { - llvm::StringMap<bool> Features; - // getHIPOffloadTargetTriple() is known to return valid value as it has - // been called successfully in the CreateOffloadingDeviceToolChains(). - auto T = - (IdStr == "amdgcnspirv") - ? llvm::Triple("spirv64-amd-amdhsa") - : *getHIPOffloadTargetTriple(C.getDriver(), C.getInputArgs()); - auto ArchStr = parseTargetID(T, IdStr, &Features); - if (!ArchStr) { - C.getDriver().Diag(clang::diag::err_drv_bad_target_id) << IdStr; - C.setContainsError(); - return StringRef(); - } - auto CanId = getCanonicalTargetID(*ArchStr, Features); - return Args.MakeArgStringRef(CanId); - }; - std::optional<std::pair<llvm::StringRef, llvm::StringRef>> getConflictOffloadArchCombination( const std::set<StringRef> &GpuArchs) override { @@ -4715,23 +4612,20 @@ void Driver::BuildActions(Compilation &C, DerivedArgList &Args, static StringRef getCanonicalArchString(Compilation &C, const llvm::opt::DerivedArgList &Args, StringRef ArchStr, - const llvm::Triple &Triple, - bool SpecificToolchain) { + const llvm::Triple &Triple) { // Lookup the CUDA / HIP architecture string. Only report an error if we were // expecting the triple to be only NVPTX / AMDGPU. OffloadArch Arch = StringToOffloadArch(getProcessorFromTargetID(Triple, ArchStr)); if (Triple.isNVPTX() && (Arch == OffloadArch::UNKNOWN || !IsNVIDIAOffloadArch(Arch))) { - if (SpecificToolchain) - C.getDriver().Diag(clang::diag::err_drv_offload_bad_gpu_arch) - << "CUDA" << ArchStr; + C.getDriver().Diag(clang::diag::err_drv_offload_bad_gpu_arch) + << "CUDA" << ArchStr; return StringRef(); } else if (Triple.isAMDGPU() && (Arch == OffloadArch::UNKNOWN || !IsAMDOffloadArch(Arch))) { - if (SpecificToolchain) - C.getDriver().Diag(clang::diag::err_drv_offload_bad_gpu_arch) - << "HIP" << ArchStr; + C.getDriver().Diag(clang::diag::err_drv_offload_bad_gpu_arch) + << "HIP" << ArchStr; return StringRef(); } @@ -4767,11 +4661,7 @@ getConflictOffloadArchCombination(const llvm::DenseSet<StringRef> &Archs, llvm::SmallVector<StringRef> Driver::getOffloadArchs(Compilation &C, const llvm::opt::DerivedArgList &Args, - Action::OffloadKind Kind, const ToolChain *TC, - bool SpecificToolchain) const { - if (!TC) - TC = &C.getDefaultToolChain(); - + Action::OffloadKind Kind, const ToolChain &TC) const { // --offload and --offload-arch options are mutually exclusive. if (Args.hasArgNoClaim(options::OPT_offload_EQ) && Args.hasArgNoClaim(options::OPT_offload_arch_EQ, @@ -4784,48 +4674,44 @@ Driver::getOffloadArchs(Compilation &C, const llvm::opt::DerivedArgList &Args, } llvm::DenseSet<StringRef> Archs; - for (auto *Arg : C.getArgsForToolChain(TC, /*BoundArch=*/"", Kind)) { + for (auto *Arg : C.getArgsForToolChain(&TC, /*BoundArch=*/"", Kind)) { // Add or remove the seen architectures in order of appearance. If an // invalid architecture is given we simply exit. if (Arg->getOption().matches(options::OPT_offload_arch_EQ)) { for (StringRef Arch : Arg->getValues()) { if (Arch == "native" || Arch.empty()) { - auto GPUsOrErr = TC->getSystemGPUArchs(Args); + auto GPUsOrErr = TC.getSystemGPUArchs(Args); if (!GPUsOrErr) { - if (!SpecificToolchain) - llvm::consumeError(GPUsOrErr.takeError()); - else - TC->getDriver().Diag(diag::err_drv_undetermined_gpu_arch) - << llvm::Triple::getArchTypeName(TC->getArch()) - << llvm::toString(GPUsOrErr.takeError()) << "--offload-arch"; + TC.getDriver().Diag(diag::err_drv_undetermined_gpu_arch) + << llvm::Triple::getArchTypeName(TC.getArch()) + << llvm::toString(GPUsOrErr.takeError()) << "--offload-arch"; continue; } for (auto ArchStr : *GPUsOrErr) { - StringRef CanonicalStr = - getCanonicalArchString(C, Args, Args.MakeArgString(ArchStr), - TC->getTriple(), SpecificToolchain); + StringRef CanonicalStr = getCanonicalArchString( + C, Args, Args.MakeArgString(ArchStr), TC.getTriple()); if (!CanonicalStr.empty()) Archs.insert(CanonicalStr); - else if (SpecificToolchain) + else return llvm::SmallVector<StringRef>(); } } else { - StringRef CanonicalStr = getCanonicalArchString( - C, Args, Arch, TC->getTriple(), SpecificToolchain); + StringRef CanonicalStr = + getCanonicalArchString(C, Args, Arch, TC.getTriple()); if (!CanonicalStr.empty()) Archs.insert(CanonicalStr); - else if (SpecificToolchain) + else return llvm::SmallVector<StringRef>(); } } } else if (Arg->getOption().matches(options::OPT_no_offload_arch_EQ)) { - for (StringRef Arch : llvm::split(Arg->getValue(), ",")) { + for (StringRef Arch : Arg->getValues()) { if (Arch == "all") { Archs.clear(); } else { - StringRef ArchStr = getCanonicalArchString( - C, Args, Arch, TC->getTriple(), SpecificToolchain); + StringRef ArchStr = + getCanonicalArchString(C, Args, Arch, TC.getTriple()); Archs.erase(ArchStr); } } @@ -4833,28 +4719,30 @@ Driver::getOffloadArchs(Compilation &C, const llvm::opt::DerivedArgList &Args, } if (auto ConflictingArchs = - getConflictOffloadArchCombination(Archs, TC->getTriple())) + getConflictOffloadArchCombination(Archs, TC.getTriple())) C.getDriver().Diag(clang::diag::err_drv_bad_offload_arch_combo) << ConflictingArchs->first << ConflictingArchs->second; - // Skip filling defaults if we're just querying what is availible. - if (SpecificToolchain && Archs.empty()) { + // Fill in the default architectures if not provided explicitly. + if (Archs.empty()) { if (Kind == Action::OFK_Cuda) { Archs.insert(OffloadArchToString(OffloadArch::CudaDefault)); } else if (Kind == Action::OFK_HIP) { - Archs.insert(OffloadArchToString(OffloadArch::HIPDefault)); + Archs.insert(OffloadArchToString(TC.getTriple().isSPIRV() + ? OffloadArch::Generic + : OffloadArch::HIPDefault)); } else if (Kind == Action::OFK_SYCL) { Archs.insert(StringRef()); } else if (Kind == Action::OFK_OpenMP) { // Accept legacy `-march` device arguments for OpenMP. - if (auto *Arg = C.getArgsForToolChain(TC, /*BoundArch=*/"", Kind) + if (auto *Arg = C.getArgsForToolChain(&TC, /*BoundArch=*/"", Kind) .getLastArg(options::OPT_march_EQ)) { Archs.insert(Arg->getValue()); } else { - auto ArchsOrErr = TC->getSystemGPUArchs(Args); + auto ArchsOrErr = TC.getSystemGPUArchs(Args); if (!ArchsOrErr) { - TC->getDriver().Diag(diag::err_drv_undetermined_gpu_arch) - << llvm::Triple::getArchTypeName(TC->getArch()) + TC.getDriver().Diag(diag::err_drv_undetermined_gpu_arch) + << llvm::Triple::getArchTypeName(TC.getArch()) << llvm::toString(ArchsOrErr.takeError()) << "--offload-arch"; } else if (!ArchsOrErr->empty()) { for (auto Arch : *ArchsOrErr) @@ -4934,7 +4822,7 @@ Action *Driver::BuildOffloadingActions(Compilation &C, // Get the product of all bound architectures and toolchains. SmallVector<std::pair<const ToolChain *, StringRef>> TCAndArchs; for (const ToolChain *TC : ToolChains) { - for (StringRef Arch : OffloadArchs.lookup(TC)) { + for (StringRef Arch : getOffloadArchs(C, C.getArgs(), Kind, *TC)) { TCAndArchs.push_back(std::make_pair(TC, Arch)); DeviceActions.push_back( C.MakeAction<InputAction>(*InputArg, InputType, CUID)); @@ -4966,7 +4854,7 @@ Action *Driver::BuildOffloadingActions(Compilation &C, if (Kind == Action::OFK_SYCL && Phase == phases::Assemble) continue; - auto TCAndArch = TCAndArchs.begin(); + auto *TCAndArch = TCAndArchs.begin(); for (Action *&A : DeviceActions) { if (A->getType() == types::TY_Nothing) continue; @@ -4998,7 +4886,13 @@ Action *Driver::BuildOffloadingActions(Compilation &C, // Compiling HIP in device-only non-RDC mode requires linking each action // individually. for (Action *&A : DeviceActions) { - if ((A->getType() != types::TY_Object && + // Special handling for the HIP SPIR-V toolchain because it doesn't use + // the SPIR-V backend yet doesn't report the output as an object. + bool IsAMDGCNSPIRV = A->getOffloadingToolChain() && + A->getOffloadingToolChain()->getTriple().getOS() == + llvm::Triple::OSType::AMDHSA && + A->getOffloadingToolChain()->getTriple().isSPIRV(); + if ((A->getType() != types::TY_Object && !IsAMDGCNSPIRV && A->getType() != types::TY_LTO_BC) || !HIPNoRDC || !offloadDeviceOnly()) continue; @@ -5006,7 +4900,7 @@ Action *Driver::BuildOffloadingActions(Compilation &C, A = C.MakeAction<LinkJobAction>(LinkerInput, types::TY_Image); } - auto TCAndArch = TCAndArchs.begin(); + auto *TCAndArch = TCAndArchs.begin(); for (Action *A : DeviceActions) { DDeps.add(*A, *TCAndArch->first, TCAndArch->second.data(), Kind); OffloadAction::DeviceDependences DDep; @@ -5054,8 +4948,9 @@ Action *Driver::BuildOffloadingActions(Compilation &C, // fatbinary for each translation unit, linking each input individually. Action *FatbinAction = C.MakeAction<LinkJobAction>(OffloadActions, types::TY_HIP_FATBIN); - DDep.add(*FatbinAction, *C.getSingleOffloadToolChain<Action::OFK_HIP>(), - nullptr, Action::OFK_HIP); + DDep.add(*FatbinAction, + *C.getOffloadToolChains<Action::OFK_HIP>().first->second, nullptr, + Action::OFK_HIP); } else { // Package all the offloading actions into a single output that can be // embedded in the host and linked. @@ -5210,7 +5105,10 @@ Action *Driver::ConstructPhaseAction( false) || (Args.hasFlag(options::OPT_offload_new_driver, options::OPT_no_offload_new_driver, false) && - !offloadDeviceOnly())) || + (!offloadDeviceOnly() || + (Input->getOffloadingToolChain() && + TargetDeviceOffloadKind == Action::OFK_HIP && + Input->getOffloadingToolChain()->getTriple().isSPIRV())))) || TargetDeviceOffloadKind == Action::OFK_OpenMP))) { types::ID Output = Args.hasArg(options::OPT_S) && diff --git a/clang/lib/Driver/ToolChain.cpp b/clang/lib/Driver/ToolChain.cpp index 481f575..47f93fa1 100644 --- a/clang/lib/Driver/ToolChain.cpp +++ b/clang/lib/Driver/ToolChain.cpp @@ -104,44 +104,6 @@ ToolChain::ToolChain(const Driver &D, const llvm::Triple &T, addIfExists(getFilePaths(), Path); } -llvm::Expected<std::unique_ptr<llvm::MemoryBuffer>> -ToolChain::executeToolChainProgram(StringRef Executable) const { - llvm::SmallString<64> OutputFile; - llvm::sys::fs::createTemporaryFile("toolchain-program", "txt", OutputFile, - llvm::sys::fs::OF_Text); - llvm::FileRemover OutputRemover(OutputFile.c_str()); - std::optional<llvm::StringRef> Redirects[] = { - {""}, - OutputFile.str(), - {""}, - }; - - std::string ErrorMessage; - int SecondsToWait = 60; - if (std::optional<std::string> Str = - llvm::sys::Process::GetEnv("CLANG_TOOLCHAIN_PROGRAM_TIMEOUT")) { - if (!llvm::to_integer(*Str, SecondsToWait)) - return llvm::createStringError(std::error_code(), - "CLANG_TOOLCHAIN_PROGRAM_TIMEOUT expected " - "an integer, got '" + - *Str + "'"); - SecondsToWait = std::max(SecondsToWait, 0); // infinite - } - if (llvm::sys::ExecuteAndWait(Executable, {Executable}, {}, Redirects, - SecondsToWait, - /*MemoryLimit=*/0, &ErrorMessage)) - return llvm::createStringError(std::error_code(), - Executable + ": " + ErrorMessage); - - llvm::ErrorOr<std::unique_ptr<llvm::MemoryBuffer>> OutputBuf = - llvm::MemoryBuffer::getFile(OutputFile.c_str()); - if (!OutputBuf) - return llvm::createStringError(OutputBuf.getError(), - "Failed to read stdout of " + Executable + - ": " + OutputBuf.getError().message()); - return std::move(*OutputBuf); -} - void ToolChain::setTripleEnvironment(llvm::Triple::EnvironmentType Env) { Triple.setEnvironment(Env); if (EffectiveTriple != llvm::Triple()) @@ -255,6 +217,18 @@ static void getAArch64MultilibFlags(const Driver &D, Result.push_back(ABIArg->getAsString(Args)); } + if (const Arg *A = Args.getLastArg(options::OPT_O_Group); + A && A->getOption().matches(options::OPT_O)) { + switch (A->getValue()[0]) { + case 's': + Result.push_back("-Os"); + break; + case 'z': + Result.push_back("-Oz"); + break; + } + } + processMultilibCustomFlags(Result, Args); } @@ -332,6 +306,19 @@ static void getARMMultilibFlags(const Driver &D, const llvm::Triple &Triple, if (Endian->getOption().matches(options::OPT_mbig_endian)) Result.push_back(Endian->getAsString(Args)); } + + if (const Arg *A = Args.getLastArg(options::OPT_O_Group); + A && A->getOption().matches(options::OPT_O)) { + switch (A->getValue()[0]) { + case 's': + Result.push_back("-Os"); + break; + case 'z': + Result.push_back("-Oz"); + break; + } + } + processMultilibCustomFlags(Result, Args); } @@ -1644,7 +1631,8 @@ void ToolChain::addSYCLIncludeArgs(const ArgList &DriverArgs, ArgStringList &CC1Args) const {} llvm::SmallVector<ToolChain::BitCodeLibraryInfo, 12> -ToolChain::getDeviceLibs(const ArgList &DriverArgs) const { +ToolChain::getDeviceLibs(const ArgList &DriverArgs, + const Action::OffloadKind DeviceOffloadingKind) const { return {}; } diff --git a/clang/lib/Driver/ToolChains/AMDGPU.cpp b/clang/lib/Driver/ToolChains/AMDGPU.cpp index 7fc34f4..0781683 100644 --- a/clang/lib/Driver/ToolChains/AMDGPU.cpp +++ b/clang/lib/Driver/ToolChains/AMDGPU.cpp @@ -31,6 +31,68 @@ using namespace clang::driver::toolchains; using namespace clang; using namespace llvm::opt; +RocmInstallationDetector::CommonBitcodeLibsPreferences:: + CommonBitcodeLibsPreferences(const Driver &D, + const llvm::opt::ArgList &DriverArgs, + StringRef GPUArch, + const Action::OffloadKind DeviceOffloadingKind, + const bool NeedsASanRT) + : ABIVer(DeviceLibABIVersion::fromCodeObjectVersion( + tools::getAMDGPUCodeObjectVersion(D, DriverArgs))) { + const auto Kind = llvm::AMDGPU::parseArchAMDGCN(GPUArch); + const unsigned ArchAttr = llvm::AMDGPU::getArchAttrAMDGCN(Kind); + + IsOpenMP = DeviceOffloadingKind == Action::OFK_OpenMP; + + const bool HasWave32 = (ArchAttr & llvm::AMDGPU::FEATURE_WAVE32); + Wave64 = + !HasWave32 || DriverArgs.hasFlag(options::OPT_mwavefrontsize64, + options::OPT_mno_wavefrontsize64, false); + + const bool IsKnownOffloading = DeviceOffloadingKind == Action::OFK_OpenMP || + DeviceOffloadingKind == Action::OFK_HIP; + + // Default to enabling f32 denormals on subtargets where fma is fast with + // denormals + const bool DefaultDAZ = + (Kind == llvm::AMDGPU::GK_NONE) + ? false + : !((ArchAttr & llvm::AMDGPU::FEATURE_FAST_FMA_F32) && + (ArchAttr & llvm::AMDGPU::FEATURE_FAST_DENORMAL_F32)); + // TODO: There are way too many flags that change this. Do we need to + // check them all? + DAZ = IsKnownOffloading + ? DriverArgs.hasFlag(options::OPT_fgpu_flush_denormals_to_zero, + options::OPT_fno_gpu_flush_denormals_to_zero, + DefaultDAZ) + : DriverArgs.hasArg(options::OPT_cl_denorms_are_zero) || DefaultDAZ; + + FiniteOnly = DriverArgs.hasArg(options::OPT_cl_finite_math_only) || + DriverArgs.hasFlag(options::OPT_ffinite_math_only, + options::OPT_fno_finite_math_only, false); + + UnsafeMathOpt = + DriverArgs.hasArg(options::OPT_cl_unsafe_math_optimizations) || + DriverArgs.hasFlag(options::OPT_funsafe_math_optimizations, + options::OPT_fno_unsafe_math_optimizations, false); + + FastRelaxedMath = DriverArgs.hasArg(options::OPT_cl_fast_relaxed_math) || + DriverArgs.hasFlag(options::OPT_ffast_math, + options::OPT_fno_fast_math, false); + + const bool DefaultSqrt = IsKnownOffloading ? true : false; + CorrectSqrt = + DriverArgs.hasArg(options::OPT_cl_fp32_correctly_rounded_divide_sqrt) || + DriverArgs.hasFlag( + options::OPT_fhip_fp32_correctly_rounded_divide_sqrt, + options::OPT_fno_hip_fp32_correctly_rounded_divide_sqrt, DefaultSqrt); + // GPU Sanitizer currently only supports ASan and is enabled through host + // ASan. + GPUSan = (DriverArgs.hasFlag(options::OPT_fgpu_sanitize, + options::OPT_fno_gpu_sanitize, true) && + NeedsASanRT); +} + void RocmInstallationDetector::scanLibDevicePath(llvm::StringRef Path) { assert(!Path.empty()); @@ -841,7 +903,7 @@ AMDGPUToolChain::getSystemGPUArchs(const ArgList &Args) const { else Program = GetProgramPath("amdgpu-arch"); - auto StdoutOrErr = executeToolChainProgram(Program); + auto StdoutOrErr = getDriver().executeProgram({Program}); if (!StdoutOrErr) return StdoutOrErr.takeError(); @@ -884,33 +946,14 @@ void ROCMToolChain::addClangTargetOptions( ABIVer)) return; - bool Wave64 = isWave64(DriverArgs, Kind); - // TODO: There are way too many flags that change this. Do we need to check - // them all? - bool DAZ = DriverArgs.hasArg(options::OPT_cl_denorms_are_zero) || - getDefaultDenormsAreZeroForTarget(Kind); - bool FiniteOnly = DriverArgs.hasArg(options::OPT_cl_finite_math_only); - - bool UnsafeMathOpt = - DriverArgs.hasArg(options::OPT_cl_unsafe_math_optimizations); - bool FastRelaxedMath = DriverArgs.hasArg(options::OPT_cl_fast_relaxed_math); - bool CorrectSqrt = - DriverArgs.hasArg(options::OPT_cl_fp32_correctly_rounded_divide_sqrt); - - // GPU Sanitizer currently only supports ASan and is enabled through host - // ASan. - bool GPUSan = DriverArgs.hasFlag(options::OPT_fgpu_sanitize, - options::OPT_fno_gpu_sanitize, true) && - getSanitizerArgs(DriverArgs).needsAsanRt(); - // Add the OpenCL specific bitcode library. llvm::SmallVector<BitCodeLibraryInfo, 12> BCLibs; BCLibs.emplace_back(RocmInstallation->getOpenCLPath().str()); // Add the generic set of libraries. BCLibs.append(RocmInstallation->getCommonBitcodeLibs( - DriverArgs, LibDeviceFile, Wave64, DAZ, FiniteOnly, UnsafeMathOpt, - FastRelaxedMath, CorrectSqrt, ABIVer, GPUSan, false)); + DriverArgs, LibDeviceFile, GpuArch, DeviceOffloadingKind, + getSanitizerArgs(DriverArgs).needsAsanRt())); for (auto [BCFile, Internalize] : BCLibs) { if (Internalize) @@ -947,35 +990,37 @@ bool RocmInstallationDetector::checkCommonBitcodeLibs( llvm::SmallVector<ToolChain::BitCodeLibraryInfo, 12> RocmInstallationDetector::getCommonBitcodeLibs( - const llvm::opt::ArgList &DriverArgs, StringRef LibDeviceFile, bool Wave64, - bool DAZ, bool FiniteOnly, bool UnsafeMathOpt, bool FastRelaxedMath, - bool CorrectSqrt, DeviceLibABIVersion ABIVer, bool GPUSan, - bool isOpenMP) const { + const llvm::opt::ArgList &DriverArgs, StringRef LibDeviceFile, + StringRef GPUArch, const Action::OffloadKind DeviceOffloadingKind, + const bool NeedsASanRT) const { llvm::SmallVector<ToolChain::BitCodeLibraryInfo, 12> BCLibs; + CommonBitcodeLibsPreferences Pref{D, DriverArgs, GPUArch, + DeviceOffloadingKind, NeedsASanRT}; + auto AddBCLib = [&](ToolChain::BitCodeLibraryInfo BCLib, bool Internalize = true) { BCLib.ShouldInternalize = Internalize; BCLibs.emplace_back(BCLib); }; auto AddSanBCLibs = [&]() { - if (GPUSan) + if (Pref.GPUSan) AddBCLib(getAsanRTLPath(), false); }; AddSanBCLibs(); AddBCLib(getOCMLPath()); - if (!isOpenMP) + if (!Pref.IsOpenMP) AddBCLib(getOCKLPath()); - else if (GPUSan && isOpenMP) + else if (Pref.GPUSan && Pref.IsOpenMP) AddBCLib(getOCKLPath(), false); - AddBCLib(getDenormalsAreZeroPath(DAZ)); - AddBCLib(getUnsafeMathPath(UnsafeMathOpt || FastRelaxedMath)); - AddBCLib(getFiniteOnlyPath(FiniteOnly || FastRelaxedMath)); - AddBCLib(getCorrectlyRoundedSqrtPath(CorrectSqrt)); - AddBCLib(getWavefrontSize64Path(Wave64)); + AddBCLib(getDenormalsAreZeroPath(Pref.DAZ)); + AddBCLib(getUnsafeMathPath(Pref.UnsafeMathOpt || Pref.FastRelaxedMath)); + AddBCLib(getFiniteOnlyPath(Pref.FiniteOnly || Pref.FastRelaxedMath)); + AddBCLib(getCorrectlyRoundedSqrtPath(Pref.CorrectSqrt)); + AddBCLib(getWavefrontSize64Path(Pref.Wave64)); AddBCLib(LibDeviceFile); - auto ABIVerPath = getABIVersionPath(ABIVer); + auto ABIVerPath = getABIVersionPath(Pref.ABIVer); if (!ABIVerPath.empty()) AddBCLib(ABIVerPath); @@ -983,9 +1028,9 @@ RocmInstallationDetector::getCommonBitcodeLibs( } llvm::SmallVector<ToolChain::BitCodeLibraryInfo, 12> -ROCMToolChain::getCommonDeviceLibNames(const llvm::opt::ArgList &DriverArgs, - const std::string &GPUArch, - bool isOpenMP) const { +ROCMToolChain::getCommonDeviceLibNames( + const llvm::opt::ArgList &DriverArgs, const std::string &GPUArch, + Action::OffloadKind DeviceOffloadingKind) const { auto Kind = llvm::AMDGPU::parseArchAMDGCN(GPUArch); const StringRef CanonArch = llvm::AMDGPU::getArchNameAMDGCN(Kind); @@ -996,33 +1041,9 @@ ROCMToolChain::getCommonDeviceLibNames(const llvm::opt::ArgList &DriverArgs, ABIVer)) return {}; - // If --hip-device-lib is not set, add the default bitcode libraries. - // TODO: There are way too many flags that change this. Do we need to check - // them all? - bool DAZ = DriverArgs.hasFlag(options::OPT_fgpu_flush_denormals_to_zero, - options::OPT_fno_gpu_flush_denormals_to_zero, - getDefaultDenormsAreZeroForTarget(Kind)); - bool FiniteOnly = DriverArgs.hasFlag( - options::OPT_ffinite_math_only, options::OPT_fno_finite_math_only, false); - bool UnsafeMathOpt = - DriverArgs.hasFlag(options::OPT_funsafe_math_optimizations, - options::OPT_fno_unsafe_math_optimizations, false); - bool FastRelaxedMath = DriverArgs.hasFlag(options::OPT_ffast_math, - options::OPT_fno_fast_math, false); - bool CorrectSqrt = DriverArgs.hasFlag( - options::OPT_fhip_fp32_correctly_rounded_divide_sqrt, - options::OPT_fno_hip_fp32_correctly_rounded_divide_sqrt, true); - bool Wave64 = isWave64(DriverArgs, Kind); - - // GPU Sanitizer currently only supports ASan and is enabled through host - // ASan. - bool GPUSan = DriverArgs.hasFlag(options::OPT_fgpu_sanitize, - options::OPT_fno_gpu_sanitize, true) && - getSanitizerArgs(DriverArgs).needsAsanRt(); - return RocmInstallation->getCommonBitcodeLibs( - DriverArgs, LibDeviceFile, Wave64, DAZ, FiniteOnly, UnsafeMathOpt, - FastRelaxedMath, CorrectSqrt, ABIVer, GPUSan, isOpenMP); + DriverArgs, LibDeviceFile, GPUArch, DeviceOffloadingKind, + getSanitizerArgs(DriverArgs).needsAsanRt()); } bool AMDGPUToolChain::shouldSkipSanitizeOption( diff --git a/clang/lib/Driver/ToolChains/AMDGPU.h b/clang/lib/Driver/ToolChains/AMDGPU.h index 08bd4fa..513c77d 100644 --- a/clang/lib/Driver/ToolChains/AMDGPU.h +++ b/clang/lib/Driver/ToolChains/AMDGPU.h @@ -147,7 +147,7 @@ public: llvm::SmallVector<BitCodeLibraryInfo, 12> getCommonDeviceLibNames(const llvm::opt::ArgList &DriverArgs, const std::string &GPUArch, - bool isOpenMP = false) const; + Action::OffloadKind DeviceOffloadingKind) const; SanitizerMask getSupportedSanitizers() const override { return SanitizerKind::Address; diff --git a/clang/lib/Driver/ToolChains/AMDGPUOpenMP.cpp b/clang/lib/Driver/ToolChains/AMDGPUOpenMP.cpp index 7ffa3f0..2b41d54 100644 --- a/clang/lib/Driver/ToolChains/AMDGPUOpenMP.cpp +++ b/clang/lib/Driver/ToolChains/AMDGPUOpenMP.cpp @@ -44,7 +44,7 @@ void AMDGPUOpenMPToolChain::addClangTargetOptions( true)) return; - for (auto BCFile : getDeviceLibs(DriverArgs)) { + for (auto BCFile : getDeviceLibs(DriverArgs, DeviceOffloadingKind)) { CC1Args.push_back(BCFile.ShouldInternalize ? "-mlink-builtin-bitcode" : "-mlink-bitcode-file"); CC1Args.push_back(DriverArgs.MakeArgString(BCFile.Path)); @@ -132,7 +132,9 @@ AMDGPUOpenMPToolChain::computeMSVCVersion(const Driver *D, } llvm::SmallVector<ToolChain::BitCodeLibraryInfo, 12> -AMDGPUOpenMPToolChain::getDeviceLibs(const llvm::opt::ArgList &Args) const { +AMDGPUOpenMPToolChain::getDeviceLibs( + const llvm::opt::ArgList &Args, + const Action::OffloadKind DeviceOffloadingKind) const { if (!Args.hasFlag(options::OPT_offloadlib, options::OPT_no_offloadlib, true)) return {}; @@ -140,8 +142,8 @@ AMDGPUOpenMPToolChain::getDeviceLibs(const llvm::opt::ArgList &Args) const { getTriple(), Args.getLastArgValue(options::OPT_march_EQ)); SmallVector<BitCodeLibraryInfo, 12> BCLibs; - for (auto BCLib : getCommonDeviceLibNames(Args, GpuArch.str(), - /*IsOpenMP=*/true)) + for (auto BCLib : + getCommonDeviceLibNames(Args, GpuArch.str(), DeviceOffloadingKind)) BCLibs.emplace_back(BCLib); return BCLibs; diff --git a/clang/lib/Driver/ToolChains/AMDGPUOpenMP.h b/clang/lib/Driver/ToolChains/AMDGPUOpenMP.h index 0536c9f..cbafdf5 100644 --- a/clang/lib/Driver/ToolChains/AMDGPUOpenMP.h +++ b/clang/lib/Driver/ToolChains/AMDGPUOpenMP.h @@ -58,7 +58,8 @@ public: const llvm::opt::ArgList &Args) const override; llvm::SmallVector<BitCodeLibraryInfo, 12> - getDeviceLibs(const llvm::opt::ArgList &Args) const override; + getDeviceLibs(const llvm::opt::ArgList &Args, + const Action::OffloadKind DeviceOffloadKind) const override; const ToolChain &HostTC; }; diff --git a/clang/lib/Driver/ToolChains/Arch/Sparc.cpp b/clang/lib/Driver/ToolChains/Arch/Sparc.cpp index 504f110..94a94f1 100644 --- a/clang/lib/Driver/ToolChains/Arch/Sparc.cpp +++ b/clang/lib/Driver/ToolChains/Arch/Sparc.cpp @@ -23,7 +23,9 @@ const char *sparc::getSparcAsmModeForCPU(StringRef Name, if (Triple.getArch() == llvm::Triple::sparcv9) { const char *DefV9CPU; - if (Triple.isOSLinux() || Triple.isOSFreeBSD() || Triple.isOSOpenBSD()) + if (Triple.isOSSolaris()) + DefV9CPU = "-Av9b"; + else if (Triple.isOSLinux() || Triple.isOSFreeBSD() || Triple.isOSOpenBSD()) DefV9CPU = "-Av9a"; else DefV9CPU = "-Av9"; @@ -35,6 +37,13 @@ const char *sparc::getSparcAsmModeForCPU(StringRef Name, .Case("niagara4", "-Av9d") .Default(DefV9CPU); } else { + const char *DefV8CPU; + + if (Triple.isOSSolaris()) + DefV8CPU = "-Av8plus"; + else + DefV8CPU = "-Av8"; + return llvm::StringSwitch<const char *>(Name) .Case("v8", "-Av8") .Case("supersparc", "-Av8") @@ -70,7 +79,7 @@ const char *sparc::getSparcAsmModeForCPU(StringRef Name, .Case("gr712rc", "-Aleon") .Case("leon4", "-Aleon") .Case("gr740", "-Aleon") - .Default("-Av8"); + .Default(DefV8CPU); } } @@ -157,6 +166,9 @@ void sparc::getSparcTargetFeatures(const Driver &D, const llvm::Triple &Triple, bool IsSparcV9ATarget = (Triple.getArch() == llvm::Triple::sparcv9) && (Triple.isOSLinux() || Triple.isOSFreeBSD() || Triple.isOSOpenBSD()); + bool IsSparcV9BTarget = Triple.isOSSolaris(); + bool IsSparcV8PlusTarget = + Triple.getArch() == llvm::Triple::sparc && Triple.isOSSolaris(); if (Arg *A = Args.getLastArg(options::OPT_mvis, options::OPT_mno_vis)) { if (A->getOption().matches(options::OPT_mvis)) Features.push_back("+vis"); @@ -171,6 +183,8 @@ void sparc::getSparcTargetFeatures(const Driver &D, const llvm::Triple &Triple, Features.push_back("+vis2"); else Features.push_back("-vis2"); + } else if (IsSparcV9BTarget) { + Features.push_back("+vis2"); } if (Arg *A = Args.getLastArg(options::OPT_mvis3, options::OPT_mno_vis3)) { @@ -191,6 +205,8 @@ void sparc::getSparcTargetFeatures(const Driver &D, const llvm::Triple &Triple, if (Arg *A = Args.getLastArg(options::OPT_mv8plus, options::OPT_mno_v8plus)) { if (A->getOption().matches(options::OPT_mv8plus)) Features.push_back("+v8plus"); + } else if (IsSparcV8PlusTarget) { + Features.push_back("+v8plus"); } if (Args.hasArg(options::OPT_ffixed_g1)) diff --git a/clang/lib/Driver/ToolChains/BareMetal.cpp b/clang/lib/Driver/ToolChains/BareMetal.cpp index e670696..497f333 100644 --- a/clang/lib/Driver/ToolChains/BareMetal.cpp +++ b/clang/lib/Driver/ToolChains/BareMetal.cpp @@ -694,9 +694,6 @@ void baremetal::Linker::ConstructJob(Compilation &C, const JobAction &JA, NeedCRTs) CmdArgs.push_back(Args.MakeArgString(TC.GetFilePath(CRTEnd))); - if (TC.getTriple().isRISCV()) - CmdArgs.push_back("-X"); - // The R_ARM_TARGET2 relocation must be treated as R_ARM_REL32 on arm*-*-elf // and arm*-*-eabi (the default is R_ARM_GOT_PREL, used on arm*-*-linux and // arm*-*-*bsd). diff --git a/clang/lib/Driver/ToolChains/Clang.cpp b/clang/lib/Driver/ToolChains/Clang.cpp index 1fc7002..7d0c142 100644 --- a/clang/lib/Driver/ToolChains/Clang.cpp +++ b/clang/lib/Driver/ToolChains/Clang.cpp @@ -97,32 +97,15 @@ forAllAssociatedToolChains(Compilation &C, const JobAction &JA, // Apply Work on all the offloading tool chains associated with the current // action. - if (JA.isHostOffloading(Action::OFK_Cuda)) - Work(*C.getSingleOffloadToolChain<Action::OFK_Cuda>()); - else if (JA.isDeviceOffloading(Action::OFK_Cuda)) - Work(*C.getSingleOffloadToolChain<Action::OFK_Host>()); - else if (JA.isHostOffloading(Action::OFK_HIP)) - Work(*C.getSingleOffloadToolChain<Action::OFK_HIP>()); - else if (JA.isDeviceOffloading(Action::OFK_HIP)) - Work(*C.getSingleOffloadToolChain<Action::OFK_Host>()); - - if (JA.isHostOffloading(Action::OFK_OpenMP)) { - auto TCs = C.getOffloadToolChains<Action::OFK_OpenMP>(); - for (auto II = TCs.first, IE = TCs.second; II != IE; ++II) - Work(*II->second); - } else if (JA.isDeviceOffloading(Action::OFK_OpenMP)) - Work(*C.getSingleOffloadToolChain<Action::OFK_Host>()); - - if (JA.isHostOffloading(Action::OFK_SYCL)) { - auto TCs = C.getOffloadToolChains<Action::OFK_SYCL>(); - for (auto II = TCs.first, IE = TCs.second; II != IE; ++II) - Work(*II->second); - } else if (JA.isDeviceOffloading(Action::OFK_SYCL)) - Work(*C.getSingleOffloadToolChain<Action::OFK_Host>()); - - // - // TODO: Add support for other offloading programming models here. - // + for (Action::OffloadKind Kind : {Action::OFK_Cuda, Action::OFK_OpenMP, + Action::OFK_HIP, Action::OFK_SYCL}) { + if (JA.isHostOffloading(Kind)) { + auto TCs = C.getOffloadToolChains(Kind); + for (auto II = TCs.first, IE = TCs.second; II != IE; ++II) + Work(*II->second); + } else if (JA.isDeviceOffloading(Kind)) + Work(*C.getSingleOffloadToolChain<Action::OFK_Host>()); + } } static bool @@ -4985,8 +4968,8 @@ void Clang::ConstructJob(Compilation &C, const JobAction &JA, else { // Host-side compilation. NormalizedTriple = - (IsCuda ? C.getSingleOffloadToolChain<Action::OFK_Cuda>() - : C.getSingleOffloadToolChain<Action::OFK_HIP>()) + (IsCuda ? C.getOffloadToolChains(Action::OFK_Cuda).first->second + : C.getOffloadToolChains(Action::OFK_HIP).first->second) ->getTriple() .normalize(); if (IsCuda) { diff --git a/clang/lib/Driver/ToolChains/Cuda.cpp b/clang/lib/Driver/ToolChains/Cuda.cpp index 2373d94..7d803be 100644 --- a/clang/lib/Driver/ToolChains/Cuda.cpp +++ b/clang/lib/Driver/ToolChains/Cuda.cpp @@ -815,7 +815,7 @@ NVPTXToolChain::getSystemGPUArchs(const ArgList &Args) const { else Program = GetProgramPath("nvptx-arch"); - auto StdoutOrErr = executeToolChainProgram(Program); + auto StdoutOrErr = getDriver().executeProgram({Program}); if (!StdoutOrErr) return StdoutOrErr.takeError(); diff --git a/clang/lib/Driver/ToolChains/HIPAMD.cpp b/clang/lib/Driver/ToolChains/HIPAMD.cpp index 5fe0f85..b4c6da0 100644 --- a/clang/lib/Driver/ToolChains/HIPAMD.cpp +++ b/clang/lib/Driver/ToolChains/HIPAMD.cpp @@ -264,7 +264,7 @@ void HIPAMDToolChain::addClangTargetOptions( return; // No DeviceLibs for SPIR-V. } - for (auto BCFile : getDeviceLibs(DriverArgs)) { + for (auto BCFile : getDeviceLibs(DriverArgs, DeviceOffloadingKind)) { CC1Args.push_back(BCFile.ShouldInternalize ? "-mlink-builtin-bitcode" : "-mlink-bitcode-file"); CC1Args.push_back(DriverArgs.MakeArgString(BCFile.Path)); @@ -355,7 +355,8 @@ VersionTuple HIPAMDToolChain::computeMSVCVersion(const Driver *D, } llvm::SmallVector<ToolChain::BitCodeLibraryInfo, 12> -HIPAMDToolChain::getDeviceLibs(const llvm::opt::ArgList &DriverArgs) const { +HIPAMDToolChain::getDeviceLibs(const llvm::opt::ArgList &DriverArgs, + Action::OffloadKind DeviceOffloadingKind) const { llvm::SmallVector<BitCodeLibraryInfo, 12> BCLibs; if (!DriverArgs.hasFlag(options::OPT_offloadlib, options::OPT_no_offloadlib, true) || @@ -397,7 +398,8 @@ HIPAMDToolChain::getDeviceLibs(const llvm::opt::ArgList &DriverArgs) const { assert(!GpuArch.empty() && "Must have an explicit GPU arch."); // Add common device libraries like ocml etc. - for (auto N : getCommonDeviceLibNames(DriverArgs, GpuArch.str())) + for (auto N : getCommonDeviceLibNames(DriverArgs, GpuArch.str(), + DeviceOffloadingKind)) BCLibs.emplace_back(N); // Add instrument lib. diff --git a/clang/lib/Driver/ToolChains/HIPAMD.h b/clang/lib/Driver/ToolChains/HIPAMD.h index 3630b11..bcc3ebb 100644 --- a/clang/lib/Driver/ToolChains/HIPAMD.h +++ b/clang/lib/Driver/ToolChains/HIPAMD.h @@ -80,7 +80,8 @@ public: void AddHIPIncludeArgs(const llvm::opt::ArgList &DriverArgs, llvm::opt::ArgStringList &CC1Args) const override; llvm::SmallVector<BitCodeLibraryInfo, 12> - getDeviceLibs(const llvm::opt::ArgList &Args) const override; + getDeviceLibs(const llvm::opt::ArgList &Args, + Action::OffloadKind DeviceOffloadKind) const override; SanitizerMask getSupportedSanitizers() const override; diff --git a/clang/lib/Driver/ToolChains/HIPSPV.cpp b/clang/lib/Driver/ToolChains/HIPSPV.cpp index 53649ca..643a67f 100644 --- a/clang/lib/Driver/ToolChains/HIPSPV.cpp +++ b/clang/lib/Driver/ToolChains/HIPSPV.cpp @@ -149,7 +149,8 @@ void HIPSPVToolChain::addClangTargetOptions( CC1Args.append( {"-fvisibility=hidden", "-fapply-global-visibility-to-externs"}); - for (const BitCodeLibraryInfo &BCFile : getDeviceLibs(DriverArgs)) + for (const BitCodeLibraryInfo &BCFile : + getDeviceLibs(DriverArgs, DeviceOffloadingKind)) CC1Args.append( {"-mlink-builtin-bitcode", DriverArgs.MakeArgString(BCFile.Path)}); } @@ -200,7 +201,9 @@ void HIPSPVToolChain::AddHIPIncludeArgs(const ArgList &DriverArgs, } llvm::SmallVector<ToolChain::BitCodeLibraryInfo, 12> -HIPSPVToolChain::getDeviceLibs(const llvm::opt::ArgList &DriverArgs) const { +HIPSPVToolChain::getDeviceLibs( + const llvm::opt::ArgList &DriverArgs, + const Action::OffloadKind DeviceOffloadingKind) const { llvm::SmallVector<ToolChain::BitCodeLibraryInfo, 12> BCLibs; if (!DriverArgs.hasFlag(options::OPT_offloadlib, options::OPT_no_offloadlib, true)) diff --git a/clang/lib/Driver/ToolChains/HIPSPV.h b/clang/lib/Driver/ToolChains/HIPSPV.h index ecd82e7..caf6924 100644 --- a/clang/lib/Driver/ToolChains/HIPSPV.h +++ b/clang/lib/Driver/ToolChains/HIPSPV.h @@ -69,7 +69,8 @@ public: void AddHIPIncludeArgs(const llvm::opt::ArgList &DriverArgs, llvm::opt::ArgStringList &CC1Args) const override; llvm::SmallVector<BitCodeLibraryInfo, 12> - getDeviceLibs(const llvm::opt::ArgList &Args) const override; + getDeviceLibs(const llvm::opt::ArgList &Args, + const Action::OffloadKind DeviceOffloadKind) const override; SanitizerMask getSupportedSanitizers() const override; diff --git a/clang/lib/Driver/ToolChains/OpenBSD.cpp b/clang/lib/Driver/ToolChains/OpenBSD.cpp index 79b1b69..8f58918 100644 --- a/clang/lib/Driver/ToolChains/OpenBSD.cpp +++ b/clang/lib/Driver/ToolChains/OpenBSD.cpp @@ -161,7 +161,7 @@ void openbsd::Linker::ConstructJob(Compilation &C, const JobAction &JA, if (Nopie || Profiling) CmdArgs.push_back("-nopie"); - if (Triple.isRISCV64()) { + if (Triple.isLoongArch64() || Triple.isRISCV64()) { CmdArgs.push_back("-X"); if (Args.hasArg(options::OPT_mno_relax)) CmdArgs.push_back("--no-relax"); diff --git a/clang/lib/Driver/ToolChains/ROCm.h b/clang/lib/Driver/ToolChains/ROCm.h index 2a09da01..ebd5443 100644 --- a/clang/lib/Driver/ToolChains/ROCm.h +++ b/clang/lib/Driver/ToolChains/ROCm.h @@ -11,6 +11,7 @@ #include "clang/Basic/Cuda.h" #include "clang/Basic/LLVM.h" +#include "clang/Driver/CommonArgs.h" #include "clang/Driver/Driver.h" #include "clang/Driver/Options.h" #include "clang/Driver/SanitizerArgs.h" @@ -18,6 +19,7 @@ #include "llvm/ADT/StringMap.h" #include "llvm/Option/ArgList.h" #include "llvm/Support/VersionTuple.h" +#include "llvm/TargetParser/TargetParser.h" #include "llvm/TargetParser/Triple.h" namespace clang { @@ -77,6 +79,24 @@ private: SPACKReleaseStr(SPACKReleaseStr.str()) {} }; + struct CommonBitcodeLibsPreferences { + CommonBitcodeLibsPreferences(const Driver &D, + const llvm::opt::ArgList &DriverArgs, + StringRef GPUArch, + const Action::OffloadKind DeviceOffloadingKind, + const bool NeedsASanRT); + + DeviceLibABIVersion ABIVer; + bool IsOpenMP; + bool Wave64; + bool DAZ; + bool FiniteOnly; + bool UnsafeMathOpt; + bool FastRelaxedMath; + bool CorrectSqrt; + bool GPUSan; + }; + const Driver &D; bool HasHIPRuntime = false; bool HasDeviceLibrary = false; @@ -175,11 +195,11 @@ public: /// Get file paths of default bitcode libraries common to AMDGPU based /// toolchains. - llvm::SmallVector<ToolChain::BitCodeLibraryInfo, 12> getCommonBitcodeLibs( - const llvm::opt::ArgList &DriverArgs, StringRef LibDeviceFile, - bool Wave64, bool DAZ, bool FiniteOnly, bool UnsafeMathOpt, - bool FastRelaxedMath, bool CorrectSqrt, DeviceLibABIVersion ABIVer, - bool GPUSan, bool isOpenMP) const; + llvm::SmallVector<ToolChain::BitCodeLibraryInfo, 12> + getCommonBitcodeLibs(const llvm::opt::ArgList &DriverArgs, + StringRef LibDeviceFile, StringRef GPUArch, + const Action::OffloadKind DeviceOffloadingKind, + const bool NeedsASanRT) const; /// Check file paths of default bitcode libraries common to AMDGPU based /// toolchains. \returns false if there are invalid or missing files. bool checkCommonBitcodeLibs(StringRef GPUArch, StringRef LibDeviceFile, diff --git a/clang/lib/Format/BreakableToken.cpp b/clang/lib/Format/BreakableToken.cpp index c36cb74..29db200 100644 --- a/clang/lib/Format/BreakableToken.cpp +++ b/clang/lib/Format/BreakableToken.cpp @@ -25,7 +25,7 @@ namespace clang { namespace format { -static constexpr StringRef Blanks = " \t\v\f\r"; +static constexpr StringRef Blanks(" \t\v\f\r"); static StringRef getLineCommentIndentPrefix(StringRef Comment, const FormatStyle &Style) { @@ -513,7 +513,7 @@ BreakableBlockComment::BreakableBlockComment( Decoration = ""; } for (size_t i = 1, e = Content.size(); i < e && !Decoration.empty(); ++i) { - const StringRef &Text = Content[i]; + const StringRef Text(Content[i]); if (i + 1 == e) { // If the last line is empty, the closing "*/" will have a star. if (Text.empty()) diff --git a/clang/lib/Format/IntegerLiteralSeparatorFixer.cpp b/clang/lib/Format/IntegerLiteralSeparatorFixer.cpp index 87823ae..80487fa 100644 --- a/clang/lib/Format/IntegerLiteralSeparatorFixer.cpp +++ b/clang/lib/Format/IntegerLiteralSeparatorFixer.cpp @@ -19,7 +19,7 @@ namespace format { enum class Base { Binary, Decimal, Hex, Other }; -static Base getBase(const StringRef IntegerLiteral) { +static Base getBase(StringRef IntegerLiteral) { assert(IntegerLiteral.size() > 1); if (IntegerLiteral[0] > '0') { @@ -164,8 +164,8 @@ IntegerLiteralSeparatorFixer::process(const Environment &Env, return {Result, 0}; } -bool IntegerLiteralSeparatorFixer::checkSeparator( - const StringRef IntegerLiteral, int DigitsPerGroup) const { +bool IntegerLiteralSeparatorFixer::checkSeparator(StringRef IntegerLiteral, + int DigitsPerGroup) const { assert(DigitsPerGroup > 0); int I = 0; @@ -184,7 +184,7 @@ bool IntegerLiteralSeparatorFixer::checkSeparator( return true; } -std::string IntegerLiteralSeparatorFixer::format(const StringRef IntegerLiteral, +std::string IntegerLiteralSeparatorFixer::format(StringRef IntegerLiteral, int DigitsPerGroup, int DigitCount, bool RemoveSeparator) const { diff --git a/clang/lib/Format/IntegerLiteralSeparatorFixer.h b/clang/lib/Format/IntegerLiteralSeparatorFixer.h index 2c158e4..e24af18 100644 --- a/clang/lib/Format/IntegerLiteralSeparatorFixer.h +++ b/clang/lib/Format/IntegerLiteralSeparatorFixer.h @@ -26,8 +26,8 @@ public: const FormatStyle &Style); private: - bool checkSeparator(const StringRef IntegerLiteral, int DigitsPerGroup) const; - std::string format(const StringRef IntegerLiteral, int DigitsPerGroup, + bool checkSeparator(StringRef IntegerLiteral, int DigitsPerGroup) const; + std::string format(StringRef IntegerLiteral, int DigitsPerGroup, int DigitCount, bool RemoveSeparator) const; char Separator; diff --git a/clang/lib/Format/ObjCPropertyAttributeOrderFixer.cpp b/clang/lib/Format/ObjCPropertyAttributeOrderFixer.cpp index 37a1807..b885942 100644 --- a/clang/lib/Format/ObjCPropertyAttributeOrderFixer.cpp +++ b/clang/lib/Format/ObjCPropertyAttributeOrderFixer.cpp @@ -66,7 +66,7 @@ void ObjCPropertyAttributeOrderFixer::sortPropertyAttributes( return; } - const StringRef Attribute{Tok->TokenText}; + const StringRef Attribute(Tok->TokenText); StringRef Value; // Also handle `getter=getFoo` attributes. diff --git a/clang/lib/Frontend/CompilerInvocation.cpp b/clang/lib/Frontend/CompilerInvocation.cpp index 3a36250..ab4384a 100644 --- a/clang/lib/Frontend/CompilerInvocation.cpp +++ b/clang/lib/Frontend/CompilerInvocation.cpp @@ -2013,8 +2013,8 @@ bool CompilerInvocation::ParseCodeGenArgs(CodeGenOptions &Opts, ArgList &Args, : llvm::codegenoptions::DebugTemplateNamesKind::Mangled); } - if (const Arg *A = Args.getLastArg(OPT_ftime_report, OPT_ftime_report_EQ, - OPT_ftime_report_json)) { + if (Args.hasArg(OPT_ftime_report, OPT_ftime_report_EQ, OPT_ftime_report_json, + OPT_stats_file_timers)) { Opts.TimePasses = true; // -ftime-report= is only for new pass manager. @@ -2026,7 +2026,7 @@ bool CompilerInvocation::ParseCodeGenArgs(CodeGenOptions &Opts, ArgList &Args, Opts.TimePassesPerRun = true; else Diags.Report(diag::err_drv_invalid_value) - << A->getAsString(Args) << A->getValue(); + << EQ->getAsString(Args) << EQ->getValue(); } if (Args.getLastArg(OPT_ftime_report_json)) diff --git a/clang/lib/Interpreter/Interpreter.cpp b/clang/lib/Interpreter/Interpreter.cpp index db6a2bb..9b71486 100644 --- a/clang/lib/Interpreter/Interpreter.cpp +++ b/clang/lib/Interpreter/Interpreter.cpp @@ -761,10 +761,18 @@ Interpreter::getSymbolAddressFromLinkerName(llvm::StringRef Name) const { llvm::Error Interpreter::Undo(unsigned N) { - if (N > getEffectivePTUSize()) + if (getEffectivePTUSize() == 0) { return llvm::make_error<llvm::StringError>("Operation failed. " - "Too many undos", + "No input left to undo", std::error_code()); + } else if (N > getEffectivePTUSize()) { + return llvm::make_error<llvm::StringError>( + llvm::formatv( + "Operation failed. Wanted to undo {0} inputs, only have {1}.", N, + getEffectivePTUSize()), + std::error_code()); + } + for (unsigned I = 0; I < N; I++) { if (IncrExecutor) { if (llvm::Error Err = IncrExecutor->removeModule(PTUs.back())) diff --git a/clang/lib/Lex/Pragma.cpp b/clang/lib/Lex/Pragma.cpp index 01c85e6..bba3c89 100644 --- a/clang/lib/Lex/Pragma.cpp +++ b/clang/lib/Lex/Pragma.cpp @@ -591,7 +591,8 @@ IdentifierInfo *Preprocessor::ParsePragmaPushOrPopMacro(Token &Tok) { } // Remember the macro string. - std::string StrVal = getSpelling(Tok); + Token StrTok = Tok; + std::string StrVal = getSpelling(StrTok); // Read the ')'. Lex(Tok); @@ -604,6 +605,15 @@ IdentifierInfo *Preprocessor::ParsePragmaPushOrPopMacro(Token &Tok) { assert(StrVal[0] == '"' && StrVal[StrVal.size()-1] == '"' && "Invalid string token!"); + if (StrVal.size() <= 2) { + Diag(StrTok.getLocation(), diag::warn_pargma_push_pop_macro_empty_string) + << SourceRange( + StrTok.getLocation(), + StrTok.getLocation().getLocWithOffset(StrTok.getLength())) + << PragmaTok.getIdentifierInfo()->isStr("pop_macro"); + return nullptr; + } + // Create a Token from the string. Token MacroTok; MacroTok.startToken(); diff --git a/clang/lib/Parse/ParseDecl.cpp b/clang/lib/Parse/ParseDecl.cpp index 893ef02..e47caeb 100644 --- a/clang/lib/Parse/ParseDecl.cpp +++ b/clang/lib/Parse/ParseDecl.cpp @@ -5695,11 +5695,10 @@ Parser::DeclGroupPtrTy Parser::ParseTopLevelStmtDecl() { Scope::CompoundStmtScope); TopLevelStmtDecl *TLSD = Actions.ActOnStartTopLevelStmtDecl(getCurScope()); StmtResult R = ParseStatementOrDeclaration(Stmts, SubStmtCtx); + Actions.ActOnFinishTopLevelStmtDecl(TLSD, R.get()); if (!R.isUsable()) R = Actions.ActOnNullStmt(Tok.getLocation()); - Actions.ActOnFinishTopLevelStmtDecl(TLSD, R.get()); - if (Tok.is(tok::annot_repl_input_end) && Tok.getAnnotationValue() != nullptr) { ConsumeAnnotationToken(); diff --git a/clang/lib/Sema/AnalysisBasedWarnings.cpp b/clang/lib/Sema/AnalysisBasedWarnings.cpp index d1400cb..829c81b 100644 --- a/clang/lib/Sema/AnalysisBasedWarnings.cpp +++ b/clang/lib/Sema/AnalysisBasedWarnings.cpp @@ -2901,8 +2901,7 @@ void clang::sema::AnalysisBasedWarnings::IssueWarnings( .setAlwaysAdd(Stmt::UnaryOperatorClass); } - bool EnableLifetimeSafetyAnalysis = !Diags.isIgnored( - diag::warn_experimental_lifetime_safety_dummy_warning, D->getBeginLoc()); + bool EnableLifetimeSafetyAnalysis = S.getLangOpts().EnableLifetimeSafety; // Install the logical handler. std::optional<LogicalErrorHandler> LEH; if (LogicalErrorHandler::hasActiveDiagnostics(Diags, D->getBeginLoc())) { @@ -3029,8 +3028,8 @@ void clang::sema::AnalysisBasedWarnings::IssueWarnings( // TODO: Enable lifetime safety analysis for other languages once it is // stable. if (EnableLifetimeSafetyAnalysis && S.getLangOpts().CPlusPlus) { - if (CFG *cfg = AC.getCFG()) - runLifetimeSafetyAnalysis(*cast<DeclContext>(D), *cfg, AC); + if (AC.getCFG()) + lifetimes::runLifetimeSafetyAnalysis(AC); } // Check for violations of "called once" parameter properties. if (S.getLangOpts().ObjC && !S.getLangOpts().CPlusPlus && diff --git a/clang/lib/Sema/SemaARM.cpp b/clang/lib/Sema/SemaARM.cpp index bd603a9..8e27fab 100644 --- a/clang/lib/Sema/SemaARM.cpp +++ b/clang/lib/Sema/SemaARM.cpp @@ -1535,4 +1535,95 @@ bool SemaARM::areLaxCompatibleSveTypes(QualType FirstType, IsLaxCompatible(SecondType, FirstType); } +bool SemaARM::checkTargetVersionAttr(const StringRef Param, + const SourceLocation Loc) { + using namespace DiagAttrParams; + + llvm::SmallVector<StringRef, 8> Features; + Param.split(Features, '+'); + for (StringRef Feat : Features) { + Feat = Feat.trim(); + if (Feat == "default") + continue; + if (!getASTContext().getTargetInfo().validateCpuSupports(Feat)) + return Diag(Loc, diag::warn_unsupported_target_attribute) + << Unsupported << None << Feat << TargetVersion; + } + return false; +} + +bool SemaARM::checkTargetClonesAttr( + SmallVectorImpl<StringRef> &Params, SmallVectorImpl<SourceLocation> &Locs, + SmallVectorImpl<SmallString<64>> &NewParams) { + using namespace DiagAttrParams; + + if (!getASTContext().getTargetInfo().hasFeature("fmv")) + return true; + + assert(Params.size() == Locs.size() && + "Mismatch between number of string parameters and locations"); + + bool HasDefault = false; + bool HasNonDefault = false; + for (unsigned I = 0, E = Params.size(); I < E; ++I) { + const StringRef Param = Params[I].trim(); + const SourceLocation &Loc = Locs[I]; + + if (Param.empty()) + return Diag(Loc, diag::warn_unsupported_target_attribute) + << Unsupported << None << "" << TargetClones; + + if (Param == "default") { + if (HasDefault) + Diag(Loc, diag::warn_target_clone_duplicate_options); + else { + NewParams.push_back(Param); + HasDefault = true; + } + continue; + } + + bool HasCodeGenImpact = false; + llvm::SmallVector<StringRef, 8> Features; + llvm::SmallVector<StringRef, 8> ValidFeatures; + Param.split(Features, '+'); + for (StringRef Feat : Features) { + Feat = Feat.trim(); + if (!getASTContext().getTargetInfo().validateCpuSupports(Feat)) { + Diag(Loc, diag::warn_unsupported_target_attribute) + << Unsupported << None << Feat << TargetClones; + continue; + } + if (getASTContext().getTargetInfo().doesFeatureAffectCodeGen(Feat)) + HasCodeGenImpact = true; + ValidFeatures.push_back(Feat); + } + + // Ignore features that don't impact code generation. + if (!HasCodeGenImpact) { + Diag(Loc, diag::warn_target_clone_no_impact_options); + continue; + } + + if (ValidFeatures.empty()) + continue; + + // Canonicalize attribute parameter. + llvm::sort(ValidFeatures); + SmallString<64> NewParam(llvm::join(ValidFeatures, "+")); + if (llvm::is_contained(NewParams, NewParam)) { + Diag(Loc, diag::warn_target_clone_duplicate_options); + continue; + } + + // Valid non-default argument. + NewParams.push_back(NewParam); + HasNonDefault = true; + } + if (!HasNonDefault) + return true; + + return false; +} + } // namespace clang diff --git a/clang/lib/Sema/SemaChecking.cpp b/clang/lib/Sema/SemaChecking.cpp index 5e523fe..c74b671 100644 --- a/clang/lib/Sema/SemaChecking.cpp +++ b/clang/lib/Sema/SemaChecking.cpp @@ -3013,6 +3013,8 @@ Sema::CheckBuiltinFunctionCall(FunctionDecl *FDecl, unsigned BuiltinID, case Builtin::BI__builtin_elementwise_maxnum: case Builtin::BI__builtin_elementwise_minimum: case Builtin::BI__builtin_elementwise_maximum: + case Builtin::BI__builtin_elementwise_minimumnum: + case Builtin::BI__builtin_elementwise_maximumnum: case Builtin::BI__builtin_elementwise_atan2: case Builtin::BI__builtin_elementwise_fmod: case Builtin::BI__builtin_elementwise_pow: diff --git a/clang/lib/Sema/SemaConcept.cpp b/clang/lib/Sema/SemaConcept.cpp index 5205ca0b..044cf5c 100644 --- a/clang/lib/Sema/SemaConcept.cpp +++ b/clang/lib/Sema/SemaConcept.cpp @@ -588,6 +588,9 @@ static bool CheckConstraintSatisfaction( return true; for (const AssociatedConstraint &AC : AssociatedConstraints) { + if (AC.isNull()) + return true; + Sema::ArgPackSubstIndexRAII _(S, AC.ArgPackSubstIndex); ExprResult Res = calculateConstraintSatisfaction( S, Template, TemplateIDRange.getBegin(), TemplateArgsLists, diff --git a/clang/lib/Sema/SemaDecl.cpp b/clang/lib/Sema/SemaDecl.cpp index 14403e6..6e66698 100644 --- a/clang/lib/Sema/SemaDecl.cpp +++ b/clang/lib/Sema/SemaDecl.cpp @@ -18476,6 +18476,10 @@ CreateNewDecl: // record. AddPushedVisibilityAttribute(New); + // If this is not a definition, process API notes for it now. + if (TUK != TagUseKind::Definition) + ProcessAPINotes(New); + if (isMemberSpecialization && !New->isInvalidDecl()) CompleteMemberSpecialization(New, Previous); @@ -20569,7 +20573,8 @@ TopLevelStmtDecl *Sema::ActOnStartTopLevelStmtDecl(Scope *S) { } void Sema::ActOnFinishTopLevelStmtDecl(TopLevelStmtDecl *D, Stmt *Statement) { - D->setStmt(Statement); + if (Statement) + D->setStmt(Statement); PopCompoundScope(); PopFunctionScopeInfo(); PopDeclContext(); diff --git a/clang/lib/Sema/SemaDeclAttr.cpp b/clang/lib/Sema/SemaDeclAttr.cpp index 78f4804..9a2950c 100644 --- a/clang/lib/Sema/SemaDeclAttr.cpp +++ b/clang/lib/Sema/SemaDeclAttr.cpp @@ -3254,9 +3254,8 @@ static void handleCodeSegAttr(Sema &S, Decl *D, const ParsedAttr &AL) { } bool Sema::checkTargetAttr(SourceLocation LiteralLoc, StringRef AttrStr) { - enum FirstParam { Unsupported, Duplicate, Unknown }; - enum SecondParam { None, CPU, Tune }; - enum ThirdParam { Target, TargetClones }; + using namespace DiagAttrParams; + if (AttrStr.contains("fpmath=")) return Diag(LiteralLoc, diag::warn_unsupported_target_attribute) << Unsupported << None << "fpmath=" << Target; @@ -3331,80 +3330,22 @@ bool Sema::checkTargetAttr(SourceLocation LiteralLoc, StringRef AttrStr) { return false; } -bool Sema::checkTargetVersionAttr(SourceLocation LiteralLoc, Decl *D, - StringRef AttrStr) { - enum FirstParam { Unsupported }; - enum SecondParam { None }; - enum ThirdParam { Target, TargetClones, TargetVersion }; - llvm::SmallVector<StringRef, 8> Features; - if (Context.getTargetInfo().getTriple().isRISCV()) { - llvm::SmallVector<StringRef, 8> AttrStrs; - AttrStr.split(AttrStrs, ';'); - - bool HasArch = false; - bool HasPriority = false; - bool HasDefault = false; - bool DuplicateAttr = false; - for (auto &AttrStr : AttrStrs) { - // Only support arch=+ext,... syntax. - if (AttrStr.starts_with("arch=+")) { - if (HasArch) - DuplicateAttr = true; - HasArch = true; - ParsedTargetAttr TargetAttr = - Context.getTargetInfo().parseTargetAttr(AttrStr); - - if (TargetAttr.Features.empty() || - llvm::any_of(TargetAttr.Features, [&](const StringRef Ext) { - return !RISCV().isValidFMVExtension(Ext); - })) - return Diag(LiteralLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << AttrStr << TargetVersion; - } else if (AttrStr.starts_with("default")) { - if (HasDefault) - DuplicateAttr = true; - HasDefault = true; - } else if (AttrStr.consume_front("priority=")) { - if (HasPriority) - DuplicateAttr = true; - HasPriority = true; - unsigned Digit; - if (AttrStr.getAsInteger(0, Digit)) - return Diag(LiteralLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << AttrStr << TargetVersion; - } else { - return Diag(LiteralLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << AttrStr << TargetVersion; - } - } - - if (((HasPriority || HasArch) && HasDefault) || DuplicateAttr || - (HasPriority && !HasArch)) - return Diag(LiteralLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << AttrStr << TargetVersion; +static void handleTargetVersionAttr(Sema &S, Decl *D, const ParsedAttr &AL) { + StringRef Param; + SourceLocation Loc; + if (!S.checkStringLiteralArgumentAttr(AL, 0, Param, &Loc)) + return; - return false; - } - AttrStr.split(Features, "+"); - for (auto &CurFeature : Features) { - CurFeature = CurFeature.trim(); - if (CurFeature == "default") - continue; - if (!Context.getTargetInfo().validateCpuSupports(CurFeature)) - return Diag(LiteralLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << CurFeature << TargetVersion; + if (S.Context.getTargetInfo().getTriple().isAArch64()) { + if (S.ARM().checkTargetVersionAttr(Param, Loc)) + return; + } else if (S.Context.getTargetInfo().getTriple().isRISCV()) { + if (S.RISCV().checkTargetVersionAttr(Param, Loc)) + return; } - return false; -} -static void handleTargetVersionAttr(Sema &S, Decl *D, const ParsedAttr &AL) { - StringRef Str; - SourceLocation LiteralLoc; - if (!S.checkStringLiteralArgumentAttr(AL, 0, Str, &LiteralLoc) || - S.checkTargetVersionAttr(LiteralLoc, D, Str)) - return; TargetVersionAttr *NewAttr = - ::new (S.Context) TargetVersionAttr(S.Context, AL, Str); + ::new (S.Context) TargetVersionAttr(S.Context, AL, Param); D->addAttr(NewAttr); } @@ -3419,158 +3360,7 @@ static void handleTargetAttr(Sema &S, Decl *D, const ParsedAttr &AL) { D->addAttr(NewAttr); } -bool Sema::checkTargetClonesAttrString( - SourceLocation LiteralLoc, StringRef Str, const StringLiteral *Literal, - Decl *D, bool &HasDefault, bool &HasCommas, bool &HasNotDefault, - SmallVectorImpl<SmallString<64>> &StringsBuffer) { - enum FirstParam { Unsupported, Duplicate, Unknown }; - enum SecondParam { None, CPU, Tune }; - enum ThirdParam { Target, TargetClones }; - HasCommas = HasCommas || Str.contains(','); - const TargetInfo &TInfo = Context.getTargetInfo(); - // Warn on empty at the beginning of a string. - if (Str.size() == 0) - return Diag(LiteralLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << "" << TargetClones; - - std::pair<StringRef, StringRef> Parts = {{}, Str}; - while (!Parts.second.empty()) { - Parts = Parts.second.split(','); - StringRef Cur = Parts.first.trim(); - SourceLocation CurLoc = - Literal->getLocationOfByte(Cur.data() - Literal->getString().data(), - getSourceManager(), getLangOpts(), TInfo); - - bool DefaultIsDupe = false; - bool HasCodeGenImpact = false; - if (Cur.empty()) - return Diag(CurLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << "" << TargetClones; - - if (TInfo.getTriple().isAArch64()) { - // AArch64 target clones specific - if (Cur == "default") { - DefaultIsDupe = HasDefault; - HasDefault = true; - if (llvm::is_contained(StringsBuffer, Cur) || DefaultIsDupe) - Diag(CurLoc, diag::warn_target_clone_duplicate_options); - else - StringsBuffer.push_back(Cur); - } else { - std::pair<StringRef, StringRef> CurParts = {{}, Cur}; - llvm::SmallVector<StringRef, 8> CurFeatures; - while (!CurParts.second.empty()) { - CurParts = CurParts.second.split('+'); - StringRef CurFeature = CurParts.first.trim(); - if (!TInfo.validateCpuSupports(CurFeature)) { - Diag(CurLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << CurFeature << TargetClones; - continue; - } - if (TInfo.doesFeatureAffectCodeGen(CurFeature)) - HasCodeGenImpact = true; - CurFeatures.push_back(CurFeature); - } - // Canonize TargetClones Attributes - llvm::sort(CurFeatures); - SmallString<64> Res; - for (auto &CurFeat : CurFeatures) { - if (!Res.empty()) - Res.append("+"); - Res.append(CurFeat); - } - if (llvm::is_contained(StringsBuffer, Res) || DefaultIsDupe) - Diag(CurLoc, diag::warn_target_clone_duplicate_options); - else if (!HasCodeGenImpact) - // Ignore features in target_clone attribute that don't impact - // code generation - Diag(CurLoc, diag::warn_target_clone_no_impact_options); - else if (!Res.empty()) { - StringsBuffer.push_back(Res); - HasNotDefault = true; - } - } - } else if (TInfo.getTriple().isRISCV()) { - // Suppress warn_target_clone_mixed_values - HasCommas = false; - - // Cur is split's parts of Str. RISC-V uses Str directly, - // so skip when encountered more than once. - if (!Str.starts_with(Cur)) - continue; - - llvm::SmallVector<StringRef, 8> AttrStrs; - Str.split(AttrStrs, ";"); - - bool IsPriority = false; - bool IsDefault = false; - for (auto &AttrStr : AttrStrs) { - // Only support arch=+ext,... syntax. - if (AttrStr.starts_with("arch=+")) { - ParsedTargetAttr TargetAttr = - Context.getTargetInfo().parseTargetAttr(AttrStr); - - if (TargetAttr.Features.empty() || - llvm::any_of(TargetAttr.Features, [&](const StringRef Ext) { - return !RISCV().isValidFMVExtension(Ext); - })) - return Diag(CurLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << Str << TargetClones; - } else if (AttrStr.starts_with("default")) { - IsDefault = true; - DefaultIsDupe = HasDefault; - HasDefault = true; - } else if (AttrStr.consume_front("priority=")) { - IsPriority = true; - unsigned Digit; - if (AttrStr.getAsInteger(0, Digit)) - return Diag(CurLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << Str << TargetClones; - } else { - return Diag(CurLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << Str << TargetClones; - } - } - - if (IsPriority && IsDefault) - return Diag(CurLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << Str << TargetClones; - - if (llvm::is_contained(StringsBuffer, Str) || DefaultIsDupe) - Diag(CurLoc, diag::warn_target_clone_duplicate_options); - StringsBuffer.push_back(Str); - } else { - // Other targets ( currently X86 ) - if (Cur.starts_with("arch=")) { - if (!Context.getTargetInfo().isValidCPUName( - Cur.drop_front(sizeof("arch=") - 1))) - return Diag(CurLoc, diag::warn_unsupported_target_attribute) - << Unsupported << CPU << Cur.drop_front(sizeof("arch=") - 1) - << TargetClones; - } else if (Cur == "default") { - DefaultIsDupe = HasDefault; - HasDefault = true; - } else if (!Context.getTargetInfo().isValidFeatureName(Cur) || - Context.getTargetInfo().getFMVPriority(Cur) == 0) - return Diag(CurLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << Cur << TargetClones; - if (llvm::is_contained(StringsBuffer, Cur) || DefaultIsDupe) - Diag(CurLoc, diag::warn_target_clone_duplicate_options); - // Note: Add even if there are duplicates, since it changes name mangling. - StringsBuffer.push_back(Cur); - } - } - if (Str.rtrim().ends_with(",")) - return Diag(LiteralLoc, diag::warn_unsupported_target_attribute) - << Unsupported << None << "" << TargetClones; - return false; -} - static void handleTargetClonesAttr(Sema &S, Decl *D, const ParsedAttr &AL) { - if (S.Context.getTargetInfo().getTriple().isAArch64() && - !S.Context.getTargetInfo().hasFeature("fmv")) - return; - // Ensure we don't combine these with themselves, since that causes some // confusing behavior. if (const auto *Other = D->getAttr<TargetClonesAttr>()) { @@ -3581,31 +3371,6 @@ static void handleTargetClonesAttr(Sema &S, Decl *D, const ParsedAttr &AL) { if (checkAttrMutualExclusion<TargetClonesAttr>(S, D, AL)) return; - SmallVector<StringRef, 2> Strings; - SmallVector<SmallString<64>, 2> StringsBuffer; - bool HasCommas = false, HasDefault = false, HasNotDefault = false; - - for (unsigned I = 0, E = AL.getNumArgs(); I != E; ++I) { - StringRef CurStr; - SourceLocation LiteralLoc; - if (!S.checkStringLiteralArgumentAttr(AL, I, CurStr, &LiteralLoc) || - S.checkTargetClonesAttrString( - LiteralLoc, CurStr, - cast<StringLiteral>(AL.getArgAsExpr(I)->IgnoreParenCasts()), D, - HasDefault, HasCommas, HasNotDefault, StringsBuffer)) - return; - } - for (auto &SmallStr : StringsBuffer) - Strings.push_back(SmallStr.str()); - - if (HasCommas && AL.getNumArgs() > 1) - S.Diag(AL.getLoc(), diag::warn_target_clone_mixed_values); - - if (!HasDefault && !S.Context.getTargetInfo().getTriple().isAArch64()) { - S.Diag(AL.getLoc(), diag::err_target_clone_must_have_default); - return; - } - // FIXME: We could probably figure out how to get this to work for lambdas // someday. if (const auto *MD = dyn_cast<CXXMethodDecl>(D)) { @@ -3617,13 +3382,34 @@ static void handleTargetClonesAttr(Sema &S, Decl *D, const ParsedAttr &AL) { } } - // No multiversion if we have default version only. - if (S.Context.getTargetInfo().getTriple().isAArch64() && !HasNotDefault) - return; + SmallVector<StringRef, 2> Params; + SmallVector<SourceLocation, 2> Locations; + for (unsigned I = 0, E = AL.getNumArgs(); I != E; ++I) { + StringRef Param; + SourceLocation Loc; + if (!S.checkStringLiteralArgumentAttr(AL, I, Param, &Loc)) + return; + Params.push_back(Param); + Locations.push_back(Loc); + } + + SmallVector<SmallString<64>, 2> NewParams; + if (S.Context.getTargetInfo().getTriple().isAArch64()) { + if (S.ARM().checkTargetClonesAttr(Params, Locations, NewParams)) + return; + } else if (S.Context.getTargetInfo().getTriple().isRISCV()) { + if (S.RISCV().checkTargetClonesAttr(Params, Locations, NewParams)) + return; + } else if (S.Context.getTargetInfo().getTriple().isX86()) { + if (S.X86().checkTargetClonesAttr(Params, Locations, NewParams)) + return; + } + Params.clear(); + for (auto &SmallStr : NewParams) + Params.push_back(SmallStr.str()); - cast<FunctionDecl>(D)->setIsMultiVersion(); TargetClonesAttr *NewAttr = ::new (S.Context) - TargetClonesAttr(S.Context, AL, Strings.data(), Strings.size()); + TargetClonesAttr(S.Context, AL, Params.data(), Params.size()); D->addAttr(NewAttr); } diff --git a/clang/lib/Sema/SemaOpenMP.cpp b/clang/lib/Sema/SemaOpenMP.cpp index 4ecc9b0..2c5d97c 100644 --- a/clang/lib/Sema/SemaOpenMP.cpp +++ b/clang/lib/Sema/SemaOpenMP.cpp @@ -2829,7 +2829,7 @@ static void checkReductionClauses(Sema &S, DSAStackTy *Stack, continue; } for (Expr *Ref : RC->varlist()) { - assert(Ref && "NULL expr in OpenMP nontemporal clause."); + assert(Ref && "NULL expr in OpenMP reduction clause."); SourceLocation ELoc; SourceRange ERange; Expr *SimpleRefExpr = Ref; @@ -7612,6 +7612,23 @@ void SemaOpenMP::ActOnOpenMPDeclareVariantDirective( return; } + // OpenMP 6.0 [9.6.2 (page 332, line 31-33, adjust_args clause, Restrictions] + // If the `need_device_addr` adjust-op modifier is present, each list item + // that appears in the clause must refer to an argument in the declaration of + // the function variant that has a reference type + if (getLangOpts().OpenMP >= 60) { + for (Expr *E : AdjustArgsNeedDeviceAddr) { + E = E->IgnoreParenImpCasts(); + if (const auto *DRE = dyn_cast<DeclRefExpr>(E)) { + if (const auto *VD = dyn_cast<VarDecl>(DRE->getDecl())) { + if (!VD->getType()->isReferenceType()) + Diag(E->getExprLoc(), + diag::err_omp_non_by_ref_need_device_addr_modifier_argument); + } + } + } + } + auto *NewAttr = OMPDeclareVariantAttr::CreateImplicit( getASTContext(), VariantRef, &TI, const_cast<Expr **>(AdjustArgsNothing.data()), AdjustArgsNothing.size(), @@ -18344,7 +18361,7 @@ OMPClause *SemaOpenMP::ActOnOpenMPSharedClause(ArrayRef<Expr *> VarList, SourceLocation EndLoc) { SmallVector<Expr *, 8> Vars; for (Expr *RefExpr : VarList) { - assert(RefExpr && "NULL expr in OpenMP lastprivate clause."); + assert(RefExpr && "NULL expr in OpenMP shared clause."); SourceLocation ELoc; SourceRange ERange; Expr *SimpleRefExpr = RefExpr; @@ -19991,7 +20008,7 @@ OMPClause *SemaOpenMP::ActOnOpenMPAlignedClause( SourceLocation LParenLoc, SourceLocation ColonLoc, SourceLocation EndLoc) { SmallVector<Expr *, 8> Vars; for (Expr *RefExpr : VarList) { - assert(RefExpr && "NULL expr in OpenMP linear clause."); + assert(RefExpr && "NULL expr in OpenMP aligned clause."); SourceLocation ELoc; SourceRange ERange; Expr *SimpleRefExpr = RefExpr; @@ -20167,7 +20184,7 @@ OMPClause *SemaOpenMP::ActOnOpenMPCopyprivateClause(ArrayRef<Expr *> VarList, SmallVector<Expr *, 8> DstExprs; SmallVector<Expr *, 8> AssignmentOps; for (Expr *RefExpr : VarList) { - assert(RefExpr && "NULL expr in OpenMP linear clause."); + assert(RefExpr && "NULL expr in OpenMP copyprivate clause."); SourceLocation ELoc; SourceRange ERange; Expr *SimpleRefExpr = RefExpr; @@ -20526,7 +20543,7 @@ OMPClause *SemaOpenMP::ActOnOpenMPDependClause( TotalDepCount = VarOffset.TotalDepCount; } else { for (Expr *RefExpr : VarList) { - assert(RefExpr && "NULL expr in OpenMP shared clause."); + assert(RefExpr && "NULL expr in OpenMP depend clause."); if (isa<DependentScopeDeclRefExpr>(RefExpr)) { // It will be analyzed later. Vars.push_back(RefExpr); @@ -23737,7 +23754,7 @@ OMPClause *SemaOpenMP::ActOnOpenMPAllocateClause( // Analyze and build list of variables. SmallVector<Expr *, 8> Vars; for (Expr *RefExpr : VarList) { - assert(RefExpr && "NULL expr in OpenMP private clause."); + assert(RefExpr && "NULL expr in OpenMP allocate clause."); SourceLocation ELoc; SourceRange ERange; Expr *SimpleRefExpr = RefExpr; @@ -23829,7 +23846,7 @@ OMPClause *SemaOpenMP::ActOnOpenMPInclusiveClause(ArrayRef<Expr *> VarList, SourceLocation EndLoc) { SmallVector<Expr *, 8> Vars; for (Expr *RefExpr : VarList) { - assert(RefExpr && "NULL expr in OpenMP nontemporal clause."); + assert(RefExpr && "NULL expr in OpenMP inclusive clause."); SourceLocation ELoc; SourceRange ERange; Expr *SimpleRefExpr = RefExpr; @@ -23870,7 +23887,7 @@ OMPClause *SemaOpenMP::ActOnOpenMPExclusiveClause(ArrayRef<Expr *> VarList, SourceLocation EndLoc) { SmallVector<Expr *, 8> Vars; for (Expr *RefExpr : VarList) { - assert(RefExpr && "NULL expr in OpenMP nontemporal clause."); + assert(RefExpr && "NULL expr in OpenMP exclusive clause."); SourceLocation ELoc; SourceRange ERange; Expr *SimpleRefExpr = RefExpr; @@ -24063,7 +24080,7 @@ OMPClause *SemaOpenMP::ActOnOpenMPAffinityClause( SourceLocation EndLoc, Expr *Modifier, ArrayRef<Expr *> Locators) { SmallVector<Expr *, 8> Vars; for (Expr *RefExpr : Locators) { - assert(RefExpr && "NULL expr in OpenMP shared clause."); + assert(RefExpr && "NULL expr in OpenMP affinity clause."); if (isa<DependentScopeDeclRefExpr>(RefExpr) || RefExpr->isTypeDependent()) { // It will be analyzed later. Vars.push_back(RefExpr); @@ -24375,7 +24392,7 @@ ExprResult SemaOpenMP::ActOnOMPArraySectionExpr( return ExprError(); } } - } else if (ColonLocFirst.isValid() && + } else if (SemaRef.getLangOpts().OpenMP < 60 && ColonLocFirst.isValid() && (OriginalTy.isNull() || (!OriginalTy->isConstantArrayType() && !OriginalTy->isVariableArrayType()))) { // OpenMP 5.0, [2.1.5 Array Sections] diff --git a/clang/lib/Sema/SemaRISCV.cpp b/clang/lib/Sema/SemaRISCV.cpp index 43f7992..994cd07 100644 --- a/clang/lib/Sema/SemaRISCV.cpp +++ b/clang/lib/Sema/SemaRISCV.cpp @@ -1635,6 +1635,116 @@ bool SemaRISCV::isValidFMVExtension(StringRef Ext) { return -1 != RISCVISAInfo::getRISCVFeaturesBitsInfo(Ext).second; } +bool SemaRISCV::checkTargetVersionAttr(const StringRef Param, + const SourceLocation Loc) { + using namespace DiagAttrParams; + + llvm::SmallVector<StringRef, 8> AttrStrs; + Param.split(AttrStrs, ';'); + + bool HasArch = false; + bool HasPriority = false; + bool HasDefault = false; + bool DuplicateAttr = false; + for (StringRef AttrStr : AttrStrs) { + AttrStr = AttrStr.trim(); + // Only support arch=+ext,... syntax. + if (AttrStr.starts_with("arch=+")) { + DuplicateAttr = HasArch; + HasArch = true; + ParsedTargetAttr TargetAttr = + getASTContext().getTargetInfo().parseTargetAttr(AttrStr); + + if (TargetAttr.Features.empty() || + llvm::any_of(TargetAttr.Features, [&](const StringRef Ext) { + return !isValidFMVExtension(Ext); + })) + return Diag(Loc, diag::warn_unsupported_target_attribute) + << Unsupported << None << AttrStr << TargetVersion; + } else if (AttrStr == "default") { + DuplicateAttr = HasDefault; + HasDefault = true; + } else if (AttrStr.consume_front("priority=")) { + DuplicateAttr = HasPriority; + HasPriority = true; + unsigned Digit; + if (AttrStr.getAsInteger(0, Digit)) + return Diag(Loc, diag::warn_unsupported_target_attribute) + << Unsupported << None << AttrStr << TargetVersion; + } else { + return Diag(Loc, diag::warn_unsupported_target_attribute) + << Unsupported << None << AttrStr << TargetVersion; + } + } + + if (((HasPriority || HasArch) && HasDefault) || DuplicateAttr || + (HasPriority && !HasArch)) + return Diag(Loc, diag::warn_unsupported_target_attribute) + << Unsupported << None << Param << TargetVersion; + + return false; +} + +bool SemaRISCV::checkTargetClonesAttr( + SmallVectorImpl<StringRef> &Params, SmallVectorImpl<SourceLocation> &Locs, + SmallVectorImpl<SmallString<64>> &NewParams) { + using namespace DiagAttrParams; + + assert(Params.size() == Locs.size() && + "Mismatch between number of string parameters and locations"); + + bool HasDefault = false; + for (unsigned I = 0, E = Params.size(); I < E; ++I) { + const StringRef Param = Params[I].trim(); + const SourceLocation &Loc = Locs[I]; + + llvm::SmallVector<StringRef, 8> AttrStrs; + Param.split(AttrStrs, ';'); + + bool IsPriority = false; + bool IsDefault = false; + for (StringRef AttrStr : AttrStrs) { + AttrStr = AttrStr.trim(); + // Only support arch=+ext,... syntax. + if (AttrStr.starts_with("arch=+")) { + ParsedTargetAttr TargetAttr = + getASTContext().getTargetInfo().parseTargetAttr(AttrStr); + + if (TargetAttr.Features.empty() || + llvm::any_of(TargetAttr.Features, [&](const StringRef Ext) { + return !isValidFMVExtension(Ext); + })) + return Diag(Loc, diag::warn_unsupported_target_attribute) + << Unsupported << None << Param << TargetClones; + } else if (AttrStr == "default") { + IsDefault = true; + HasDefault = true; + } else if (AttrStr.consume_front("priority=")) { + IsPriority = true; + unsigned Digit; + if (AttrStr.getAsInteger(0, Digit)) + return Diag(Loc, diag::warn_unsupported_target_attribute) + << Unsupported << None << Param << TargetClones; + } else { + return Diag(Loc, diag::warn_unsupported_target_attribute) + << Unsupported << None << Param << TargetClones; + } + } + + if (IsPriority && IsDefault) + return Diag(Loc, diag::warn_unsupported_target_attribute) + << Unsupported << None << Param << TargetClones; + + if (llvm::is_contained(NewParams, Param)) + Diag(Loc, diag::warn_target_clone_duplicate_options); + NewParams.push_back(Param); + } + if (!HasDefault) + return Diag(Locs[0], diag::err_target_clone_must_have_default); + + return false; +} + SemaRISCV::SemaRISCV(Sema &S) : SemaBase(S) {} } // namespace clang diff --git a/clang/lib/Sema/SemaX86.cpp b/clang/lib/Sema/SemaX86.cpp index 5c149bd..850bcb1 100644 --- a/clang/lib/Sema/SemaX86.cpp +++ b/clang/lib/Sema/SemaX86.cpp @@ -954,6 +954,11 @@ bool SemaX86::CheckBuiltinFunctionCall(const TargetInfo &TI, unsigned BuiltinID, l = 0; u = 15; break; + case X86::BI__builtin_ia32_prefetchi: + i = 1; + l = 2; // _MM_HINT_T1 + u = 3; // _MM_HINT_T0 + break; } // Note that we don't force a hard error on the range check here, allowing @@ -1056,4 +1061,61 @@ void SemaX86::handleForceAlignArgPointerAttr(Decl *D, const ParsedAttr &AL) { X86ForceAlignArgPointerAttr(getASTContext(), AL)); } +bool SemaX86::checkTargetClonesAttr( + SmallVectorImpl<StringRef> &Params, SmallVectorImpl<SourceLocation> &Locs, + SmallVectorImpl<SmallString<64>> &NewParams) { + using namespace DiagAttrParams; + + assert(Params.size() == Locs.size() && + "Mismatch between number of string parameters and locations"); + + bool HasDefault = false; + bool HasComma = false; + for (unsigned I = 0, E = Params.size(); I < E; ++I) { + const StringRef Param = Params[I].trim(); + const SourceLocation &Loc = Locs[I]; + + if (Param.empty() || Param.ends_with(',')) + return Diag(Loc, diag::warn_unsupported_target_attribute) + << Unsupported << None << "" << TargetClones; + + if (Param.contains(',')) + HasComma = true; + + StringRef LHS; + StringRef RHS = Param; + do { + std::tie(LHS, RHS) = RHS.split(','); + LHS = LHS.trim(); + const SourceLocation &CurLoc = + Loc.getLocWithOffset(LHS.data() - Param.data()); + + if (LHS.starts_with("arch=")) { + if (!getASTContext().getTargetInfo().isValidCPUName( + LHS.drop_front(sizeof("arch=") - 1))) + return Diag(CurLoc, diag::warn_unsupported_target_attribute) + << Unsupported << CPU << LHS.drop_front(sizeof("arch=") - 1) + << TargetClones; + } else if (LHS == "default") + HasDefault = true; + else if (!getASTContext().getTargetInfo().isValidFeatureName(LHS) || + getASTContext().getTargetInfo().getFMVPriority(LHS) == 0) + return Diag(CurLoc, diag::warn_unsupported_target_attribute) + << Unsupported << None << LHS << TargetClones; + + if (llvm::is_contained(NewParams, LHS)) + Diag(CurLoc, diag::warn_target_clone_duplicate_options); + // Note: Add even if there are duplicates, since it changes name mangling. + NewParams.push_back(LHS); + } while (!RHS.empty()); + } + if (HasComma && Params.size() > 1) + Diag(Locs[0], diag::warn_target_clone_mixed_values); + + if (!HasDefault) + return Diag(Locs[0], diag::err_target_clone_must_have_default); + + return false; +} + } // namespace clang diff --git a/clang/lib/Tooling/DependencyScanning/DependencyScanningWorker.cpp b/clang/lib/Tooling/DependencyScanning/DependencyScanningWorker.cpp index 9bd8547..8ce2706 100644 --- a/clang/lib/Tooling/DependencyScanning/DependencyScanningWorker.cpp +++ b/clang/lib/Tooling/DependencyScanning/DependencyScanningWorker.cpp @@ -24,7 +24,6 @@ #include "clang/Tooling/DependencyScanning/DependencyScanningService.h" #include "clang/Tooling/DependencyScanning/InProcessModuleCache.h" #include "clang/Tooling/DependencyScanning/ModuleDepCollector.h" -#include "clang/Tooling/Tooling.h" #include "llvm/ADT/IntrusiveRefCntPtr.h" #include "llvm/Support/Allocator.h" #include "llvm/Support/Error.h" @@ -376,25 +375,23 @@ public: /// A clang tool that runs the preprocessor in a mode that's optimized for /// dependency scanning for the given compiler invocation. -class DependencyScanningAction : public tooling::ToolAction { +class DependencyScanningAction { public: DependencyScanningAction( DependencyScanningService &Service, StringRef WorkingDirectory, DependencyConsumer &Consumer, DependencyActionController &Controller, llvm::IntrusiveRefCntPtr<DependencyScanningWorkerFilesystem> DepFS, - bool DisableFree, std::optional<StringRef> ModuleName = std::nullopt) + std::optional<StringRef> ModuleName = std::nullopt) : Service(Service), WorkingDirectory(WorkingDirectory), Consumer(Consumer), Controller(Controller), DepFS(std::move(DepFS)), - DisableFree(DisableFree), ModuleName(ModuleName) {} + ModuleName(ModuleName) {} bool runInvocation(std::shared_ptr<CompilerInvocation> Invocation, - FileManager *DriverFileMgr, + IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS, std::shared_ptr<PCHContainerOperations> PCHContainerOps, - DiagnosticConsumer *DiagConsumer) override { + DiagnosticConsumer *DiagConsumer) { // Make a deep copy of the original Clang invocation. CompilerInvocation OriginalInvocation(*Invocation); - // Restore the value of DisableFree, which may be modified by Tooling. - OriginalInvocation.getFrontendOpts().DisableFree = DisableFree; if (any(Service.getOptimizeArgs() & ScanningOptimizations::Macros)) canonicalizeDefines(OriginalInvocation.getPreprocessorOpts()); @@ -419,8 +416,8 @@ public: // Create the compiler's actual diagnostics engine. sanitizeDiagOpts(ScanInstance.getDiagnosticOpts()); assert(!DiagConsumerFinished && "attempt to reuse finished consumer"); - ScanInstance.createDiagnostics(DriverFileMgr->getVirtualFileSystem(), - DiagConsumer, /*ShouldOwnClient=*/false); + ScanInstance.createDiagnostics(*FS, DiagConsumer, + /*ShouldOwnClient=*/false); if (!ScanInstance.hasDiagnostics()) return false; @@ -431,6 +428,7 @@ public: ScanInstance.getHeaderSearchOpts().BuildSessionTimestamp = Service.getBuildSessionTimestamp(); + ScanInstance.getFrontendOpts().DisableFree = false; ScanInstance.getFrontendOpts().GenerateGlobalModuleIndex = false; ScanInstance.getFrontendOpts().UseGlobalModuleIndex = false; // This will prevent us compiling individual modules asynchronously since @@ -441,9 +439,9 @@ public: any(Service.getOptimizeArgs() & ScanningOptimizations::VFS); // Support for virtual file system overlays. - auto FS = createVFSFromCompilerInvocation( - ScanInstance.getInvocation(), ScanInstance.getDiagnostics(), - DriverFileMgr->getVirtualFileSystemPtr()); + FS = createVFSFromCompilerInvocation(ScanInstance.getInvocation(), + ScanInstance.getDiagnostics(), + std::move(FS)); // Create a new FileManager to match the invocation's FileSystemOptions. auto *FileMgr = ScanInstance.createFileManager(FS); @@ -554,9 +552,6 @@ public: if (Result) setLastCC1Arguments(std::move(OriginalInvocation)); - // Propagate the statistics to the parent FileManager. - DriverFileMgr->AddStats(ScanInstance.getFileManager()); - return Result; } @@ -584,7 +579,6 @@ private: DependencyConsumer &Consumer; DependencyActionController &Controller; llvm::IntrusiveRefCntPtr<DependencyScanningWorkerFilesystem> DepFS; - bool DisableFree; std::optional<StringRef> ModuleName; std::optional<CompilerInstance> ScanInstanceStorage; std::shared_ptr<ModuleDepCollector> MDC; @@ -669,15 +663,14 @@ llvm::Error DependencyScanningWorker::computeDependencies( } static bool forEachDriverJob( - ArrayRef<std::string> ArgStrs, DiagnosticsEngine &Diags, FileManager &FM, + ArrayRef<std::string> ArgStrs, DiagnosticsEngine &Diags, + IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS, llvm::function_ref<bool(const driver::Command &Cmd)> Callback) { SmallVector<const char *, 256> Argv; Argv.reserve(ArgStrs.size()); for (const std::string &Arg : ArgStrs) Argv.push_back(Arg.c_str()); - llvm::vfs::FileSystem *FS = &FM.getVirtualFileSystem(); - std::unique_ptr<driver::Driver> Driver = std::make_unique<driver::Driver>( Argv[0], llvm::sys::getDefaultTargetTriple(), Diags, "clang LLVM compiler", FS); @@ -687,7 +680,8 @@ static bool forEachDriverJob( bool CLMode = driver::IsClangCL( driver::getDriverMode(Argv[0], ArrayRef(Argv).slice(1))); - if (llvm::Error E = driver::expandResponseFiles(Argv, CLMode, Alloc, FS)) { + if (llvm::Error E = + driver::expandResponseFiles(Argv, CLMode, Alloc, FS.get())) { Diags.Report(diag::err_drv_expand_response_file) << llvm::toString(std::move(E)); return false; @@ -710,17 +704,25 @@ static bool forEachDriverJob( static bool createAndRunToolInvocation( std::vector<std::string> CommandLine, DependencyScanningAction &Action, - FileManager &FM, + IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS, std::shared_ptr<clang::PCHContainerOperations> &PCHContainerOps, DiagnosticsEngine &Diags, DependencyConsumer &Consumer) { // Save executable path before providing CommandLine to ToolInvocation std::string Executable = CommandLine[0]; - ToolInvocation Invocation(std::move(CommandLine), &Action, &FM, - PCHContainerOps); - Invocation.setDiagnosticConsumer(Diags.getClient()); - Invocation.setDiagnosticOptions(&Diags.getDiagnosticOptions()); - if (!Invocation.run()) + + llvm::opt::ArgStringList Argv; + for (const std::string &Str : ArrayRef(CommandLine).drop_front()) + Argv.push_back(Str.c_str()); + + auto Invocation = std::make_shared<CompilerInvocation>(); + if (!CompilerInvocation::CreateFromArgs(*Invocation, Argv, Diags)) { + // FIXME: Should we just go on like cc1_main does? + return false; + } + + if (!Action.runInvocation(std::move(Invocation), std::move(FS), + PCHContainerOps, Diags.getClient())) return false; std::vector<std::string> Args = Action.takeLastCC1Arguments(); @@ -733,37 +735,24 @@ bool DependencyScanningWorker::scanDependencies( DependencyConsumer &Consumer, DependencyActionController &Controller, DiagnosticConsumer &DC, llvm::IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS, std::optional<StringRef> ModuleName) { - auto FileMgr = - llvm::makeIntrusiveRefCnt<FileManager>(FileSystemOptions{}, FS); - std::vector<const char *> CCommandLine(CommandLine.size(), nullptr); llvm::transform(CommandLine, CCommandLine.begin(), [](const std::string &Str) { return Str.c_str(); }); auto DiagOpts = CreateAndPopulateDiagOpts(CCommandLine); sanitizeDiagOpts(*DiagOpts); - IntrusiveRefCntPtr<DiagnosticsEngine> Diags = - CompilerInstance::createDiagnostics(FileMgr->getVirtualFileSystem(), - *DiagOpts, &DC, - /*ShouldOwnClient=*/false); - - // Although `Diagnostics` are used only for command-line parsing, the - // custom `DiagConsumer` might expect a `SourceManager` to be present. - SourceManager SrcMgr(*Diags, *FileMgr); - Diags->setSourceManager(&SrcMgr); - // DisableFree is modified by Tooling for running - // in-process; preserve the original value, which is - // always true for a driver invocation. - bool DisableFree = true; + auto Diags = CompilerInstance::createDiagnostics(*FS, *DiagOpts, &DC, + /*ShouldOwnClient=*/false); + DependencyScanningAction Action(Service, WorkingDirectory, Consumer, - Controller, DepFS, DisableFree, ModuleName); + Controller, DepFS, ModuleName); bool Success = false; if (CommandLine[1] == "-cc1") { - Success = createAndRunToolInvocation(CommandLine, Action, *FileMgr, + Success = createAndRunToolInvocation(CommandLine, Action, FS, PCHContainerOps, *Diags, Consumer); } else { Success = forEachDriverJob( - CommandLine, *Diags, *FileMgr, [&](const driver::Command &Cmd) { + CommandLine, *Diags, FS, [&](const driver::Command &Cmd) { if (StringRef(Cmd.getCreator().getName()) != "clang") { // Non-clang command. Just pass through to the dependency // consumer. @@ -782,7 +771,7 @@ bool DependencyScanningWorker::scanDependencies( // system to ensure that any file system requests that // are made by the driver do not go through the // dependency scanning filesystem. - return createAndRunToolInvocation(std::move(Argv), Action, *FileMgr, + return createAndRunToolInvocation(std::move(Argv), Action, FS, PCHContainerOps, *Diags, Consumer); }); } diff --git a/clang/lib/Tooling/Inclusions/HeaderIncludes.cpp b/clang/lib/Tooling/Inclusions/HeaderIncludes.cpp index 9f10ee1..2b5a293 100644 --- a/clang/lib/Tooling/Inclusions/HeaderIncludes.cpp +++ b/clang/lib/Tooling/Inclusions/HeaderIncludes.cpp @@ -285,8 +285,7 @@ HeaderIncludes::HeaderIncludes(StringRef FileName, StringRef Code, MaxInsertOffset(MinInsertOffset + getMaxHeaderInsertionOffset( FileName, Code.drop_front(MinInsertOffset), Style)), - MainIncludeFound(false), - Categories(Style, FileName) { + MainIncludeFound(false), Categories(Style, FileName) { // Add 0 for main header and INT_MAX for headers that are not in any // category. Priorities = {0, INT_MAX}; diff --git a/clang/lib/Tooling/Inclusions/Stdlib/StandardLibrary.cpp b/clang/lib/Tooling/Inclusions/Stdlib/StandardLibrary.cpp index b88e6db..807a8d8 100644 --- a/clang/lib/Tooling/Inclusions/Stdlib/StandardLibrary.cpp +++ b/clang/lib/Tooling/Inclusions/Stdlib/StandardLibrary.cpp @@ -131,7 +131,7 @@ static int initialize(Lang Language) { Mapping->SymbolNames[SymIndex] = { QName.data(), NSLen, static_cast<unsigned int>(QName.size() - NSLen)}; if (!HeaderName.empty()) - Mapping->SymbolHeaderIDs[SymIndex].push_back(AddHeader(HeaderName)); + Mapping->SymbolHeaderIDs[SymIndex].push_back(AddHeader(HeaderName)); NSSymbolMap &NSSymbols = AddNS(QName.take_front(NSLen)); NSSymbols.try_emplace(QName.drop_front(NSLen), SymIndex); @@ -236,7 +236,7 @@ std::optional<Symbol> Symbol::named(llvm::StringRef Scope, llvm::StringRef Name, return std::nullopt; } std::optional<Header> Symbol::header() const { - const auto& Headers = getMappingPerLang(Language)->SymbolHeaderIDs[ID]; + const auto &Headers = getMappingPerLang(Language)->SymbolHeaderIDs[ID]; if (Headers.empty()) return std::nullopt; return Header(Headers.front(), Language); diff --git a/clang/test/APINotes/Inputs/Headers/SwiftImportAs.apinotes b/clang/test/APINotes/Inputs/Headers/SwiftImportAs.apinotes index 66fc46e..c096822 100644 --- a/clang/test/APINotes/Inputs/Headers/SwiftImportAs.apinotes +++ b/clang/test/APINotes/Inputs/Headers/SwiftImportAs.apinotes @@ -19,6 +19,10 @@ Tags: SwiftReleaseOp: release SwiftRetainOp: retain SwiftDefaultOwnership: unretained +- Name: OpaqueRefCountedType + SwiftImportAs: reference + SwiftReleaseOp: ORCRelease + SwiftRetainOp: ORCRetain - Name: NonCopyableType SwiftCopyable: false SwiftConformsTo: MySwiftModule.MySwiftNonCopyableProtocol diff --git a/clang/test/APINotes/Inputs/Headers/SwiftImportAs.h b/clang/test/APINotes/Inputs/Headers/SwiftImportAs.h index 20b8f04..5f817ac 100644 --- a/clang/test/APINotes/Inputs/Headers/SwiftImportAs.h +++ b/clang/test/APINotes/Inputs/Headers/SwiftImportAs.h @@ -23,3 +23,9 @@ struct EscapableType { int value; }; struct RefCountedTypeWithDefaultConvention {}; inline void retain(RefCountedType *x) {} inline void release(RefCountedType *x) {} + +struct OpaqueRefCountedType; +struct OpaqueRefCountedType; // redeclaration + +inline void ORCRetain(struct OpaqueRefCountedType *x); +inline void ORCRelease(struct OpaqueRefCountedType *x); diff --git a/clang/test/APINotes/swift-import-as.cpp b/clang/test/APINotes/swift-import-as.cpp index 929f924..179170f 100644 --- a/clang/test/APINotes/swift-import-as.cpp +++ b/clang/test/APINotes/swift-import-as.cpp @@ -3,6 +3,7 @@ // RUN: %clang_cc1 -fmodules -fblocks -fimplicit-module-maps -fmodules-cache-path=%t/ModulesCache -fdisable-module-hash -fapinotes-modules -I %S/Inputs/Headers %s -x c++ -ast-dump -ast-dump-filter ImmortalRefType | FileCheck -check-prefix=CHECK-IMMORTAL %s // RUN: %clang_cc1 -fmodules -fblocks -fimplicit-module-maps -fmodules-cache-path=%t/ModulesCache -fdisable-module-hash -fapinotes-modules -I %S/Inputs/Headers %s -x c++ -ast-dump -ast-dump-filter RefCountedType | FileCheck -check-prefix=CHECK-REF-COUNTED %s // RUN: %clang_cc1 -fmodules -fblocks -fimplicit-module-maps -fmodules-cache-path=%t/ModulesCache -fdisable-module-hash -fapinotes-modules -I %S/Inputs/Headers %s -x c++ -ast-dump -ast-dump-filter RefCountedTypeWithDefaultConvention | FileCheck -check-prefix=CHECK-REF-COUNTED-DEFAULT %s +// RUN: %clang_cc1 -fmodules -fblocks -fimplicit-module-maps -fmodules-cache-path=%t/ModulesCache -fdisable-module-hash -fapinotes-modules -I %S/Inputs/Headers %s -x c++ -ast-dump -ast-dump-filter OpaqueRefCountedType | FileCheck -check-prefix=CHECK-OPAQUE-REF-COUNTED %s // RUN: %clang_cc1 -fmodules -fblocks -fimplicit-module-maps -fmodules-cache-path=%t/ModulesCache -fdisable-module-hash -fapinotes-modules -I %S/Inputs/Headers %s -x c++ -ast-dump -ast-dump-filter NonCopyableType | FileCheck -check-prefix=CHECK-NON-COPYABLE %s // RUN: %clang_cc1 -fmodules -fblocks -fimplicit-module-maps -fmodules-cache-path=%t/ModulesCache -fdisable-module-hash -fapinotes-modules -I %S/Inputs/Headers %s -x c++ -ast-dump -ast-dump-filter CopyableType | FileCheck -check-prefix=CHECK-COPYABLE %s // RUN: %clang_cc1 -fmodules -fblocks -fimplicit-module-maps -fmodules-cache-path=%t/ModulesCache -fdisable-module-hash -fapinotes-modules -I %S/Inputs/Headers %s -x c++ -ast-dump -ast-dump-filter NonEscapableType | FileCheck -check-prefix=CHECK-NON-ESCAPABLE %s @@ -34,6 +35,20 @@ // CHECK-REF-COUNTED-DEFAULT: SwiftAttrAttr {{.+}} <<invalid sloc>> "release:release" // CHECK-REF-COUNTED-DEFAULT: SwiftAttrAttr {{.+}} <<invalid sloc>> "returned_as_unretained_by_default" +// CHECK-OPAQUE-REF-COUNTED: Dumping OpaqueRefCountedType: +// CHECK-OPAQUE-REF-COUNTED-NEXT: CXXRecordDecl {{.+}} imported in SwiftImportAs{{.*}}struct OpaqueRefCountedType +// CHECK-OPAQUE-REF-COUNTED: SwiftAttrAttr {{.+}} <<invalid sloc>> "import_reference" +// CHECK-OPAQUE-REF-COUNTED: SwiftAttrAttr {{.+}} <<invalid sloc>> "retain:ORCRetain" +// CHECK-OPAQUE-REF-COUNTED: SwiftAttrAttr {{.+}} <<invalid sloc>> "release:ORCRelease" +// CHECK-OPAQUE-REF-COUNTED-NOT: SwiftAttrAttr {{.+}} <<invalid sloc>> "release:ORCRelease" + +// CHECK-OPAQUE-REF-COUNTED: Dumping OpaqueRefCountedType: +// CHECK-OPAQUE-REF-COUNTED-NEXT: CXXRecordDecl {{.+}} imported in SwiftImportAs{{.*}}struct OpaqueRefCountedType +// CHECK-OPAQUE-REF-COUNTED: SwiftAttrAttr {{.+}} <<invalid sloc>> "import_reference" +// CHECK-OPAQUE-REF-COUNTED: SwiftAttrAttr {{.+}} <<invalid sloc>> "retain:ORCRetain" +// CHECK-OPAQUE-REF-COUNTED: SwiftAttrAttr {{.+}} <<invalid sloc>> "release:ORCRelease" + +// CHECK-OPAQUE-REF-COUNTED-NOT: SwiftAttrAttr {{.+}} <<invalid sloc>> "release: // CHECK-NON-COPYABLE: Dumping NonCopyableType: // CHECK-NON-COPYABLE-NEXT: CXXRecordDecl {{.+}} imported in SwiftImportAs {{.+}} struct NonCopyableType // CHECK-NON-COPYABLE: SwiftAttrAttr {{.+}} <<invalid sloc>> "conforms_to:MySwiftModule.MySwiftNonCopyableProtocol" diff --git a/clang/test/AST/ByteCode/builtin-bit-cast.cpp b/clang/test/AST/ByteCode/builtin-bit-cast.cpp index 3c5e89d7..bc356b0 100644 --- a/clang/test/AST/ByteCode/builtin-bit-cast.cpp +++ b/clang/test/AST/ByteCode/builtin-bit-cast.cpp @@ -22,6 +22,10 @@ typedef __INTPTR_TYPE__ intptr_t; static_assert(sizeof(int) == 4); static_assert(sizeof(long long) == 8); + +constexpr bool test_bad_bool = __builtin_bit_cast(bool, (char)0xff); // both-error {{must be initialized by a constant expression}} \ + // both-note {{value 255 cannot be represented in type 'bool'}} + template <class To, class From> constexpr To bit_cast(const From &from) { static_assert(sizeof(To) == sizeof(From)); diff --git a/clang/test/AST/ByteCode/unions.cpp b/clang/test/AST/ByteCode/unions.cpp index 7cfd0d6..139e318 100644 --- a/clang/test/AST/ByteCode/unions.cpp +++ b/clang/test/AST/ByteCode/unions.cpp @@ -79,10 +79,9 @@ namespace DefaultInit { constexpr U1 u1; /// OK. - constexpr int foo() { // expected-error {{never produces a constant expression}} + constexpr int foo() { U1 u; - return u.a; // both-note {{read of member 'a' of union with active member 'b'}} \ - // expected-note {{read of member 'a' of union with active member 'b'}} + return u.a; // both-note {{read of member 'a' of union with active member 'b'}} } static_assert(foo() == 42); // both-error {{not an integral constant expression}} \ // both-note {{in call to}} @@ -861,6 +860,73 @@ namespace CopyCtorMutable { // both-note {{in call}} } + +namespace NonTrivialCtor { + struct A { int x = 1; constexpr int f() { return 1; } }; + struct B : A { int y = 1; constexpr int g() { return 2; } }; + struct C { + int x; + constexpr virtual int f() = 0; + }; + struct D : C { + int y; + constexpr virtual int f() override { return 3; } + }; + + union U { + int n; + B b; + D d; + }; + + consteval int test(int which) { + if (which == 0) {} + + U u{.n = 5}; + assert_active(u); + assert_active(u.n); + assert_inactive(u.b); + + switch (which) { + case 0: + u.b.x = 10; // both-note {{assignment to member 'b' of union with active member 'n'}} + return u.b.f(); + case 1: + u.b.y = 10; // both-note {{assignment to member 'b' of union with active member 'n'}} + return u.b.g(); + case 2: + u.d.x = 10; // both-note {{assignment to member 'd' of union with active member 'n'}} + return u.d.f(); + case 3: + u.d.y = 10; // both-note {{assignment to member 'd' of union with active member 'n'}} + return u.d.f(); + } + + return 1; + } + static_assert(test(0)); // both-error {{not an integral constant expression}} \ + // both-note {{in call}} + static_assert(test(1)); // both-error {{not an integral constant expression}} \ + // both-note {{in call}} + static_assert(test(2)); // both-error {{not an integral constant expression}} \ + // both-note {{in call}} + static_assert(test(3)); // both-error {{not an integral constant expression}} \ + // both-note {{in call}} + +} + +namespace PrimitiveFieldInitActivates { + /// The initializer of a needs the field to be active _before_ it's visited. + template<int> struct X {}; + union V { + int a, b; + constexpr V(X<0>) : a(a = 1) {} // ok + constexpr V(X<2>) : a() { b = 1; } // ok + }; + constinit V v0 = X<0>(); + constinit V v2 = X<2>(); +} + #endif namespace AddressComparison { diff --git a/clang/test/AST/ast-dump-APValue-lvalue.cpp b/clang/test/AST/ast-dump-APValue-lvalue.cpp index 51d22a5..f4cf2f5 100644 --- a/clang/test/AST/ast-dump-APValue-lvalue.cpp +++ b/clang/test/AST/ast-dump-APValue-lvalue.cpp @@ -67,6 +67,10 @@ void Test(int (&arr)[10]) { // CHECK-NEXT: | |-value: LValue Base=TypeInfoLValue typeid(int), Null=0, Offset=0, HasPath=1, PathLength=0, Path=() constexpr int(MP::*pmi) = (int MP::*)&P::x; - // CHECK: `-VarDecl {{.*}} <col:{{.*}}, col:{{.*}}> col:{{.*}} pmi 'int (MP::*const)' constexpr cinit - // CHECK-NEXT: |-value: MemberPointer MP::x + // CHECK: | `-VarDecl {{.*}} <col:{{.*}}, col:{{.*}}> col:{{.*}} pmi 'int (MP::*const)' constexpr cinit + // CHECK-NEXT: | |-value: MemberPointer MP::x + + constexpr int(MP::*pmn) = (int MP::*)nullptr; + // CHECK: `-VarDecl {{.*}} <col:{{.*}}, col:{{.*}}> col:{{.*}} pmn 'int (MP::*const)' constexpr cinit + // CHECK-NEXT: |-value: MemberPointer null } diff --git a/clang/test/CIR/CodeGen/bitfields.c b/clang/test/CIR/CodeGen/bitfields.c index a73c076..869a7c9 100644 --- a/clang/test/CIR/CodeGen/bitfields.c +++ b/clang/test/CIR/CodeGen/bitfields.c @@ -315,3 +315,51 @@ void unOp(S* s) { // OGCG: [[TMP12:%.*]] = shl i64 [[TMP8]], 62 // OGCG: [[TMP13:%.*]] = ashr i64 [[TMP12]], 62 // OGCG: [[TMP14:%.*]] = trunc i64 [[TMP13]] to i32 + +void binOp(S* s) { + s->d |= 42; +} + +// CIR: cir.func {{.*@binOp}} +// CIR: [[TMP0:%.*]] = cir.const #cir.int<42> : !s32i +// CIR: [[TMP1:%.*]] = cir.get_member {{.*}}[0] {name = "d"} : !cir.ptr<!rec_S> -> !cir.ptr<!u64i> +// CIR: [[TMP2:%.*]] = cir.get_bitfield align(4) (#bfi_d, [[TMP1]] : !cir.ptr<!u64i>) -> !s32i +// CIR: [[TMP3:%.*]] = cir.binop(or, [[TMP2]], [[TMP0]]) : !s32i +// CIR: cir.set_bitfield align(4) (#bfi_d, [[TMP1]] : !cir.ptr<!u64i>, [[TMP3]] : !s32i) + +// LLVM: define {{.*@binOp}} +// LLVM: [[TMP0:%.*]] = load ptr, ptr {{.*}}, align 8 +// LLVM: [[TMP1:%.*]] = getelementptr %struct.S, ptr [[TMP0]], i32 0, i32 0 +// LLVM: [[TMP2:%.*]] = load i64, ptr [[TMP1]], align 4 +// LLVM: [[TMP3:%.*]] = shl i64 [[TMP2]], 13 +// LLVM: [[TMP4:%.*]] = ashr i64 [[TMP3]], 62 +// LLVM: [[TMP5:%.*]] = trunc i64 [[TMP4]] to i32 +// LLVM: [[TMP6:%.*]] = or i32 [[TMP5]], 42 +// LLVM: [[TMP7:%.*]] = zext i32 [[TMP6]] to i64 +// LLVM: [[TMP8:%.*]] = load i64, ptr [[TMP1]], align 4 +// LLVM: [[TMP9:%.*]] = and i64 [[TMP7]], 3 +// LLVM: [[TMP10:%.*]] = shl i64 [[TMP9]], 49 +// LLVM: [[TMP11:%.*]] = and i64 [[TMP8]], -1688849860263937 +// LLVM: [[TMP12:%.*]] = or i64 [[TMP11]], [[TMP10]] +// LLVM: store i64 [[TMP12]], ptr [[TMP1]], align 4 +// LLVM: [[TMP13:%.*]] = shl i64 [[TMP9]], 62 +// LLVM: [[TMP14:%.*]] = ashr i64 [[TMP13]], 62 +// LLVM: [[TMP15:%.*]] = trunc i64 [[TMP14]] to i32 + +// OGCG: define {{.*@binOp}} +// OGCG: [[TMP0:%.*]] = load ptr, ptr %s.addr, align 8 +// OGCG: [[TMP1:%.*]] = load i64, ptr [[TMP0]], align 4 +// OGCG: [[TMP2:%.*]] = shl i64 [[TMP1]], 13 +// OGCG: [[TMP3:%.*]] = ashr i64 [[TMP2]], 62 +// OGCG: [[TMP4:%.*]] = trunc i64 [[TMP3]] to i32 +// OGCG: [[TMP5:%.*]] = or i32 [[TMP4]], 42 +// OGCG: [[TMP6:%.*]] = zext i32 [[TMP5]] to i64 +// OGCG: [[TMP7:%.*]] = load i64, ptr [[TMP0]], align 4 +// OGCG: [[TMP8:%.*]] = and i64 [[TMP6]], 3 +// OGCG: [[TMP9:%.*]] = shl i64 [[TMP8]], 49 +// OGCG: [[TMP10:%.*]] = and i64 [[TMP7]], -1688849860263937 +// OGCG: [[TMP11:%.*]] = or i64 [[TMP10]], [[TMP9]] +// OGCG: store i64 [[TMP11]], ptr [[TMP0]], align 4 +// OGCG: [[TMP12:%.*]] = shl i64 [[TMP8]], 62 +// OGCG: [[TMP13:%.*]] = ashr i64 [[TMP12]], 62 +// OGCG: [[TMP14:%.*]] = trunc i64 [[TMP13]] to i32 diff --git a/clang/test/CIR/CodeGen/complex-cast.cpp b/clang/test/CIR/CodeGen/complex-cast.cpp new file mode 100644 index 0000000..94f47e4 --- /dev/null +++ b/clang/test/CIR/CodeGen/complex-cast.cpp @@ -0,0 +1,328 @@ +// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -fclangir -emit-cir -mmlir --mlir-print-ir-before=cir-canonicalize -o %t.cir %s 2>&1 | FileCheck --check-prefix=CIR-BEFORE %s +// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -fclangir -emit-cir -mmlir --mlir-print-ir-after=cir-lowering-prepare -o %t.cir %s 2>&1 | FileCheck --check-prefixes=CIR-AFTER %s +// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -Wno-unused-value -fclangir -emit-llvm %s -o %t-cir.ll +// RUN: FileCheck --input-file=%t-cir.ll %s -check-prefix=LLVM +// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -Wno-unused-value -emit-llvm %s -o %t.ll +// RUN: FileCheck --input-file=%t.ll %s -check-prefix=OGCG + +double _Complex cd; +float _Complex cf; +int _Complex ci; +short _Complex cs; +double sd; +int si; +bool b; + +void scalar_to_complex() { + cd = sd; + ci = si; + cd = si; + ci = sd; +} + +// CIR-BEFORE: %[[FP_TO_COMPLEX:.*]] = cir.cast(float_to_complex, %{{.*}} : !cir.double), !cir.complex<!cir.double> + +// CIR-AFTER: %[[REAL:.*]] = cir.load{{.*}} %{{.*}} : !cir.ptr<!cir.double>, !cir.double +// CIR-AFTER-NEXT: %[[IMAG:.*]] = cir.const #cir.fp<0.000000e+00> : !cir.double +// CIR-AFTER-NEXT: %{{.*}} = cir.complex.create %[[REAL]], %[[IMAG]] : !cir.double -> !cir.complex<!cir.double> + +// LLVM: %[[REAL:.*]] = load double, ptr {{.*}}, align 8 +// LLVM-NEXT: %[[TMP:.*]] = insertvalue { double, double } undef, double %[[REAL]], 0 +// LLVM-NEXT: %[[COMPLEX:.*]] = insertvalue { double, double } %[[TMP]], double 0.000000e+00, 1 +// LLVM-NEXT: store { double, double } %[[COMPLEX]], ptr {{.*}}, align 8 + +// OGCG: %[[REAL:.*]] = load double, ptr {{.*}}, align 8 +// OGCG: store double %[[REAL]], ptr {{.*}}, align 8 +// OGCG: store double 0.000000e+00, ptr getelementptr inbounds nuw ({ double, double }, ptr @cd, i32 0, i32 1), align 8 + +// CIR-BEFORE: %[[INT_TO_COMPLEX:.*]] = cir.cast(int_to_complex, %{{.*}} : !s32i), !cir.complex<!s32i> + +// CIR-AFTER: %[[REAL:.*]] = cir.load{{.*}} %{{.*}} : !cir.ptr<!s32i>, !s32i +// CIR-AFTER-NEXT: %[[IMAG:.*]] = cir.const #cir.int<0> : !s32i +// CIR-AFTER-NEXT: %{{.*}} = cir.complex.create %[[REAL]], %[[IMAG]] : !s32i -> !cir.complex<!s32i> + +// LLVM: %[[REAL:.*]] = load i32, ptr {{.*}}, align 4 +// LLVM-NEXT: %[[TMP:.*]] = insertvalue { i32, i32 } undef, i32 %[[REAL]], 0 +// LLVM-NEXT: %[[COMPLEX:.*]] = insertvalue { i32, i32 } %[[TMP]], i32 0, 1 +// LLVM-NEXT: store { i32, i32 } %[[COMPLEX]], ptr {{.*}}, align 4 + +// OGCG: %[[REAL:.*]] = load i32, ptr {{.*}}, align 4 +// OGCG: store i32 %[[REAL]], ptr {{.*}}, align 4 +// OGCG: store i32 0, ptr getelementptr inbounds nuw ({ i32, i32 }, ptr @ci, i32 0, i32 1), align 4 + +// CIR-BEFORE: %[[INT_TO_FP:.*]] = cir.cast(int_to_float, %{{.*}} : !s32i), !cir.double +// CIR-BEFORE: %[[FP_TO_COMPLEX:.*]] = cir.cast(float_to_complex, %[[INT_TO_FP]] : !cir.double), !cir.complex<!cir.double> + +// CIR-AFTER: %[[TMP:.*]] = cir.load{{.*}} %{{.*}} : !cir.ptr<!s32i>, !s32i +// CIR-AFTER-NEXT: %[[REAL:.*]] = cir.cast(int_to_float, %[[TMP]] : !s32i), !cir.double +// CIR-AFTER-NEXT: %[[IMAG:.*]] = cir.const #cir.fp<0.000000e+00> : !cir.double +// CIR-AFTER-NEXT: %{{.*}} = cir.complex.create %[[REAL]], %[[IMAG]] : !cir.double -> !cir.complex<!cir.double> + +// LLVM: %[[TMP:.*]] = load i32, ptr {{.*}}, align 4 +// LLVM-NEXT: %[[REAL:.*]] = sitofp i32 %[[TMP]] to double +// LLVM-NEXT: %[[TMP_2:.*]] = insertvalue { double, double } undef, double %[[REAL]], 0 +// LLVM-NEXT: %[[COMPLEX:.*]] = insertvalue { double, double } %[[TMP_2]], double 0.000000e+00, 1 +// LLVM-NEXT: store { double, double } %[[COMPLEX]], ptr {{.*}}, align 8 + +// OGCG: %[[TMP:.*]] = load i32, ptr {{.*}}, align 4 +// OGCG: %[[REAL:.*]] = sitofp i32 %[[TMP]] to double +// OGCG: store double %[[REAL]], ptr {{.*}}, align 8 +// OGCG: store double 0.000000e+00, ptr getelementptr inbounds nuw ({ double, double }, ptr {{.*}}, i32 0, i32 1), align 8 + +// CIR-BEFORE: %[[FP_TO_INT:.*]] = cir.cast(float_to_int, %{{.*}} : !cir.double), !s32i +// CIR-BEFORE: %[[INT_TO_COMPLEX:.*]] = cir.cast(int_to_complex, %[[FP_TO_INT]] : !s32i), !cir.complex<!s32i> + +// CIR-AFTER: %[[TMP:.*]] = cir.load{{.*}} %{{.*}} : !cir.ptr<!cir.double>, !cir.double +// CIR-AFTER-NEXT: %[[REAL:.*]] = cir.cast(float_to_int, %[[TMP]] : !cir.double), !s32i +// CIR-AFTER-NEXT: %[[IMAG:.*]] = cir.const #cir.int<0> : !s32i +// CIR-AFTER-NEXT: %{{.*}} = cir.complex.create %[[REAL]], %[[IMAG]] : !s32i -> !cir.complex<!s32i> + +// LLVM: %[[TMP:.*]] = load double, ptr {{.*}}, align 8 +// LLVM-NEXT: %[[REAL:.*]] = fptosi double %[[TMP]] to i32 +// LLVM-NEXT: %[[TMP_2:.*]] = insertvalue { i32, i32 } undef, i32 %[[REAL]], 0 +// LLVM-NEXT: %[[COMPLEX:.*]] = insertvalue { i32, i32 } %[[TMP_2]], i32 0, 1 +// LLVM-NEXT: store { i32, i32 } %[[COMPLEX]], ptr {{.*}}, align 4 + +// OGCG: %[[TMP:.*]] = load double, ptr {{.*}}, align 8 +// OGCG: %[[REAL:.*]] = fptosi double %[[TMP]] to i32 +// OGCG: store i32 %[[REAL]], ptr {{.*}}, align 4 +// OGCG: store i32 0, ptr getelementptr inbounds nuw ({ i32, i32 }, ptr {{.*}}, i32 0, i32 1), align 4 + +void scalar_to_complex_explicit() { + cd = (double _Complex)sd; + ci = (int _Complex)si; + cd = (double _Complex)si; + ci = (int _Complex)sd; +} + +// CIR-BEFORE: %[[FP_TO_COMPLEX:.*]] = cir.cast(float_to_complex, %{{.*}} : !cir.double), !cir.complex<!cir.double> + +// CIR-AFTER: %[[REAL:.*]] = cir.load{{.*}} %{{.*}} : !cir.ptr<!cir.double>, !cir.double +// CIR-AFTER-NEXT: %[[IMAG:.*]] = cir.const #cir.fp<0.000000e+00> : !cir.double +// CIR-AFTER-NEXT: %{{.*}} = cir.complex.create %[[REAL]], %[[IMAG]] : !cir.double -> !cir.complex<!cir.double> + +// LLVM: %[[REAL:.*]] = load double, ptr {{.*}}, align 8 +// LLVM-NEXT: %[[TMP:.*]] = insertvalue { double, double } undef, double %[[REAL]], 0 +// LLVM-NEXT: %[[COMPLEX:.*]] = insertvalue { double, double } %[[TMP]], double 0.000000e+00, 1 +// LLVM-NEXT: store { double, double } %[[COMPLEX]], ptr {{.*}}, align 8 + +// OGCG: %[[REAL:.*]] = load double, ptr {{.*}}, align 8 +// OGCG: store double %[[REAL]], ptr {{.*}}, align 8 +// OGCG: store double 0.000000e+00, ptr getelementptr inbounds nuw ({ double, double }, ptr @cd, i32 0, i32 1), align 8 + +// CIR-BEFORE: %[[INT_TO_COMPLEX:.*]] = cir.cast(int_to_complex, %{{.*}} : !s32i), !cir.complex<!s32i> + +// CIR-AFTER: %[[REAL:.*]] = cir.load{{.*}} %{{.*}} : !cir.ptr<!s32i>, !s32i +// CIR-AFTER-NEXT: %[[IMAG:.*]] = cir.const #cir.int<0> : !s32i +// CIR-AFTER-NEXT: %{{.*}} = cir.complex.create %[[REAL]], %[[IMAG]] : !s32i -> !cir.complex<!s32i> + +// LLVM: %[[REAL:.*]] = load i32, ptr {{.*}}, align 4 +// LLVM-NEXT: %[[TMP:.*]] = insertvalue { i32, i32 } undef, i32 %[[REAL]], 0 +// LLVM-NEXT: %[[COMPLEX:.*]] = insertvalue { i32, i32 } %[[TMP]], i32 0, 1 +// LLVM-NEXT: store { i32, i32 } %[[COMPLEX]], ptr {{.*}}, align 4 + +// OGCG: %[[REAL:.*]] = load i32, ptr {{.*}}, align 4 +// OGCG: store i32 %[[REAL]], ptr {{.*}}, align 4 +// OGCG: store i32 0, ptr getelementptr inbounds nuw ({ i32, i32 }, ptr @ci, i32 0, i32 1), align 4 + +// CIR-BEFORE: %[[INT_TO_FP:.*]] = cir.cast(int_to_float, %{{.*}} : !s32i), !cir.double +// CIR-BEFORE: %[[FP_TO_COMPLEX:.*]] = cir.cast(float_to_complex, %[[INT_TO_FP]] : !cir.double), !cir.complex<!cir.double> + +// CIR-AFTER: %[[TMP:.*]] = cir.load{{.*}} %{{.*}} : !cir.ptr<!s32i>, !s32i +// CIR-AFTER-NEXT: %[[REAL:.*]] = cir.cast(int_to_float, %[[TMP]] : !s32i), !cir.double +// CIR-AFTER-NEXT: %[[IMAG:.*]] = cir.const #cir.fp<0.000000e+00> : !cir.double +// CIR-AFTER-NEXT: %{{.*}} = cir.complex.create %[[REAL]], %[[IMAG]] : !cir.double -> !cir.complex<!cir.double> + +// LLVM: %[[TMP:.*]] = load i32, ptr {{.*}}, align 4 +// LLVM-NEXT: %[[REAL:.*]] = sitofp i32 %[[TMP]] to double +// LLVM-NEXT: %[[TMP_2:.*]] = insertvalue { double, double } undef, double %[[REAL]], 0 +// LLVM-NEXT: %[[COMPLEX:.*]] = insertvalue { double, double } %[[TMP_2]], double 0.000000e+00, 1 +// LLVM-NEXT: store { double, double } %[[COMPLEX]], ptr {{.*}}, align 8 + +// OGCG: %[[TMP:.*]] = load i32, ptr {{.*}}, align 4 +// OGCG: %[[REAL:.*]] = sitofp i32 %[[TMP]] to double +// OGCG: store double %[[REAL]], ptr {{.*}}, align 8 +// OGCG: store double 0.000000e+00, ptr getelementptr inbounds nuw ({ double, double }, ptr {{.*}}, i32 0, i32 1), align 8 + +// CIR-BEFORE: %[[FP_TO_INT:.*]] = cir.cast(float_to_int, %{{.*}} : !cir.double), !s32i +// CIR-BEFORE: %[[INT_TO_COMPLEX:.*]] = cir.cast(int_to_complex, %[[FP_TO_INT]] : !s32i), !cir.complex<!s32i> + +// CIR-AFTER: %[[TMP:.*]] = cir.load{{.*}} %{{.*}} : !cir.ptr<!cir.double>, !cir.double +// CIR-AFTER-NEXT: %[[REAL:.*]] = cir.cast(float_to_int, %[[TMP]] : !cir.double), !s32i +// CIR-AFTER-NEXT: %[[IMAG:.*]] = cir.const #cir.int<0> : !s32i +// CIR-AFTER-NEXT: %{{.*}} = cir.complex.create %[[REAL]], %[[IMAG]] : !s32i -> !cir.complex<!s32i> + +// LLVM: %[[TMP:.*]] = load double, ptr {{.*}}, align 8 +// LLVM-NEXT: %[[REAL:.*]] = fptosi double %[[TMP]] to i32 +// LLVM-NEXT: %[[TMP_2:.*]] = insertvalue { i32, i32 } undef, i32 %[[REAL]], 0 +// LLVM-NEXT: %[[COMPLEX:.*]] = insertvalue { i32, i32 } %[[TMP_2]], i32 0, 1 +// LLVM-NEXT: store { i32, i32 } %[[COMPLEX]], ptr {{.*}}, align 4 + +// OGCG: %[[TMP:.*]] = load double, ptr {{.*}}, align 8 +// OGCG: %[[REAL:.*]] = fptosi double %[[TMP]] to i32 +// OGCG: store i32 %[[REAL]], ptr {{.*}}, align 4 +// OGCG: store i32 0, ptr getelementptr inbounds nuw ({ i32, i32 }, ptr {{.*}}, i32 0, i32 1), align 4 + +void complex_to_scalar() { + sd = (double)cd; + si = (int)ci; + sd = (double)ci; + si = (int)cd; +} + +// CIR-BEFORE: %[[FP_TO_COMPLEX_REAL:.*]] = cir.cast(float_complex_to_real, %{{.*}} : !cir.complex<!cir.double>), !cir.double + +// CIR-AFTER: %{{.*}} = cir.complex.real %{{.*}} : !cir.complex<!cir.double> -> !cir.double + +// LLVM: %[[REAL:.*]] = extractvalue { double, double } %{{.*}}, 0 +// LLVM: store double %[[REAL]], ptr {{.*}}, align 8 + +// OGCG: %[[REAL:.*]] = load double, ptr {{.*}}, align 8 +// OGCG: store double %[[REAL]], ptr {{.*}}, align 8 + +// CIR-BEFORE: %[[INT_COMPLEX_TO_REAL:.*]] = cir.cast(int_complex_to_real, %{{.*}} : !cir.complex<!s32i>), !s32i + +// CIR-AFTER: %{{.*}} = cir.complex.real %{{.*}} : !cir.complex<!s32i> -> !s32i + +// LLVM: %[[REAL:.*]] = extractvalue { i32, i32 } %{{.*}}, 0 +// LLVM: store i32 %[[REAL]], ptr {{.*}}, align 4 + +// OGCG: %[[REAL:.*]] = load i32, ptr {{.*}}, align 4 +// OGCG: store i32 %[[REAL]], ptr {{.*}}, align 4 + +// CIR-BEFORE: %[[INT_COMPLEX_TO_REAL:.*]] = cir.cast(int_complex_to_real, %{{.*}} : !cir.complex<!s32i>), !s32i +// CIR-BEFORE: %[[INT_TO_FP:.*]] = cir.cast(int_to_float, %[[INT_COMPLEX_TO_REAL]] : !s32i), !cir.double + +// CIR-AFTER: %[[REAL:.*]] = cir.complex.real %{{.*}} : !cir.complex<!s32i> -> !s32i +// CIR-AFTER-NEXT: %{{.*}} = cir.cast(int_to_float, %[[REAL]] : !s32i), !cir.double + +// LLVM: %[[REAL:.*]] = extractvalue { i32, i32 } %{{.+}}, 0 +// LLVM-NEXT: %[[REAL_TO_DOUBLE:.*]] = sitofp i32 %[[REAL]] to double +// LLVM-NEXT: store double %[[REAL_TO_DOUBLE]], ptr {{.*}}, align 8 + +// OGCG: %[[REAL:.*]] = load i32, ptr {{.*}}, align 4 +// OGCG: %[[INT_TO_FP:.*]] = sitofp i32 %[[REAL]] to double +// OGCG: store double %[[INT_TO_FP]], ptr {{.*}}, align 8 + +// CIR-BEFORE: %[[FP_TO_COMPLEX_REAL:.*]] = cir.cast(float_complex_to_real, %{{.*}} : !cir.complex<!cir.double>), !cir.double +// CIR-BEFORE: %[[FP_TO_INT:.*]] = cir.cast(float_to_int, %[[FP_TO_COMPLEX_REAL]] : !cir.double), !s32i + +// CIR-AFTER: %[[REAL:.*]] = cir.complex.real %{{.*}} : !cir.complex<!cir.double> -> !cir.double +// CIR-AFTER-NEXT: %{{.*}} = cir.cast(float_to_int, %[[REAL]] : !cir.double), !s32i + +// LLVM: %[[REAL:.*]] = extractvalue { double, double } %{{.+}}, 0 +// LLVM-NEXT: %[[REAL_TO_INT:.*]] = fptosi double %[[REAL]] to i32 +// LLVM-NEXT: store i32 %[[REAL_TO_INT]], ptr {{.*}}, align 4 + +// OGCG: %[[REAL:.*]] = load double, ptr {{.*}}, align 8 +// OGCG: %[[FP_TO_INT:.*]] = fptosi double %[[REAL]] to i32 +// OGCG: store i32 %[[FP_TO_INT]], ptr {{.*}}, align 4 + +void complex_to_bool() { + b = (bool)cd; + b = (bool)ci; +} + +// CIR-BEFORE: %[[FP_COMPLEX_TO_BOOL:.*]] = cir.cast(float_complex_to_bool, %{{.*}} : !cir.complex<!cir.double>), !cir.bool + +// CIR-AFTER: %[[REAL:.*]] = cir.complex.real %{{.*}} : !cir.complex<!cir.double> -> !cir.double +// CIR-AFTER-NEXT: %[[IMAG:.*]] = cir.complex.imag %{{.*}} : !cir.complex<!cir.double> -> !cir.double +// CIR-AFTER-NEXT: %[[REAL_TO_BOOL:.*]] = cir.cast(float_to_bool, %[[REAL]] : !cir.double), !cir.bool +// CIR-AFTER-NEXT: %[[IMAG_TO_BOOL:.*]] = cir.cast(float_to_bool, %[[IMAG]] : !cir.double), !cir.bool +// CIR-AFTER-NEXT: %[[CONST_TRUE:.*]] = cir.const #true +// CIR-AFTER-NEXT: %{{.*}} = cir.select if %[[REAL_TO_BOOL]] then %[[CONST_TRUE]] else %[[IMAG_TO_BOOL]] : (!cir.bool, !cir.bool, !cir.bool) -> !cir.bool + +// LLVM: %[[REAL:.*]] = extractvalue { double, double } %{{.*}}, 0 +// LLVM-NEXT: %[[IMAG:.*]] = extractvalue { double, double } %{{.*}}, 1 +// LLVM-NEXT: %[[REAL_TO_BOOL:.*]] = fcmp une double %[[REAL]], 0.000000e+00 +// LLVM-NEXT: %[[IMAG_TO_BOOL:.*]] = fcmp une double %[[IMAG]], 0.000000e+00 +// LLVM-NEXT: %[[OR:.*]] = or i1 %[[REAL_TO_BOOL]], %[[IMAG_TO_BOOL]] +// LLVM-NEXT: %[[RESULT:.*]] = zext i1 %[[OR]] to i8 +// LLVM-NEXT: store i8 %[[RESULT]], ptr {{.*}}, align 1 + +// OGCG: %[[REAL:.*]] = load double, ptr {{.*}}, align 8 +// OGCG: %[[IMAG:.*]] = load double, ptr getelementptr inbounds nuw ({ double, double }, ptr {{.*}}, i32 0, i32 1), align 8 +// OGCG: %[[REAL_TO_BOOL:.*]] = fcmp une double %[[REAL]], 0.000000e+00 +// OGCG: %[[IMAG_TO_BOOL:.*]] = fcmp une double %[[IMAG]], 0.000000e+00 +// OGCG: %[[COMPLEX_TO_BOOL:.*]] = or i1 %[[REAL_TO_BOOL]], %[[IMAG_TO_BOOL]] +// OGCG: %[[BOOL_TO_INT:.*]] = zext i1 %[[COMPLEX_TO_BOOL]] to i8 +// OGCG: store i8 %[[BOOL_TO_INT]], ptr {{.*}}, align 1 + +// CIR-BEFORE: %[[INT_COMPLEX_TO_BOOL:.*]] = cir.cast(int_complex_to_bool, %{{.*}} : !cir.complex<!s32i>), !cir.bool + +// CIR-AFTER: %[[REAL:.*]] = cir.complex.real %{{.*}} : !cir.complex<!s32i> -> !s32i +// CIR-AFTER-NEXT: %[[IMAG:.*]] = cir.complex.imag %{{.*}} : !cir.complex<!s32i> -> !s32i +// CIR-AFTER-NEXT: %[[REAL_TO_BOOL:.*]] = cir.cast(int_to_bool, %[[REAL]] : !s32i), !cir.bool +// CIR-AFTER-NEXT: %[[IMAG_TO_BOOL:.*]] = cir.cast(int_to_bool, %[[IMAG]] : !s32i), !cir.bool +// CIR-AFTER-NEXT: %[[CONST_TRUE:.*]] = cir.const #true +// CIR-AFTER-NEXT: %{{.+}} = cir.select if %[[REAL_TO_BOOL]] then %[[CONST_TRUE]] else %[[IMAG_TO_BOOL]] : (!cir.bool, !cir.bool, !cir.bool) -> !cir.bool + +// LLVM: %[[REAL:.*]] = extractvalue { i32, i32 } %{{.*}}, 0 +// LLVM-NEXT: %[[IMAG:.*]] = extractvalue { i32, i32 } %{{.*}}, 1 +// LLVM-NEXT: %[[REAL_TO_BOOL:.*]] = icmp ne i32 %[[REAL]], 0 +// LLVM-NEXT: %[[IMAG_TO_BOOL:.*]] = icmp ne i32 %[[IMAG]], 0 +// LLVM-NEXT: %[[OR:.*]] = or i1 %[[REAL_TO_BOOL]], %[[IMAG_TO_BOOL]] +// LLVM-NEXT: %[[RESULT:.*]] = zext i1 %[[OR]] to i8 +// LLVM-NEXT: store i8 %[[RESULT]], ptr {{.*}}, align 1 + +// OGCG: %[[REAL:.*]] = load i32, ptr {{.*}}, align 4 +// OGCG: %[[IMAG:.*]] = load i32, ptr getelementptr inbounds nuw ({ i32, i32 }, ptr {{.*}}, i32 0, i32 1), align 4 +// OGCG: %[[REAL_TO_BOOL:.*]] = icmp ne i32 %[[REAL]], 0 +// OGCG: %[[IMAG_TO_BOOL:.*]] = icmp ne i32 %[[IMAG]], 0 +// OGCG: %[[COMPLEX_TO_BOOL:.*]] = or i1 %[[REAL_TO_BOOL]], %[[IMAG_TO_BOOL]] +// OGCG: %[[BOOL_TO_INT:.*]] = zext i1 %[[COMPLEX_TO_BOOL]] to i8 +// OGCG: store i8 %[[BOOL_TO_INT]], ptr {{.*}}, align 1 + +void complex_to_complex_cast() { + cd = cf; + ci = cs; +} + +// CIR-BEFORE: %[[TMP:.*]] = cir.load{{.*}} %{{.*}} : !cir.ptr<!cir.complex<!cir.float>>, !cir.complex<!cir.float> +// CIR-BEFORE: %[[FP_COMPLEX:.*]] = cir.cast(float_complex, %[[TMP]] : !cir.complex<!cir.float>), !cir.complex<!cir.double> + +// CIR-AFTER: %[[REAL:.*]] = cir.complex.real %{{.*}} : !cir.complex<!cir.float> -> !cir.float +// CIR-AFTER: %[[IMAG:.*]] = cir.complex.imag %{{.*}} : !cir.complex<!cir.float> -> !cir.float +// CIR-AFTER: %[[REAL_FP_CAST:.*]] = cir.cast(floating, %[[REAL]] : !cir.float), !cir.double +// CIR-AFTER: %[[IMAG_FP_CAST:.*]] = cir.cast(floating, %[[IMAG]] : !cir.float), !cir.double +// CIR-AFTER: %{{.*}} = cir.complex.create %[[REAL_FP_CAST]], %[[IMAG_FP_CAST]] : !cir.double -> !cir.complex<!cir.double> + +// LLVM: %[[REAL:.*]] = extractvalue { float, float } %{{.*}}, 0 +// LLVM: %[[IMAG:.*]] = extractvalue { float, float } %{{.*}}, 1 +// LLVM: %[[REAL_FP_CAST:.*]] = fpext float %[[REAL]] to double +// LLVM: %[[IMAG_FP_CAST:.*]] = fpext float %[[IMAG]] to double +// LLVM: %[[TMP:.*]] = insertvalue { double, double } undef, double %[[REAL_FP_CAST]], 0 +// LLVM: %[[COMPLEX:.*]] = insertvalue { double, double } %[[TMP]], double %[[IMAG_FP_CAST]], 1 +// LLVM: store { double, double } %[[COMPLEX]], ptr {{.*}}, align 8 + +// OGCG: %[[REAL:.*]] = load float, ptr {{.*}}, align 4 +// OGCG: %[[IMAG:.*]] = load float, ptr getelementptr inbounds nuw ({ float, float }, ptr {{.*}}, i32 0, i32 1), align 4 +// OGCG: %[[REAL_FP_CAST:.*]] = fpext float %[[REAL]] to double +// OGCG: %[[IMAG_FP_CAST:.*]] = fpext float %[[IMAG]] to double +// OGCG: store double %[[REAL_FP_CAST]], ptr {{.*}}, align 8 +// OGCG: store double %[[IMAG_FP_CAST]], ptr getelementptr inbounds nuw ({ double, double }, ptr {{.*}}, i32 0, i32 1), align 8 + +// CIR-BEFORE: %[[TMP:.*]] = cir.load{{.*}} %{{.*}} : !cir.ptr<!cir.complex<!s16i>>, !cir.complex<!s16i> +// CIR-BEFORE: %[[INT_COMPLEX:.*]] = cir.cast(int_complex, %[[TMP]] : !cir.complex<!s16i>), !cir.complex<!s32i> + +// CIR-AFTER: %[[REAL:.*]] = cir.complex.real %{{.*}} : !cir.complex<!s16i> -> !s16i +// CIR-AFTER: %[[IMAG:.*]] = cir.complex.imag %{{.*}} : !cir.complex<!s16i> -> !s16i +// CIR-AFTER: %[[REAL_INT_CAST:.*]] = cir.cast(integral, %[[REAL]] : !s16i), !s32i +// CIR-AFTER: %[[IMAG_INT_CAST:.*]] = cir.cast(integral, %[[IMAG]] : !s16i), !s32i +// CIR-AFTER: %{{.*}} = cir.complex.create %[[REAL_INT_CAST]], %[[IMAG_INT_CAST]] : !s32i -> !cir.complex<!s32i> + +// LLVM: %[[REAL:.*]] = extractvalue { i16, i16 } %{{.*}}, 0 +// LLVM: %[[IMAG:.*]] = extractvalue { i16, i16 } %{{.*}}, 1 +// LLVM: %[[REAL_INT_CAST:.*]] = sext i16 %[[REAL]] to i32 +// LLVM: %[[IMAG_INT_CAST:.*]] = sext i16 %[[IMAG]] to i32 +// LLVM: %[[TMP:.*]] = insertvalue { i32, i32 } undef, i32 %[[REAL_INT_CAST]], 0 +// LLVM: %[[COMPLEX:.*]] = insertvalue { i32, i32 } %[[TMP]], i32 %[[IMAG_INT_CAST]], 1 +// LLVM: store { i32, i32 } %[[COMPLEX]], ptr {{.*}}, align 4 + +// OGCG: %[[REAL:.*]] = load i16, ptr {{.*}}, align 2 +// OGCG: %[[IMAG:.*]] = load i16, ptr getelementptr inbounds nuw ({ i16, i16 }, ptr {{.*}}, i32 0, i32 1), align 2 +// OGCG: %[[REAL_INT_CAST:.*]] = sext i16 %[[REAL]] to i32 +// OGCG: %[[IMAG_INT_CAST:.*]] = sext i16 %[[IMAG]] to i32 +// OGCG: store i32 %[[REAL_INT_CAST]], ptr {{.*}}, align 4 +// OGCG: store i32 %[[IMAG_INT_CAST]], ptr getelementptr inbounds nuw ({ i32, i32 }, ptr {{.*}}, i32 0, i32 1), align 4 + diff --git a/clang/test/CodeGen/AArch64/neon-intrinsics.c b/clang/test/CodeGen/AArch64/neon-intrinsics.c index 6304245..035e1ca 100644 --- a/clang/test/CodeGen/AArch64/neon-intrinsics.c +++ b/clang/test/CodeGen/AArch64/neon-intrinsics.c @@ -8585,7 +8585,7 @@ uint32x2_t test_vqshrun_n_s64(int64x2_t a) { // CHECK-NEXT: [[SHUFFLE_I:%.*]] = shufflevector <8 x i8> [[A]], <8 x i8> [[VQSHRUN_N3]], <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15> // CHECK-NEXT: ret <16 x i8> [[SHUFFLE_I]] // -int8x16_t test_vqshrun_high_n_s16(int8x8_t a, int16x8_t b) { +uint8x16_t test_vqshrun_high_n_s16(uint8x8_t a, int16x8_t b) { return vqshrun_high_n_s16(a, b, 3); } @@ -8598,7 +8598,7 @@ int8x16_t test_vqshrun_high_n_s16(int8x8_t a, int16x8_t b) { // CHECK-NEXT: [[SHUFFLE_I:%.*]] = shufflevector <4 x i16> [[A]], <4 x i16> [[VQSHRUN_N3]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> // CHECK-NEXT: ret <8 x i16> [[SHUFFLE_I]] // -int16x8_t test_vqshrun_high_n_s32(int16x4_t a, int32x4_t b) { +uint16x8_t test_vqshrun_high_n_s32(uint16x4_t a, int32x4_t b) { return vqshrun_high_n_s32(a, b, 9); } @@ -8611,7 +8611,7 @@ int16x8_t test_vqshrun_high_n_s32(int16x4_t a, int32x4_t b) { // CHECK-NEXT: [[SHUFFLE_I:%.*]] = shufflevector <2 x i32> [[A]], <2 x i32> [[VQSHRUN_N3]], <4 x i32> <i32 0, i32 1, i32 2, i32 3> // CHECK-NEXT: ret <4 x i32> [[SHUFFLE_I]] // -int32x4_t test_vqshrun_high_n_s64(int32x2_t a, int64x2_t b) { +uint32x4_t test_vqshrun_high_n_s64(uint32x2_t a, int64x2_t b) { return vqshrun_high_n_s64(a, b, 19); } @@ -8810,7 +8810,7 @@ uint32x2_t test_vqrshrun_n_s64(int64x2_t a) { // CHECK-NEXT: [[SHUFFLE_I:%.*]] = shufflevector <8 x i8> [[A]], <8 x i8> [[VQRSHRUN_N3]], <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15> // CHECK-NEXT: ret <16 x i8> [[SHUFFLE_I]] // -int8x16_t test_vqrshrun_high_n_s16(int8x8_t a, int16x8_t b) { +uint8x16_t test_vqrshrun_high_n_s16(uint8x8_t a, int16x8_t b) { return vqrshrun_high_n_s16(a, b, 3); } @@ -8823,7 +8823,7 @@ int8x16_t test_vqrshrun_high_n_s16(int8x8_t a, int16x8_t b) { // CHECK-NEXT: [[SHUFFLE_I:%.*]] = shufflevector <4 x i16> [[A]], <4 x i16> [[VQRSHRUN_N3]], <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7> // CHECK-NEXT: ret <8 x i16> [[SHUFFLE_I]] // -int16x8_t test_vqrshrun_high_n_s32(int16x4_t a, int32x4_t b) { +uint16x8_t test_vqrshrun_high_n_s32(uint16x4_t a, int32x4_t b) { return vqrshrun_high_n_s32(a, b, 9); } @@ -8836,7 +8836,7 @@ int16x8_t test_vqrshrun_high_n_s32(int16x4_t a, int32x4_t b) { // CHECK-NEXT: [[SHUFFLE_I:%.*]] = shufflevector <2 x i32> [[A]], <2 x i32> [[VQRSHRUN_N3]], <4 x i32> <i32 0, i32 1, i32 2, i32 3> // CHECK-NEXT: ret <4 x i32> [[SHUFFLE_I]] // -int32x4_t test_vqrshrun_high_n_s64(int32x2_t a, int64x2_t b) { +uint32x4_t test_vqrshrun_high_n_s64(uint32x2_t a, int64x2_t b) { return vqrshrun_high_n_s64(a, b, 19); } diff --git a/clang/test/CodeGen/X86/prefetchi-error.c b/clang/test/CodeGen/X86/prefetchi-error.c new file mode 100644 index 0000000..31494f7 --- /dev/null +++ b/clang/test/CodeGen/X86/prefetchi-error.c @@ -0,0 +1,7 @@ +// RUN: %clang_cc1 %s -ffreestanding -triple=x86_64-unknown-unknown -target-feature +prefetchi -fsyntax-only -verify + +#include <immintrin.h> + +void test_invalid_prefetchi(void* p) { + __builtin_ia32_prefetchi(p, 1); // expected-error {{argument value 1 is outside the valid range [2, 3]}} +} diff --git a/clang/test/CodeGen/builtin-maximumnum-minimumnum.c b/clang/test/CodeGen/builtin-maximumnum-minimumnum.c new file mode 100644 index 0000000..ea9d2e7 --- /dev/null +++ b/clang/test/CodeGen/builtin-maximumnum-minimumnum.c @@ -0,0 +1,171 @@ +// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 5 +// RUN: %clang_cc1 -x c++ -std=c++20 -disable-llvm-passes -O3 -triple x86_64 %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK + +typedef _Float16 half8 __attribute__((ext_vector_type(8))); +typedef __bf16 bf16x8 __attribute__((ext_vector_type(8))); +typedef float float4 __attribute__((ext_vector_type(4))); +typedef double double2 __attribute__((ext_vector_type(2))); +typedef long double ldouble2 __attribute__((ext_vector_type(2))); + +// CHECK-LABEL: define dso_local noundef <8 x half> @_Z7pfmin16Dv8_DF16_S_( +// CHECK-SAME: <8 x half> noundef [[A:%.*]], <8 x half> noundef [[B:%.*]]) #[[ATTR0:[0-9]+]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[A_ADDR:%.*]] = alloca <8 x half>, align 16 +// CHECK-NEXT: [[B_ADDR:%.*]] = alloca <8 x half>, align 16 +// CHECK-NEXT: store <8 x half> [[A]], ptr [[A_ADDR]], align 16, !tbaa [[TBAA2:![0-9]+]] +// CHECK-NEXT: store <8 x half> [[B]], ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP0:%.*]] = load <8 x half>, ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP1:%.*]] = load <8 x half>, ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[ELT_MINIMUMNUM:%.*]] = call <8 x half> @llvm.minimumnum.v8f16(<8 x half> [[TMP0]], <8 x half> [[TMP1]]) +// CHECK-NEXT: ret <8 x half> [[ELT_MINIMUMNUM]] +// +half8 pfmin16(half8 a, half8 b) { + return __builtin_elementwise_minimumnum(a, b); +} +// CHECK-LABEL: define dso_local noundef <8 x bfloat> @_Z8pfmin16bDv8_DF16bS_( +// CHECK-SAME: <8 x bfloat> noundef [[A:%.*]], <8 x bfloat> noundef [[B:%.*]]) #[[ATTR0]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[A_ADDR:%.*]] = alloca <8 x bfloat>, align 16 +// CHECK-NEXT: [[B_ADDR:%.*]] = alloca <8 x bfloat>, align 16 +// CHECK-NEXT: store <8 x bfloat> [[A]], ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: store <8 x bfloat> [[B]], ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP0:%.*]] = load <8 x bfloat>, ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP1:%.*]] = load <8 x bfloat>, ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[ELT_MINIMUMNUM:%.*]] = call <8 x bfloat> @llvm.minimumnum.v8bf16(<8 x bfloat> [[TMP0]], <8 x bfloat> [[TMP1]]) +// CHECK-NEXT: ret <8 x bfloat> [[ELT_MINIMUMNUM]] +// +bf16x8 pfmin16b(bf16x8 a, bf16x8 b) { + return __builtin_elementwise_minimumnum(a, b); +} +// CHECK-LABEL: define dso_local noundef <4 x float> @_Z7pfmin32Dv4_fS_( +// CHECK-SAME: <4 x float> noundef [[A:%.*]], <4 x float> noundef [[B:%.*]]) #[[ATTR0]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[A_ADDR:%.*]] = alloca <4 x float>, align 16 +// CHECK-NEXT: [[B_ADDR:%.*]] = alloca <4 x float>, align 16 +// CHECK-NEXT: store <4 x float> [[A]], ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: store <4 x float> [[B]], ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP0:%.*]] = load <4 x float>, ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP1:%.*]] = load <4 x float>, ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[ELT_MINIMUMNUM:%.*]] = call <4 x float> @llvm.minimumnum.v4f32(<4 x float> [[TMP0]], <4 x float> [[TMP1]]) +// CHECK-NEXT: ret <4 x float> [[ELT_MINIMUMNUM]] +// +float4 pfmin32(float4 a, float4 b) { + return __builtin_elementwise_minimumnum(a, b); +} +// CHECK-LABEL: define dso_local noundef <2 x double> @_Z7pfmin64Dv2_dS_( +// CHECK-SAME: <2 x double> noundef [[A:%.*]], <2 x double> noundef [[B:%.*]]) #[[ATTR0]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[A_ADDR:%.*]] = alloca <2 x double>, align 16 +// CHECK-NEXT: [[B_ADDR:%.*]] = alloca <2 x double>, align 16 +// CHECK-NEXT: store <2 x double> [[A]], ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: store <2 x double> [[B]], ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP0:%.*]] = load <2 x double>, ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP1:%.*]] = load <2 x double>, ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[ELT_MINIMUMNUM:%.*]] = call <2 x double> @llvm.minimumnum.v2f64(<2 x double> [[TMP0]], <2 x double> [[TMP1]]) +// CHECK-NEXT: ret <2 x double> [[ELT_MINIMUMNUM]] +// +double2 pfmin64(double2 a, double2 b) { + return __builtin_elementwise_minimumnum(a, b); +} +// CHECK-LABEL: define dso_local noundef <2 x x86_fp80> @_Z7pfmin80Dv2_eS_( +// CHECK-SAME: ptr noundef byval(<2 x x86_fp80>) align 32 [[TMP0:%.*]], ptr noundef byval(<2 x x86_fp80>) align 32 [[TMP1:%.*]]) #[[ATTR2:[0-9]+]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[A_ADDR:%.*]] = alloca <2 x x86_fp80>, align 32 +// CHECK-NEXT: [[B_ADDR:%.*]] = alloca <2 x x86_fp80>, align 32 +// CHECK-NEXT: [[A:%.*]] = load <2 x x86_fp80>, ptr [[TMP0]], align 32, !tbaa [[TBAA2]] +// CHECK-NEXT: [[B:%.*]] = load <2 x x86_fp80>, ptr [[TMP1]], align 32, !tbaa [[TBAA2]] +// CHECK-NEXT: store <2 x x86_fp80> [[A]], ptr [[A_ADDR]], align 32, !tbaa [[TBAA2]] +// CHECK-NEXT: store <2 x x86_fp80> [[B]], ptr [[B_ADDR]], align 32, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP2:%.*]] = load <2 x x86_fp80>, ptr [[A_ADDR]], align 32, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP3:%.*]] = load <2 x x86_fp80>, ptr [[B_ADDR]], align 32, !tbaa [[TBAA2]] +// CHECK-NEXT: [[ELT_MINIMUMNUM:%.*]] = call <2 x x86_fp80> @llvm.minimumnum.v2f80(<2 x x86_fp80> [[TMP2]], <2 x x86_fp80> [[TMP3]]) +// CHECK-NEXT: ret <2 x x86_fp80> [[ELT_MINIMUMNUM]] +// +ldouble2 pfmin80(ldouble2 a, ldouble2 b) { + return __builtin_elementwise_minimumnum(a, b); +} + +// CHECK-LABEL: define dso_local noundef <8 x half> @_Z7pfmax16Dv8_DF16_S_( +// CHECK-SAME: <8 x half> noundef [[A:%.*]], <8 x half> noundef [[B:%.*]]) #[[ATTR0]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[A_ADDR:%.*]] = alloca <8 x half>, align 16 +// CHECK-NEXT: [[B_ADDR:%.*]] = alloca <8 x half>, align 16 +// CHECK-NEXT: store <8 x half> [[A]], ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: store <8 x half> [[B]], ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP0:%.*]] = load <8 x half>, ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP1:%.*]] = load <8 x half>, ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[ELT_MAXIMUMNUM:%.*]] = call <8 x half> @llvm.maximumnum.v8f16(<8 x half> [[TMP0]], <8 x half> [[TMP1]]) +// CHECK-NEXT: ret <8 x half> [[ELT_MAXIMUMNUM]] +// +half8 pfmax16(half8 a, half8 b) { + return __builtin_elementwise_maximumnum(a, b); +} +// CHECK-LABEL: define dso_local noundef <8 x bfloat> @_Z8pfmax16bDv8_DF16bS_( +// CHECK-SAME: <8 x bfloat> noundef [[A:%.*]], <8 x bfloat> noundef [[B:%.*]]) #[[ATTR0]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[A_ADDR:%.*]] = alloca <8 x bfloat>, align 16 +// CHECK-NEXT: [[B_ADDR:%.*]] = alloca <8 x bfloat>, align 16 +// CHECK-NEXT: store <8 x bfloat> [[A]], ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: store <8 x bfloat> [[B]], ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP0:%.*]] = load <8 x bfloat>, ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP1:%.*]] = load <8 x bfloat>, ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[ELT_MAXIMUMNUM:%.*]] = call <8 x bfloat> @llvm.maximumnum.v8bf16(<8 x bfloat> [[TMP0]], <8 x bfloat> [[TMP1]]) +// CHECK-NEXT: ret <8 x bfloat> [[ELT_MAXIMUMNUM]] +// +bf16x8 pfmax16b(bf16x8 a, bf16x8 b) { + return __builtin_elementwise_maximumnum(a, b); +} +// CHECK-LABEL: define dso_local noundef <4 x float> @_Z7pfmax32Dv4_fS_( +// CHECK-SAME: <4 x float> noundef [[A:%.*]], <4 x float> noundef [[B:%.*]]) #[[ATTR0]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[A_ADDR:%.*]] = alloca <4 x float>, align 16 +// CHECK-NEXT: [[B_ADDR:%.*]] = alloca <4 x float>, align 16 +// CHECK-NEXT: store <4 x float> [[A]], ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: store <4 x float> [[B]], ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP0:%.*]] = load <4 x float>, ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP1:%.*]] = load <4 x float>, ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[ELT_MAXIMUMNUM:%.*]] = call <4 x float> @llvm.maximumnum.v4f32(<4 x float> [[TMP0]], <4 x float> [[TMP1]]) +// CHECK-NEXT: ret <4 x float> [[ELT_MAXIMUMNUM]] +// +float4 pfmax32(float4 a, float4 b) { + return __builtin_elementwise_maximumnum(a, b); +} +// CHECK-LABEL: define dso_local noundef <2 x double> @_Z7pfmax64Dv2_dS_( +// CHECK-SAME: <2 x double> noundef [[A:%.*]], <2 x double> noundef [[B:%.*]]) #[[ATTR0]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[A_ADDR:%.*]] = alloca <2 x double>, align 16 +// CHECK-NEXT: [[B_ADDR:%.*]] = alloca <2 x double>, align 16 +// CHECK-NEXT: store <2 x double> [[A]], ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: store <2 x double> [[B]], ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP0:%.*]] = load <2 x double>, ptr [[A_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP1:%.*]] = load <2 x double>, ptr [[B_ADDR]], align 16, !tbaa [[TBAA2]] +// CHECK-NEXT: [[ELT_MAXIMUMNUM:%.*]] = call <2 x double> @llvm.maximumnum.v2f64(<2 x double> [[TMP0]], <2 x double> [[TMP1]]) +// CHECK-NEXT: ret <2 x double> [[ELT_MAXIMUMNUM]] +// +double2 pfmax64(double2 a, double2 b) { + return __builtin_elementwise_maximumnum(a, b); +} + +// CHECK-LABEL: define dso_local noundef <2 x x86_fp80> @_Z7pfmax80Dv2_eS_( +// CHECK-SAME: ptr noundef byval(<2 x x86_fp80>) align 32 [[TMP0:%.*]], ptr noundef byval(<2 x x86_fp80>) align 32 [[TMP1:%.*]]) #[[ATTR2]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[A_ADDR:%.*]] = alloca <2 x x86_fp80>, align 32 +// CHECK-NEXT: [[B_ADDR:%.*]] = alloca <2 x x86_fp80>, align 32 +// CHECK-NEXT: [[A:%.*]] = load <2 x x86_fp80>, ptr [[TMP0]], align 32, !tbaa [[TBAA2]] +// CHECK-NEXT: [[B:%.*]] = load <2 x x86_fp80>, ptr [[TMP1]], align 32, !tbaa [[TBAA2]] +// CHECK-NEXT: store <2 x x86_fp80> [[A]], ptr [[A_ADDR]], align 32, !tbaa [[TBAA2]] +// CHECK-NEXT: store <2 x x86_fp80> [[B]], ptr [[B_ADDR]], align 32, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP2:%.*]] = load <2 x x86_fp80>, ptr [[A_ADDR]], align 32, !tbaa [[TBAA2]] +// CHECK-NEXT: [[TMP3:%.*]] = load <2 x x86_fp80>, ptr [[B_ADDR]], align 32, !tbaa [[TBAA2]] +// CHECK-NEXT: [[ELT_MINIMUMNUM:%.*]] = call <2 x x86_fp80> @llvm.minimumnum.v2f80(<2 x x86_fp80> [[TMP2]], <2 x x86_fp80> [[TMP3]]) +// CHECK-NEXT: ret <2 x x86_fp80> [[ELT_MINIMUMNUM]] +// +ldouble2 pfmax80(ldouble2 a, ldouble2 b) { + return __builtin_elementwise_minimumnum(a, b); +} + +//. +// CHECK: [[TBAA2]] = !{[[META3:![0-9]+]], [[META3]], i64 0} +// CHECK: [[META3]] = !{!"omnipotent char", [[META4:![0-9]+]], i64 0} +// CHECK: [[META4]] = !{!"Simple C++ TBAA"} +//. diff --git a/clang/test/CodeGenCXX/microsoft-abi-eh-async.cpp b/clang/test/CodeGenCXX/microsoft-abi-eh-async.cpp new file mode 100644 index 0000000..b831737 --- /dev/null +++ b/clang/test/CodeGenCXX/microsoft-abi-eh-async.cpp @@ -0,0 +1,209 @@ +// REQUIRES: x86-registered-target + +// RUN: %clang_cl -c --target=x86_64-windows-msvc /EHa -O2 /GS- \ +// RUN: -Xclang=-import-call-optimization \ +// RUN: /clang:-S /clang:-o- -- %s 2>&1 \ +// RUN: | FileCheck %s + +#ifdef __clang__ +#define NO_TAIL __attribute((disable_tail_calls)) +#else +#define NO_TAIL +#endif + +void might_throw(); +void other_func(int x); + +void does_not_throw() noexcept(true); + +extern "C" void __declspec(dllimport) some_dll_import(); + +class HasDtor { + int x; + char foo[40]; + +public: + explicit HasDtor(int x); + ~HasDtor(); +}; + +class BadError { +public: + int errorCode; +}; + +void normal_has_regions() { + // CHECK-LABEL: .def "?normal_has_regions@@YAXXZ" + // CHECK: .seh_endprologue + + // <-- state -1 (none) + { + HasDtor hd{42}; + + // <-- state goes from -1 to 0 + // because state changes, we expect the HasDtor::HasDtor() call to have a NOP + // CHECK: call "??0HasDtor@@QEAA@H@Z" + // CHECK-NEXT: nop + + might_throw(); + // CHECK: call "?might_throw@@YAXXZ" + // CHECK-NEXT: nop + + // <-- state goes from 0 to -1 because we're about to call HasDtor::~HasDtor() + // CHECK: call "??1HasDtor@@QEAA@XZ" + // <-- state -1 + } + + // <-- state -1 + other_func(10); + // CHECK: call "?other_func@@YAXH@Z" + // CHECK-NEXT: nop + // CHECK: .seh_startepilogue + + // <-- state -1 +} + +// This tests a tail call to a destructor. +void case_dtor_arg_empty_body(HasDtor x) +{ + // CHECK-LABEL: .def "?case_dtor_arg_empty_body@@YAXVHasDtor@@@Z" + // CHECK: jmp "??1HasDtor@@QEAA@XZ" +} + +int case_dtor_arg_empty_with_ret(HasDtor x) +{ + // CHECK-LABEL: .def "?case_dtor_arg_empty_with_ret@@YAHVHasDtor@@@Z" + // CHECK: .seh_endprologue + + // CHECK: call "??1HasDtor@@QEAA@XZ" + // CHECK-NOT: nop + + // The call to HasDtor::~HasDtor() should NOT have a NOP because the + // following "mov eax, 100" instruction is in the same EH state. + + return 100; + + // CHECK: mov eax, 100 + // CHECK: .seh_startepilogue + // CHECK: .seh_endepilogue + // CHECK: .seh_endproc +} + +int case_noexcept_dtor(HasDtor x) noexcept(true) +{ + // CHECK: .def "?case_noexcept_dtor@@YAHVHasDtor@@@Z" + // CHECK: call "??1HasDtor@@QEAA@XZ" + // CHECK-NEXT: mov eax, 100 + // CHECK: .seh_startepilogue + return 100; +} + +void case_except_simple_call() NO_TAIL +{ + does_not_throw(); +} +// CHECK-LABEL: .def "?case_except_simple_call@@YAXXZ" +// CHECK: .seh_endprologue +// CHECK-NEXT: call "?does_not_throw@@YAXXZ" +// CHECK-NEXT: nop +// CHECK-NEXT: .seh_startepilogue +// CHECK: .seh_endproc + +void case_noexcept_simple_call() noexcept(true) NO_TAIL +{ + does_not_throw(); +} +// CHECK-LABEL: .def "?case_noexcept_simple_call@@YAXXZ" +// CHECK: .seh_endprologue +// CHECK-NEXT: call "?does_not_throw@@YAXXZ" +// CHECK-NEXT: nop +// CHECK-NEXT: .seh_startepilogue +// CHECK: .seh_endepilogue +// CHECK-NEXT: ret +// CHECK-NEXT: .seh_endproc + +// This tests that the destructor is called right before SEH_BeginEpilogue, +// but in a function that has a return value. Loading the return value +// counts as a real instruction, so there is no need for a NOP after the +// dtor call. +int case_dtor_arg_calls_no_throw(HasDtor x) +{ + does_not_throw(); // no NOP expected + return 100; +} +// CHECK-LABEL: .def "?case_dtor_arg_calls_no_throw@@YAHVHasDtor@@@Z" +// CHECK: .seh_endprologue +// CHECK: "?does_not_throw@@YAXXZ" +// CHECK-NEXT: nop +// CHECK: "??1HasDtor@@QEAA@XZ" +// CHECK-NEXT: mov eax, 100 +// CHECK: .seh_startepilogue +// CHECK: .seh_endproc + +// Check the behavior of CALLs that are at the end of MBBs. If a CALL is within +// a non-null EH state (state -1) and is at the end of an MBB, then we expect +// to find an EH_LABEL after the CALL. This causes us to insert a NOP, which +// is the desired result. +void case_dtor_runs_after_join(int x) { + // CHECK-LABEL: .def "?case_dtor_runs_after_join@@YAXH@Z" + // CHECK: .seh_endprologue + + // <-- EH state -1 + + // ctor call does not need a NOP, because it has real instructions after it + HasDtor hd{42}; + // CHECK: call "??0HasDtor@@QEAA@H@Z" + // CHECK-NEXT: nop + // CHECK: test + + // <-- EH state transition from -1 0 + if (x) { + might_throw(); // <-- NOP expected (at end of BB w/ EH_LABEL) + // CHECK: call "?might_throw@@YAXXZ" + // CHECK-NEXT: nop + } else { + other_func(10); // <-- NOP expected (at end of BB w/ EH_LABEL) + // CHECK: call "?other_func@@YAXH@Z" + // CHECK-NEXT: nop + } + does_not_throw(); + // <-- EH state transition 0 to -1 + // ~HasDtor() runs + + // CHECK: .seh_endproc + + // CHECK: "$ip2state$?case_dtor_runs_after_join@@YAXH@Z": + // CHECK-NEXT: .long [[func_begin:.Lfunc_begin([0-9]+)@IMGREL]] + // CHECK-NEXT: .long -1 + // CHECK-NEXT: .long [[tmp1:.Ltmp([0-9]+)]]@IMGREL + // CHECK-NEXT: .long 0 + // CHECK-NEXT: .long [[tmp2:.Ltmp([0-9]+)]]@IMGREL + // CHECK-NEXT: .long -1 +} + + +// Check the behavior of NOP padding around tail calls. +// We do not expect to insert NOPs around tail calls. +// However, the first call (to other_func()) does get a NOP +// because it comes before .seh_startepilogue. +void case_tail_call_no_eh(bool b) { + // tail call; no NOP padding after JMP + if (b) { + does_not_throw(); + // <-- no NOP here + return; + } + + other_func(20); + // <-- NOP does get inserted here +} +// CHECK-LABEL: .def "?case_tail_call_no_eh@@YAX_N@Z" +// CHECK: test +// CHECK-NEXT: je .LBB +// CHECK: jmp "?does_not_throw@@YAXXZ" +// CHECK-SAME: TAILCALL +// CHECK-NEXT: .LBB +// CHECK-NEXT: mov ecx, 20 +// CHECK-NEXT: jmp "?other_func@@YAXH@Z" +// CHECK-SAME: TAILCALL +// CHECK-NEXT: # -- End function diff --git a/clang/test/CodeGenCXX/microsoft-abi-eh-disabled.cpp b/clang/test/CodeGenCXX/microsoft-abi-eh-disabled.cpp new file mode 100644 index 0000000..744f863 --- /dev/null +++ b/clang/test/CodeGenCXX/microsoft-abi-eh-disabled.cpp @@ -0,0 +1,140 @@ +// REQUIRES: x86-registered-target +// RUN: %clang_cl -c --target=x86_64-windows-msvc -EHs-c- -O2 -GS- \ +// RUN: -Xclang=-import-call-optimization \ +// RUN: -clang:-S -clang:-o- -- %s 2>&1 \ +// RUN: | FileCheck %s + +#ifdef __clang__ +#define NO_TAIL __attribute((disable_tail_calls)) +#else +#define NO_TAIL +#endif + +void might_throw(); +void other_func(int x); + +void does_not_throw() noexcept(true); + +extern "C" void __declspec(dllimport) some_dll_import(); + +class HasDtor { + int x; + char foo[40]; + +public: + explicit HasDtor(int x); + ~HasDtor(); +}; + +void normal_has_regions() { + { + HasDtor hd{42}; + + // because state changes, we expect the HasDtor::HasDtor() call to have a NOP + might_throw(); + } + + other_func(10); +} +// CHECK-LABEL: .def "?normal_has_regions@@YAXXZ" +// CHECK: .seh_endprologue +// CHECK: call "??0HasDtor@@QEAA@H@Z" +// CHECK-NEXT: call "?might_throw@@YAXXZ" +// CHECK-NEXT: mov +// CHECK: call "??1HasDtor@@QEAA@XZ" +// CHECK-NEXT: mov ecx, 10 +// CHECK-NEXT: call "?other_func@@YAXH@Z" +// CHECK-NEXT: nop +// CHECK-NEXT: .seh_startepilogue +// CHECK-NOT: "$ip2state$?normal_has_regions@@YAXXZ" + +// This tests a tail call to a destructor. +void case_dtor_arg_empty_body(HasDtor x) +{ +} +// CHECK-LABEL: .def "?case_dtor_arg_empty_body@@YAXVHasDtor@@@Z" +// CHECK: jmp "??1HasDtor@@QEAA@XZ" + +int case_dtor_arg_empty_with_ret(HasDtor x) +{ + // The call to HasDtor::~HasDtor() should NOT have a NOP because the + // following "mov eax, 100" instruction is in the same EH state. + return 100; +} +// CHECK-LABEL: .def "?case_dtor_arg_empty_with_ret@@YAHVHasDtor@@@Z" +// CHECK: .seh_endprologue +// CHECK: call "??1HasDtor@@QEAA@XZ" +// CHECK-NOT: nop +// CHECK: mov eax, 100 +// CHECK: .seh_startepilogue +// CHECK: .seh_endepilogue +// CHECK: .seh_endproc + +void case_except_simple_call() NO_TAIL +{ + does_not_throw(); +} + +// This tests that the destructor is called right before SEH_BeginEpilogue, +// but in a function that has a return value. +int case_dtor_arg_calls_no_throw(HasDtor x) +{ + does_not_throw(); // no NOP expected + return 100; +} + +// Check the behavior of CALLs that are at the end of MBBs. If a CALL is within +// a non-null EH state (state -1) and is at the end of an MBB, then we expect +// to find an EH_LABEL after the CALL. This causes us to insert a NOP, which +// is the desired result. +void case_dtor_runs_after_join(int x) { + + // ctor call does not need a NOP, because it has real instructions after it + HasDtor hd{42}; + + if (x) { + might_throw(); + } else { + other_func(10); + } + does_not_throw(); + // ~HasDtor() runs +} + +// CHECK-LABEL: .def "?case_dtor_runs_after_join@@YAXH@Z" +// CHECK: .seh_endprologue +// CHECK: call "??0HasDtor@@QEAA@H@Z" +// CHECK-NEXT: test +// CHECK: call "?might_throw@@YAXXZ" +// CHECK-NEXT: jmp +// CHECK: call "?other_func@@YAXH@Z" +// CHECK-NEXT: .LBB +// CHECK: call "?does_not_throw@@YAXXZ" +// CHECK-NEXT: lea +// CHECK-NEXT: call "??1HasDtor@@QEAA@XZ" +// CHECK-NEXT: nop +// CHECK-NEXT: .seh_startepilogue +// CHECK-NOT: "$ip2state$?case_dtor_runs_after_join@@YAXH@Z": + + +// Check the behavior of NOP padding around tail calls. +// We do not expect to insert NOPs around tail calls. +// However, the first call (to other_func()) does get a NOP +// because it comes before .seh_startepilogue. +void case_tail_call_no_eh() { + // ordinary call + other_func(10); + + // tail call; no NOP padding after JMP + does_not_throw(); +} + +// CHECK-LABEL: .def "?case_tail_call_no_eh@@YAXXZ" +// CHECK: .seh_endprologue +// CHECK: call "?other_func@@YAXH@Z" +// CHECK-NEXT: nop +// CHECK-NEXT: .seh_startepilogue +// CHECK: .seh_endepilogue +// CHECK: jmp "?does_not_throw@@YAXXZ" +// CHECK-NOT: nop +// CHECK: .seh_endproc diff --git a/clang/test/CodeGenCXX/microsoft-abi-eh-ip2state.cpp b/clang/test/CodeGenCXX/microsoft-abi-eh-ip2state.cpp new file mode 100644 index 0000000..0b7b406 --- /dev/null +++ b/clang/test/CodeGenCXX/microsoft-abi-eh-ip2state.cpp @@ -0,0 +1,242 @@ +// REQUIRES: x86-registered-target +// RUN: %clang_cl -c --target=x86_64-windows-msvc -O2 -EHsc -GS- \ +// RUN: -Xclang=-import-call-optimization \ +// RUN: -clang:-S -clang:-o- -- %s 2>&1 \ +// RUN: | FileCheck %s + +#ifdef __clang__ +#define NO_TAIL __attribute((disable_tail_calls)) +#else +#define NO_TAIL +#endif + +void might_throw(); +void other_func(int x); + +void does_not_throw() noexcept(true); + +extern "C" void __declspec(dllimport) some_dll_import(); + +class HasDtor { + int x; + char foo[40]; + +public: + explicit HasDtor(int x); + ~HasDtor(); +}; + +class BadError { +public: + int errorCode; +}; + +// Verify that when NOP padding for IP2State is active *and* Import Call +// Optimization is active that we see both forms of NOP padding. +void case_calls_dll_import() NO_TAIL { + some_dll_import(); +} +// CHECK-LABEL: .def "?case_calls_dll_import@@YAXXZ" +// CHECK: .seh_endprologue +// CHECK: .Limpcall{{[0-9]+}}: +// CHECK-NEXT: rex64 +// CHECK-NEXT: call __imp_some_dll_import +// CHECK-NEXT: nop dword ptr {{\[.*\]}} +// CHECK-NEXT: nop +// CHECK-NEXT: .seh_startepilogue + +void normal_has_regions() { + + // <-- state -1 (none) + { + HasDtor hd{42}; + + // <-- state goes from -1 to 0 + // because state changes, we expect the HasDtor::HasDtor() call to have a NOP + + might_throw(); + + // <-- state goes from 0 to -1 because we're about to call HasDtor::~HasDtor() + // <-- state -1 + } + + // <-- state -1 + other_func(10); + + // <-- state -1 +} +// CHECK-LABEL: .def "?normal_has_regions@@YAXXZ" +// CHECK: .seh_endprologue +// CHECK: call "??0HasDtor@@QEAA@H@Z" +// CHECK-NEXT: nop +// CHECK: call "?might_throw@@YAXXZ" +// CHECK-NEXT: nop +// CHECK: call "??1HasDtor@@QEAA@XZ" +// CHECK: call "?other_func@@YAXH@Z" +// CHECK-NEXT: nop +// CHECK: .seh_startepilogue + +// This tests a tail call to a destructor. +void case_dtor_arg_empty_body(HasDtor x) +{ +} +// CHECK-LABEL: .def "?case_dtor_arg_empty_body@@YAXVHasDtor@@@Z" +// CHECK: jmp "??1HasDtor@@QEAA@XZ" + +int case_dtor_arg_empty_with_ret(HasDtor x) +{ + // CHECK-LABEL: .def "?case_dtor_arg_empty_with_ret@@YAHVHasDtor@@@Z" + // CHECK: .seh_endprologue + + // CHECK: call "??1HasDtor@@QEAA@XZ" + // CHECK-NOT: nop + + // The call to HasDtor::~HasDtor() should NOT have a NOP because the + // following "mov eax, 100" instruction is in the same EH state. + + return 100; + + // CHECK: mov eax, 100 + // CHECK: .seh_startepilogue + // CHECK: .seh_endepilogue + // CHECK: .seh_endproc +} + +int case_noexcept_dtor(HasDtor x) noexcept(true) +{ + // CHECK: .def "?case_noexcept_dtor@@YAHVHasDtor@@@Z" + // CHECK: call "??1HasDtor@@QEAA@XZ" + // CHECK-NEXT: mov eax, 100 + // CHECK-NEXT: .seh_startepilogue + return 100; +} + +// Simple call of a function that can throw +void case_except_simple_call() NO_TAIL +{ + might_throw(); +} +// CHECK-LABEL: .def "?case_except_simple_call@@YAXXZ" +// CHECK: .seh_endprologue +// CHECK-NEXT: call "?might_throw@@YAXXZ" +// CHECK-NEXT: nop +// CHECK-NEXT: .seh_startepilogue + +// Simple call of a function that cannot throw, in a noexcept context. +void case_noexcept_simple_call() noexcept(true) NO_TAIL +{ + does_not_throw(); +} +// CHECK-LABEL: .def "?case_noexcept_simple_call@@YAXXZ" +// CHECK: .seh_endprologue +// CHECK-NEXT: call "?does_not_throw@@YAXXZ" +// CHECK-NEXT: nop +// CHECK-NEXT: .seh_startepilogue + + +// This tests that the destructor is called right before SEH_BeginEpilogue, +// but in a function that has a return value. +int case_dtor_arg_calls_no_throw(HasDtor x) +{ + does_not_throw(); // no NOP expected + return 100; +} + +// Check the behavior of CALLs that are at the end of MBBs. If a CALL is within +// a non-null EH state (state -1) and is at the end of an MBB, then we expect +// to find an EH_LABEL after the CALL. This causes us to insert a NOP, which +// is the desired result. +void case_dtor_runs_after_join(int x) { + // CHECK-LABEL: .def "?case_dtor_runs_after_join@@YAXH@Z" + // CHECK: .seh_endprologue + + // <-- EH state -1 + + // ctor call does not need a NOP, because it has real instructions after it + HasDtor hd{42}; + // CHECK: call "??0HasDtor@@QEAA@H@Z" + // CHECK-NEXT: test + + // <-- EH state transition from -1 0 + if (x) { + might_throw(); // <-- NOP expected (at end of BB w/ EH_LABEL) + // CHECK: call "?might_throw@@YAXXZ" + // CHECK-NEXT: nop + } else { + other_func(10); // <-- NOP expected (at end of BB w/ EH_LABEL) + // CHECK: call "?other_func@@YAXH@Z" + // CHECK-NEXT: nop + } + does_not_throw(); + // <-- EH state transition 0 to -1 + // ~HasDtor() runs + + // CHECK: .seh_endproc + + // CHECK: "$ip2state$?case_dtor_runs_after_join@@YAXH@Z": + // CHECK-NEXT: .long [[func_begin:.Lfunc_begin([0-9]+)@IMGREL]] + // CHECK-NEXT: .long -1 + // CHECK-NEXT: .long [[tmp1:.Ltmp([0-9]+)]]@IMGREL + // CHECK-NEXT: .long 0 + // CHECK-NEXT: .long [[tmp2:.Ltmp([0-9]+)]]@IMGREL + // CHECK-NEXT: .long -1 +} + + +// Check the behavior of NOP padding around tail calls. +// We do not expect to insert NOPs around tail calls. +// However, the first call (to other_func()) does get a NOP +// because it comes before .seh_startepilogue. +void case_tail_call_no_eh() { + // CHECK-LABEL: .def "?case_tail_call_no_eh@@YAXXZ" + // CHECK: .seh_endprologue + + // ordinary call + other_func(10); + // CHECK: call "?other_func@@YAXH@Z" + // CHECK-NEXT: nop + + // tail call; no NOP padding after JMP + does_not_throw(); + + // CHECK: .seh_startepilogue + // CHECK: .seh_endepilogue + // CHECK: jmp "?does_not_throw@@YAXXZ" + // CHECK-NOT: nop + // CHECK: .seh_endproc +} + + +// Check the behavior of a try/catch +int case_try_catch() { + // CHECK-LABEL: .def "?case_try_catch@@YAHXZ" + // CHECK: .seh_endprologue + + // Because of the EH_LABELs, the ctor and other_func() get NOPs. + + int result = 0; + try { + // CHECK: call "??0HasDtor@@QEAA@H@Z" + // CHECK-NEXT: nop + HasDtor hd{20}; + + // CHECK: call "?other_func@@YAXH@Z" + // CHECK-NEXT: nop + other_func(10); + + // CHECK: call "??1HasDtor@@QEAA@XZ" + // CHECK: mov + } catch (BadError& e) { + result = 1; + } + return result; + + // CHECK: .seh_endproc + + // CHECK: .def "?dtor$4@?0??case_try_catch@@YAHXZ@4HA" + // CHECK: .seh_endprologue + // CHECK: call "??1HasDtor@@QEAA@XZ" + // CHECK-NEXT: nop + // CHECK: .seh_startepilogue + // CHECK: .seh_endproc +} diff --git a/clang/test/DebugInfo/KeyInstructions/asm.c b/clang/test/DebugInfo/KeyInstructions/asm.c new file mode 100644 index 0000000..2b33016 --- /dev/null +++ b/clang/test/DebugInfo/KeyInstructions/asm.c @@ -0,0 +1,59 @@ +// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 5 +// RUN: %clang_cc1 -triple aarch64 -target-feature +ls64 -O0 -emit-llvm -x c %s -o - -gkey-instructions -debug-info-kind=line-tables-only -gno-column-info | FileCheck %s +// Partially copied from clang/test/CodeGen/AArch64/ls64-inline-asm.c + +// Check the inline asm call and result store are Key and distinct atoms. + +struct foo { unsigned long long x[8]; }; +// CHECK-LABEL: define dso_local void @load( +// CHECK-SAME: ptr noundef [[OUTPUT:%.*]], ptr noundef [[ADDR:%.*]]) #[[ATTR0:[0-9]+]] !dbg [[DBG5:![0-9]+]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[OUTPUT_ADDR:%.*]] = alloca ptr, align 8 +// CHECK-NEXT: [[ADDR_ADDR:%.*]] = alloca ptr, align 8 +// CHECK-NEXT: store ptr [[OUTPUT]], ptr [[OUTPUT_ADDR]], align 8 +// CHECK-NEXT: store ptr [[ADDR]], ptr [[ADDR_ADDR]], align 8 +// CHECK-NEXT: [[TMP0:%.*]] = load ptr, ptr [[OUTPUT_ADDR]], align 8, !dbg [[DBG9:![0-9]+]] +// CHECK-NEXT: [[TMP1:%.*]] = load ptr, ptr [[ADDR_ADDR]], align 8, !dbg [[DBG9]] +// CHECK-NEXT: [[TMP2:%.*]] = call i512 asm sideeffect "ld64b $0,[$1]", "=r,r,~{memory}"(ptr [[TMP1]]) #[[ATTR1:[0-9]+]], !dbg [[DBG10:![0-9]+]], !srcloc [[META11:![0-9]+]] +// CHECK-NEXT: store i512 [[TMP2]], ptr [[TMP0]], align 8, !dbg [[DBG12:![0-9]+]] +// CHECK-NEXT: ret void, !dbg [[DBG13:![0-9]+]] +// +void load(struct foo *output, void *addr) { + __asm__ volatile ("ld64b %0,[%1]" : "=r" (*output) : "r" (addr) : "memory"); +} + +// CHECK-LABEL: define dso_local void @load2( +// CHECK-SAME: ptr noundef [[OUTPUT:%.*]], ptr noundef [[ADDR:%.*]]) #[[ATTR0]] !dbg [[DBG14:![0-9]+]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[OUTPUT_ADDR:%.*]] = alloca ptr, align 8 +// CHECK-NEXT: [[ADDR_ADDR:%.*]] = alloca ptr, align 8 +// CHECK-NEXT: store ptr [[OUTPUT]], ptr [[OUTPUT_ADDR]], align 8 +// CHECK-NEXT: store ptr [[ADDR]], ptr [[ADDR_ADDR]], align 8 +// CHECK-NEXT: [[TMP0:%.*]] = load ptr, ptr [[OUTPUT_ADDR]], align 8, !dbg [[DBG15:![0-9]+]] +// CHECK-NEXT: [[TMP1:%.*]] = load ptr, ptr [[ADDR_ADDR]], align 8, !dbg [[DBG15]] +// CHECK-NEXT: [[TMP2:%.*]] = call i32 asm sideeffect "ld64b $0,[$1]", "=r,r,~{memory}"(ptr [[TMP1]]) #[[ATTR1]], !dbg [[DBG16:![0-9]+]], !srcloc [[META17:![0-9]+]] +// CHECK-NEXT: store i32 [[TMP2]], ptr [[TMP0]], align 4, !dbg [[DBG18:![0-9]+]] +// CHECK-NEXT: ret void, !dbg [[DBG19:![0-9]+]] +// +void load2(int *output, void *addr) { + __asm__ volatile ("ld64b %0,[%1]" : "=r" (*output) : "r" (addr) : "memory"); +} +//. +// CHECK: [[META0:![0-9]+]] = distinct !DICompileUnit(language: DW_LANG_C11, file: [[META1:![0-9]+]], producer: "{{.*}}clang version {{.*}}", isOptimized: false, runtimeVersion: 0, emissionKind: LineTablesOnly, splitDebugInlining: false, nameTableKind: None) +// CHECK: [[META1]] = !DIFile(filename: "{{.*}}<stdin>", directory: {{.*}}) +// CHECK: [[DBG5]] = distinct !DISubprogram(name: "load", scope: [[META6:![0-9]+]], file: [[META6]], line: 21, type: [[META7:![0-9]+]], scopeLine: 21, flags: DIFlagPrototyped, spFlags: DISPFlagDefinition, unit: [[META0]], keyInstructions: true) +// CHECK: [[META6]] = !DIFile(filename: "{{.*}}asm.c", directory: {{.*}}) +// CHECK: [[META7]] = !DISubroutineType(types: [[META8:![0-9]+]]) +// CHECK: [[META8]] = !{} +// CHECK: [[DBG9]] = !DILocation(line: 22, scope: [[DBG5]]) +// CHECK: [[DBG10]] = !DILocation(line: 22, scope: [[DBG5]], atomGroup: 1, atomRank: 1) +// CHECK: [[META11]] = !{i64 1458} +// CHECK: [[DBG12]] = !DILocation(line: 22, scope: [[DBG5]], atomGroup: 2, atomRank: 1) +// CHECK: [[DBG13]] = !DILocation(line: 23, scope: [[DBG5]], atomGroup: 3, atomRank: 1) +// CHECK: [[DBG14]] = distinct !DISubprogram(name: "load2", scope: [[META6]], file: [[META6]], line: 38, type: [[META7]], scopeLine: 38, flags: DIFlagPrototyped, spFlags: DISPFlagDefinition, unit: [[META0]], keyInstructions: true) +// CHECK: [[DBG15]] = !DILocation(line: 39, scope: [[DBG14]]) +// CHECK: [[DBG16]] = !DILocation(line: 39, scope: [[DBG14]], atomGroup: 1, atomRank: 1) +// CHECK: [[META17]] = !{i64 2501} +// CHECK: [[DBG18]] = !DILocation(line: 39, scope: [[DBG14]], atomGroup: 2, atomRank: 1) +// CHECK: [[DBG19]] = !DILocation(line: 40, scope: [[DBG14]], atomGroup: 3, atomRank: 1) +//. diff --git a/clang/test/Driver/amdgpu-hip-system-arch.c b/clang/test/Driver/amdgpu-hip-system-arch.c index 9c27bc0..12e298a 100644 --- a/clang/test/Driver/amdgpu-hip-system-arch.c +++ b/clang/test/Driver/amdgpu-hip-system-arch.c @@ -14,14 +14,14 @@ // RUN: | FileCheck %s --check-prefix=NO-OUTPUT-ERROR // RUN: not %clang -### --target=x86_64-unknown-linux-gnu -nogpulib --offload-new-driver --offload-arch=native --amdgpu-arch-tool=%t/amdgpu_arch_fail -x hip %s 2>&1 \ // RUN: | FileCheck %s --check-prefix=NO-OUTPUT-ERROR -// NO-OUTPUT-ERROR: error: cannot determine amdgcn architecture{{.*}}; consider passing it via '--offload-arch' +// NO-OUTPUT-ERROR: error: cannot determine hip architecture{{.*}}; consider passing it via '--offload-arch' // case when amdgpu-arch does not return anything with successful execution // RUN: not %clang -### --target=x86_64-unknown-linux-gnu -nogpulib --offload-arch=native --amdgpu-arch-tool=%t/amdgpu_arch_empty -x hip %s 2>&1 \ // RUN: | FileCheck %s --check-prefix=EMPTY-OUTPUT // RUN: not %clang -### --target=x86_64-unknown-linux-gnu -nogpulib --offload-new-driver --offload-arch=native --amdgpu-arch-tool=%t/amdgpu_arch_empty -x hip %s 2>&1 \ // RUN: | FileCheck %s --check-prefix=EMPTY-OUTPUT -// EMPTY-OUTPUT: error: cannot determine amdgcn architecture: No AMD GPU detected in the system; consider passing it via '--offload-arch' +// EMPTY-OUTPUT: error: cannot determine hip architecture: No GPU detected in the system; consider passing it via '--offload-arch' // case when amdgpu-arch returns a gfx906 GPU. // RUN: %clang -### --target=x86_64-unknown-linux-gnu -nogpuinc -nogpulib --offload-arch=native --amdgpu-arch-tool=%t/amdgpu_arch_gfx906 -x hip %s 2>&1 \ @@ -36,4 +36,4 @@ // RUN: --offload-arch=native --amdgpu-arch-tool=%t/amdgpu_arch_gfx906 \ // RUN: -x hip %s 2>&1 | \ // RUN: FileCheck %s --check-prefix=BAD-TIMEOUT -// BAD-TIMEOUT: clang: error: cannot determine amdgcn architecture: CLANG_TOOLCHAIN_PROGRAM_TIMEOUT expected an integer, got 'foo'; consider passing it via '--offload-arch'; environment variable CLANG_TOOLCHAIN_PROGRAM_TIMEOUT specifies the tool timeout (integer secs, <=0 is infinite) +// BAD-TIMEOUT: clang: error: cannot determine hip architecture: CLANG_TOOLCHAIN_PROGRAM_TIMEOUT expected an integer, got 'foo'; consider passing it via '--offload-arch'; environment variable CLANG_TOOLCHAIN_PROGRAM_TIMEOUT specifies the tool timeout (integer secs, <=0 is infinite) diff --git a/clang/test/Driver/baremetal.cpp b/clang/test/Driver/baremetal.cpp index 4dc3201..adb59e1 100644 --- a/clang/test/Driver/baremetal.cpp +++ b/clang/test/Driver/baremetal.cpp @@ -257,7 +257,7 @@ // CHECK-RV64-SAME:"{{.*}}.o" // CHECK-RV64-SAME: "{{[^"]*}}libclang_rt.builtins.a" // CHECK-RV64-SAME: "-lc" -// CHECK-RV64-SAME: "-X" "-o" "{{.*}}.tmp.out" +// CHECK-RV64-SAME: "-o" "{{.*}}.tmp.out" // RUN: %clangxx %s -### --target=riscv64-unknown-elf 2>&1 \ // RUN: --sysroot=%S/Inputs/basic_riscv64_tree/riscv64-unknown-elf \ @@ -271,7 +271,7 @@ // CHECK-RV64-DEFAULTCXX-SAME: "-lc++" "-lm" // CHECK-RV64-DEFAULTCXX-SAME: "{{[^"]*}}libclang_rt.builtins.a" // CHECK-RV64-DEFAULTCXX-SAME: "-lc" -// CHECK-RV64-DEFAULTCXX-SAME: "-X" "-o" "a.out" +// CHECK-RV64-DEFAULTCXX-SAME: "-o" "a.out" // RUN: %clangxx %s -### --target=riscv64-unknown-elf 2>&1 \ // RUN: --sysroot=%S/Inputs/basic_riscv64_tree/riscv64-unknown-elf \ @@ -288,7 +288,7 @@ // CHECK-RV64-LIBCXX-SAME: "-lc++" "-lm" // CHECK-RV64-LIBCXX-SAME: "{{[^"]*}}libclang_rt.builtins.a" // CHECK-RV64-LIBCXX-SAME: "-lc" -// CHECK-RV64-LIBCXX-SAME: "-X" "-o" "a.out" +// CHECK-RV64-LIBCXX-SAME: "-o" "a.out" // RUN: %clangxx %s -### 2>&1 --target=riscv64-unknown-elf \ // RUN: --sysroot=%S/Inputs/basic_riscv64_tree/riscv64-unknown-elf \ @@ -305,7 +305,7 @@ // CHECK-RV64-LIBSTDCXX-SAME: "-lstdc++" "-lm" // CHECK-RV64-LIBSTDCXX-SAME: "{{[^"]*}}libclang_rt.builtins.a" // CHECK-RV64-LIBSTDCXX-SAME: "-lc" -// CHECK-RV64-LIBSTDCXX-SAME: "-X" "-o" "a.out" +// CHECK-RV64-LIBSTDCXX-SAME: "-o" "a.out" // RUN: %clang %s -### 2>&1 --target=riscv32-unknown-elf \ // RUN: -L some/directory/user/asked/for \ @@ -325,7 +325,7 @@ // CHECK-RV32-SAME: "{{.*}}.o" // CHECK-RV32-SAME: "{{[^"]*}}libclang_rt.builtins.a" // CHECK-RV32-SAME: "-lc" -// CHECK-RV32-SAME: "-X" "-o" "a.out" +// CHECK-RV32-SAME: "-o" "a.out" // RUN: %clangxx %s -### 2>&1 --target=riscv32-unknown-elf \ // RUN: --sysroot=%S/Inputs/basic_riscv32_tree/riscv32-unknown-elf \ @@ -339,7 +339,7 @@ // CHECK-RV32-DEFAULTCXX-SAME: "-lc++" "-lm" // CHECK-RV32-DEFAULTCXX-SAME: "{{[^"]*}}libclang_rt.builtins.a" // CHECK-RV32-DEFAULTCXX-SAME: "-lc" -// CHECK-RV32-DEFAULTCXX-SAME: "-X" "-o" "a.out" +// CHECK-RV32-DEFAULTCXX-SAME: "-o" "a.out" // RUN: %clangxx %s -### 2>&1 --target=riscv32-unknown-elf \ // RUN: --sysroot=%S/Inputs/basic_riscv32_tree/riscv32-unknown-elf \ @@ -355,7 +355,7 @@ // CHECK-RV32-LIBCXX-SAME: "{{.*}}.o" // CHECK-RV32-LIBCXX-SAME: "-lc++" "-lm" // CHECK-RV32-LIBCXX-SAME: "{{[^"]*}}libclang_rt.builtins.a" -// CHECK-RV32-LIBCXX-SAME: "-X" "-o" "a.out" +// CHECK-RV32-LIBCXX-SAME: "-o" "a.out" // RUN: %clangxx %s -### 2>&1 --target=riscv32-unknown-elf \ // RUN: --sysroot=%S/Inputs/basic_riscv32_tree/riscv32-unknown-elf \ @@ -372,7 +372,7 @@ // CHECK-RV32-LIBSTDCXX-SAME: "-lstdc++" "-lm" // CHECK-RV32-LIBSTDCXX-SAME: "{{[^"]*}}libclang_rt.builtins.a" // CHECK-RV32-LIBSTDCXX-SAME: "-lc" -// CHECK-RV32-LIBSTDCXX-SAME: "-X" "-o" "a.out" +// CHECK-RV32-LIBSTDCXX-SAME: "-o" "a.out" // RUN: %clang %s -### 2>&1 --target=riscv64-unknown-elf \ // RUN: -nostdlibinc -nobuiltininc \ diff --git a/clang/test/Driver/cuda-phases.cu b/clang/test/Driver/cuda-phases.cu index 8b91a1d..220a320 100644 --- a/clang/test/Driver/cuda-phases.cu +++ b/clang/test/Driver/cuda-phases.cu @@ -324,8 +324,8 @@ // RUN: -ccc-print-phases --offload-arch=sm_999 -fgpu-rdc -c %s 2>&1 \ // RUN: | FileCheck -check-prefix=INVALID-ARCH %s // INVALID-ARCH: error: unsupported CUDA gpu architecture: sm_999 -// INVALID-ARCH-NEXT: 0: input, "[[INPUT:.+]]", cuda, (host-cuda) -// INVALID-ARCH-NEXT: 1: preprocessor, {0}, cuda-cpp-output, (host-cuda) -// INVALID-ARCH-NEXT: 2: compiler, {1}, ir, (host-cuda) -// INVALID-ARCH-NEXT: 3: backend, {2}, assembler, (host-cuda) -// INVALID-ARCH-NEXT: 4: assembler, {3}, object, (host-cuda) +// INVALID-ARCH: 0: input, "[[INPUT:.+]]", cuda +// INVALID-ARCH-NEXT: 1: preprocessor, {0}, cuda-cpp-output +// INVALID-ARCH-NEXT: 2: compiler, {1}, ir +// INVALID-ARCH-NEXT: 3: backend, {2}, assembler +// INVALID-ARCH-NEXT: 4: assembler, {3}, object diff --git a/clang/test/Driver/hip-inputs.hip b/clang/test/Driver/hip-inputs.hip index 2d4cc31..a8e25ad 100644 --- a/clang/test/Driver/hip-inputs.hip +++ b/clang/test/Driver/hip-inputs.hip @@ -15,5 +15,5 @@ // RUN: --hip-link %S/Inputs/hip_multiple_inputs/a.cu 2>&1 \ // RUN: | FileCheck -check-prefix=MIX %s -// CHECK-NOT: error: mixed CUDA and HIP compilation is not supported -// MIX: error: mixed CUDA and HIP compilation is not supported +// CHECK-NOT: error: mixed CUDA and HIP offloading compilation is not supported +// MIX: error: mixed CUDA and HIP offloading compilation is not supported diff --git a/clang/test/Driver/hip-invalid-target-id.hip b/clang/test/Driver/hip-invalid-target-id.hip index 555043f..ad942e4 100644 --- a/clang/test/Driver/hip-invalid-target-id.hip +++ b/clang/test/Driver/hip-invalid-target-id.hip @@ -4,7 +4,7 @@ // RUN: --rocm-path=%S/Inputs/rocm \ // RUN: %s 2>&1 | FileCheck -check-prefix=NOPLUS %s -// NOPLUS: error: invalid target ID 'gfx908xnack' +// NOPLUS: error: unsupported HIP gpu architecture: gfx908xnack // RUN: not %clang -### --target=x86_64-linux-gnu \ // RUN: -x hip --offload-arch=gfx900 \ @@ -22,7 +22,7 @@ // RUN: --rocm-path=%S/Inputs/rocm \ // RUN: %s 2>&1 | FileCheck -check-prefix=UNK %s -// UNK: error: invalid target ID 'gfx908:unknown+' +// UNK: error: unsupported HIP gpu architecture: gfx900+xnack // RUN: not %clang -### --target=x86_64-linux-gnu \ // RUN: -x hip --offload-arch=gfx908 \ @@ -31,7 +31,7 @@ // RUN: --rocm-path=%S/Inputs/rocm \ // RUN: %s 2>&1 | FileCheck -check-prefix=MIXED %s -// MIXED: error: invalid target ID 'gfx908:sramecc+:unknown+' +// MIXED: error: unsupported HIP gpu architecture: gfx900+xnack // RUN: not %clang -### --target=x86_64-linux-gnu \ // RUN: -x hip --offload-arch=gfx908 \ @@ -55,7 +55,7 @@ // RUN: --rocm-path=%S/Inputs/rocm \ // RUN: %s 2>&1 | FileCheck -check-prefix=NOCOLON %s -// NOCOLON: error: invalid target ID 'gfx900+xnack' +// NOCOLON: error: unsupported HIP gpu architecture: gfx900+xnack // RUN: not %clang -### --target=x86_64-linux-gnu \ // RUN: -x hip --offload-arch=gfx908 \ diff --git a/clang/test/Driver/hip-options.hip b/clang/test/Driver/hip-options.hip index 4fb5571..ba23bc2 100644 --- a/clang/test/Driver/hip-options.hip +++ b/clang/test/Driver/hip-options.hip @@ -115,11 +115,6 @@ // OMP-NOT: "-cc1"{{.*}} "-triple" "amdgcn-amd-amdhsa" {{.*}} "-fopenmp" // OMP: "-cc1"{{.*}} "-triple" "x86_64-unknown-linux-gnu" {{.*}} "-fopenmp" -// RUN: not %clang --target=x86_64-unknown-linux-gnu -nogpuinc -nogpulib \ -// RUN: --offload-arch=gfx906 -fopenmp=libomp -fopenmp-targets=amdgcn %s 2>&1 \ -// RUN: | FileCheck -check-prefix=OMPTGT %s -// OMPTGT: unsupported option '--offload-targets=' for language mode 'HIP' - // Check -Xoffload-linker option is passed to lld. // RUN: %clang -### --target=x86_64-unknown-linux-gnu -nogpuinc -nogpulib \ diff --git a/clang/test/Driver/hip-phases.hip b/clang/test/Driver/hip-phases.hip index d8a58b7..0ad5d76 100644 --- a/clang/test/Driver/hip-phases.hip +++ b/clang/test/Driver/hip-phases.hip @@ -675,3 +675,25 @@ // DEVICE-ONLY-NEXT: 2: compiler, {1}, ir, (device-hip, gfx90a) // DEVICE-ONLY-NEXT: 3: backend, {2}, ir, (device-hip, gfx90a) // DEVICE-ONLY-NEXT: 4: offload, "device-hip (amdgcn-amd-amdhsa:gfx90a)" {3}, none + +// +// Test the new driver bundling SPIR-V targets. +// +// RUN: %clang -### --target=x86_64-linux-gnu --offload-new-driver -ccc-print-phases \ +// RUN: --offload-device-only --offload-arch=amdgcnspirv,gfx1030 %s 2>&1 \ +// RUN: | FileCheck -check-prefix=SPIRV-ONLY %s +// SPIRV-ONLY: 0: input, "[[INPUT:.+]]", hip, (device-hip, gfx1030) +// SPIRV-ONLY-NEXT: 1: preprocessor, {0}, hip-cpp-output, (device-hip, gfx1030) +// SPIRV-ONLY-NEXT: 2: compiler, {1}, ir, (device-hip, gfx1030) +// SPIRV-ONLY-NEXT: 3: backend, {2}, assembler, (device-hip, gfx1030) +// SPIRV-ONLY-NEXT: 4: assembler, {3}, object, (device-hip, gfx1030) +// SPIRV-ONLY-NEXT: 5: linker, {4}, image, (device-hip, gfx1030) +// SPIRV-ONLY-NEXT: 6: offload, "device-hip (amdgcn-amd-amdhsa:gfx1030)" {5}, image +// SPIRV-ONLY-NEXT: 7: input, "[[INPUT]]", hip, (device-hip, amdgcnspirv) +// SPIRV-ONLY-NEXT: 8: preprocessor, {7}, hip-cpp-output, (device-hip, amdgcnspirv) +// SPIRV-ONLY-NEXT: 9: compiler, {8}, ir, (device-hip, amdgcnspirv) +// SPIRV-ONLY-NEXT: 10: backend, {9}, ir, (device-hip, amdgcnspirv) +// SPIRV-ONLY-NEXT: 11: linker, {10}, image, (device-hip, amdgcnspirv) +// SPIRV-ONLY-NEXT: 12: offload, "device-hip (spirv64-amd-amdhsa:amdgcnspirv)" {11}, image +// SPIRV-ONLY-NEXT: 13: linker, {6, 12}, hip-fatbin, (device-hip) +// SPIRV-ONLY-NEXT: 14: offload, "device-hip (amdgcn-amd-amdhsa)" {13}, none diff --git a/clang/test/Driver/invalid-offload-options.cpp b/clang/test/Driver/invalid-offload-options.cpp index 48d5310..6048a3c 100644 --- a/clang/test/Driver/invalid-offload-options.cpp +++ b/clang/test/Driver/invalid-offload-options.cpp @@ -1,29 +1,7 @@ // UNSUPPORTED: system-windows -// RUN: not %clang -### -x hip --target=x86_64-linux-gnu --offload= \ -// RUN: --hip-path=%S/Inputs/hipspv -nogpuinc -nogpulib %s \ -// RUN: 2>&1 | FileCheck --check-prefix=INVALID-TARGET %s // RUN: not %clang -### -x hip --target=x86_64-linux-gnu --offload=foo \ // RUN: --hip-path=%S/Inputs/hipspv -nogpuinc -nogpulib %s \ // RUN: 2>&1 | FileCheck --check-prefix=INVALID-TARGET %s // INVALID-TARGET: error: invalid or unsupported offload target: '{{.*}}' - -// In the future we should be able to specify multiple targets for HIP -// compilation but currently it is not supported. -// -// RUN: not %clang -### -x hip --target=x86_64-linux-gnu --offload=foo,bar \ -// RUN: --hip-path=%S/Inputs/hipspv -nogpuinc -nogpulib %s \ -// RUN: 2>&1 | FileCheck --check-prefix=TOO-MANY-TARGETS %s -// RUN: not %clang -### -x hip --target=x86_64-linux-gnu \ -// RUN: --offload=foo --offload=bar \ -// RUN: --hip-path=%S/Inputs/hipspv -nogpuinc -nogpulib %s \ -// RUN: 2>&1 | FileCheck --check-prefix=TOO-MANY-TARGETS %s - -// TOO-MANY-TARGETS: error: only one offload target is supported - -// RUN: not %clang -### -x hip --target=x86_64-linux-gnu -nogpuinc -nogpulib \ -// RUN: --offload=amdgcn-amd-amdhsa --offload-arch=gfx900 %s \ -// RUN: 2>&1 | FileCheck --check-prefix=OFFLOAD-ARCH-MIX %s - -// OFFLOAD-ARCH-MIX: error: option '--offload-arch' cannot be specified with '--offload' diff --git a/clang/test/Driver/nvptx-cuda-system-arch.c b/clang/test/Driver/nvptx-cuda-system-arch.c index c54eeac..2d4eca8 100644 --- a/clang/test/Driver/nvptx-cuda-system-arch.c +++ b/clang/test/Driver/nvptx-cuda-system-arch.c @@ -16,14 +16,14 @@ // RUN: | FileCheck %s --check-prefix=NO-OUTPUT-ERROR // RUN: not %clang -### --target=x86_64-unknown-linux-gnu -nogpulib --offload-new-driver --offload-arch=native --nvptx-arch-tool=%t/nvptx_arch_fail -x cuda %s 2>&1 \ // RUN: | FileCheck %s --check-prefix=NO-OUTPUT-ERROR -// NO-OUTPUT-ERROR: error: cannot determine nvptx64 architecture{{.*}}; consider passing it via '--offload-arch' +// NO-OUTPUT-ERROR: error: cannot determine cuda architecture{{.*}}; consider passing it via '--offload-arch' // case when nvptx-arch does not return anything with successful execution // RUN: not %clang -### --target=x86_64-unknown-linux-gnu -nogpulib --offload-arch=native --nvptx-arch-tool=%t/nvptx_arch_empty -x cuda %s 2>&1 \ // RUN: | FileCheck %s --check-prefix=EMPTY-OUTPUT // RUN: not %clang -### --target=x86_64-unknown-linux-gnu -nogpulib --offload-new-driver --offload-arch=native --nvptx-arch-tool=%t/nvptx_arch_empty -x cuda %s 2>&1 \ // RUN: | FileCheck %s --check-prefix=EMPTY-OUTPUT -// EMPTY-OUTPUT: error: cannot determine nvptx64 architecture: No NVIDIA GPU detected in the system; consider passing it via '--offload-arch' +// EMPTY-OUTPUT: error: cannot determine cuda architecture: No GPU detected in the system; consider passing it via '--offload-arch' // case when nvptx-arch does not return anything with successful execution // RUN: %clang -### --target=x86_64-unknown-linux-gnu -nogpulib --offload-arch=native --nvptx-arch-tool=%t/nvptx_arch_sm_70 -x cuda --cuda-path=%S/Inputs/CUDA_102/usr/local/cuda %s 2>&1 \ @@ -49,4 +49,4 @@ // RUN: --offload-arch=native --nvptx-arch-tool=%t/nvptx_arch_sm_70 \ // RUN: --cuda-path=%S/Inputs/CUDA_102/usr/local/cuda -x cuda %s 2>&1 | \ // RUN: FileCheck %s --check-prefix=BAD-TIMEOUT -// BAD-TIMEOUT: clang: error: cannot determine nvptx64 architecture: CLANG_TOOLCHAIN_PROGRAM_TIMEOUT expected an integer, got 'foo'; consider passing it via '--offload-arch'; environment variable CLANG_TOOLCHAIN_PROGRAM_TIMEOUT specifies the tool timeout (integer secs, <=0 is infinite) +// BAD-TIMEOUT: clang: error: cannot determine cuda architecture: CLANG_TOOLCHAIN_PROGRAM_TIMEOUT expected an integer, got 'foo'; consider passing it via '--offload-arch'; environment variable CLANG_TOOLCHAIN_PROGRAM_TIMEOUT specifies the tool timeout (integer secs, <=0 is infinite) diff --git a/clang/test/Driver/offload-target.c b/clang/test/Driver/offload-target.c new file mode 100644 index 0000000..23a2cf2 --- /dev/null +++ b/clang/test/Driver/offload-target.c @@ -0,0 +1,22 @@ +// RUN: %clang -### -fsycl --offload-targets=spirv64 -nogpuinc %s -ccc-print-bindings 2>&1 \ +// RUN: | FileCheck %s -check-prefix=SYCL +// SYCL: "spirv64" - "clang", inputs: ["[[INPUT:.+]]"], output: "[[SYCL_BC:.+]]" + +// RUN: %clang -### --offload-targets=amdgcn-amd-amdhsa -nogpulib -nogpuinc -x hip %s -ccc-print-bindings 2>&1 \ +// RUN: | FileCheck %s -check-prefix=HIP +// HIP: "amdgcn-amd-amdhsa" - "clang", inputs: ["[[INPUT:.+]]"], output: "[[AMD_OBJ:.+]]" + +// RUN: %clang -### --offload-targets=nvptx64-nvidia-cuda -nogpulib -nogpuinc -x cuda %s -ccc-print-bindings 2>&1 \ +// RUN: | FileCheck %s -check-prefix=CUDA +// CUDA: "nvptx64-nvidia-cuda" - "clang", inputs: ["[[INPUT:.+]]"], output: "[[NV_OBJ:.+]]" + +// RUN: %clang -### --offload-targets=amdgcn-amd-amdhsa,nvptx64-nvidia-cuda -fopenmp=libomp \ +// RUN: -Xarch_amdgcn --offload-arch=gfx90a -Xarch_nvptx64 --offload-arch=sm_89 \ +// RUN: -nogpulib -nogpuinc %s -ccc-print-bindings 2>&1 \ +// RUN: | FileCheck %s -check-prefix=OPENMP +// OPENMP: "amdgcn-amd-amdhsa" - "clang", inputs: ["[[INPUT:.+]]"], output: "[[AMD_OBJ:.+]]" +// OPENMP: "nvptx64-nvidia-cuda" - "clang", inputs: ["[[INPUT]]"], output: "[[NV_OBJ:.+]]" + +// RUN: %clang -### --offload-targets=spirv64-amd-amdhsa -nogpulib -nogpuinc -x hip %s -ccc-print-bindings 2>&1 \ +// RUN: | FileCheck %s -check-prefix=HIPSPIRV +// HIPSPIRV: "spirv64-amd-amdhsa" - "clang", inputs: ["[[INPUT:.+]]"], output: "[[AMD_OBJ:.+]]" diff --git a/clang/test/Driver/openbsd.c b/clang/test/Driver/openbsd.c index 6639e9d..1f12cfc 100644 --- a/clang/test/Driver/openbsd.c +++ b/clang/test/Driver/openbsd.c @@ -127,9 +127,12 @@ // UNWIND-TABLES: "-funwind-tables=2" // NO-UNWIND-TABLES-NOT: "-funwind-tables=2" -// Check that the -X and --no-relax flags are passed to the linker on riscv64 +// Check that the -X and --no-relax flags are passed to the linker +// RUN: %clang --target=loongarch64-unknown-openbsd -mno-relax -### %s 2>&1 \ +// RUN: | FileCheck --check-prefix=LA64-FLAGS %s // RUN: %clang --target=riscv64-unknown-openbsd -mno-relax -### %s 2>&1 \ // RUN: | FileCheck -check-prefix=RISCV64-FLAGS %s +// LA64-FLAGS: "-X" "--no-relax" // RISCV64-FLAGS: "-X" "--no-relax" // Check passing LTO flags to the linker diff --git a/clang/test/Driver/openmp-offload.c b/clang/test/Driver/openmp-offload.c index 162ff20..64d45f9 100644 --- a/clang/test/Driver/openmp-offload.c +++ b/clang/test/Driver/openmp-offload.c @@ -7,7 +7,7 @@ /// Check whether an invalid OpenMP target is specified: // RUN: not %clang -### -fopenmp=libomp -fopenmp-targets=aaa-bbb-ccc-ddd %s 2>&1 \ // RUN: | FileCheck -check-prefix=CHK-INVALID-TARGET %s -// CHK-INVALID-TARGET: error: OpenMP target is invalid: 'aaa-bbb-ccc-ddd' +// CHK-INVALID-TARGET: error: invalid or unsupported offload target: 'aaa-bbb-ccc-ddd' /// ########################################################################### @@ -18,15 +18,6 @@ /// ########################################################################### -/// Check error for no -fopenmp option -// RUN: not %clang -### -fopenmp-targets=powerpc64le-ibm-linux-gnu %s 2>&1 \ -// RUN: | FileCheck -check-prefix=CHK-NO-FOPENMP %s -// RUN: not %clang -### -fopenmp=libgomp -fopenmp-targets=powerpc64le-ibm-linux-gnu %s 2>&1 \ -// RUN: | FileCheck -check-prefix=CHK-NO-FOPENMP %s -// CHK-NO-FOPENMP: error: '-fopenmp-targets' must be used in conjunction with a '-fopenmp' option compatible with offloading; e.g., '-fopenmp=libomp' or '-fopenmp=libiomp5' - -/// ########################################################################### - /// Check warning for duplicate offloading targets. // RUN: %clang -### -ccc-print-phases -fopenmp=libomp -fopenmp-targets=powerpc64le-ibm-linux-gnu,powerpc64le-ibm-linux-gnu %s 2>&1 \ // RUN: | FileCheck -check-prefix=CHK-DUPLICATES %s diff --git a/clang/test/Driver/openmp-system-arch.c b/clang/test/Driver/openmp-system-arch.c index b18ecf3..167b07a 100644 --- a/clang/test/Driver/openmp-system-arch.c +++ b/clang/test/Driver/openmp-system-arch.c @@ -24,13 +24,7 @@ // RUN: not %clang -### --target=x86_64-unknown-linux-gnu -nogpulib -fopenmp=libomp --offload-arch=native \ // RUN: --nvptx-arch-tool=%t/nvptx_arch_empty --amdgpu-arch-tool=%t/amdgpu_arch_empty %s 2>&1 \ // RUN: | FileCheck %s --check-prefix=NO-OUTPUT-ERROR -// RUN: not %clang -### --target=x86_64-unknown-linux-gnu -nogpulib -fopenmp=libomp --offload-arch= \ -// RUN: --nvptx-arch-tool=%t/nvptx_arch_fail --amdgpu-arch-tool=%t/amdgpu_arch_fail %s 2>&1 \ -// RUN: | FileCheck %s --check-prefix=NO-OUTPUT-ERROR -// RUN: not %clang -### --target=x86_64-unknown-linux-gnu -nogpulib -fopenmp=libomp --offload-arch= \ -// RUN: --nvptx-arch-tool=%t/nvptx_arch_empty --amdgpu-arch-tool=%t/amdgpu_arch_empty %s 2>&1 \ -// RUN: | FileCheck %s --check-prefix=NO-OUTPUT-ERROR -// NO-OUTPUT-ERROR: error: failed to deduce triple for target architecture 'native'; specify the triple using '-fopenmp-targets' and '-Xopenmp-target' instead +// NO-OUTPUT-ERROR: error: cannot determine openmp architecture // case when amdgpu-arch succeeds. // RUN: %clang -### --target=x86_64-unknown-linux-gnu -nogpulib -fopenmp=libomp --offload-arch=native \ diff --git a/clang/test/Driver/print-multi-selection-flags.c b/clang/test/Driver/print-multi-selection-flags.c index 8cf8f04bb..b1a0a29 100644 --- a/clang/test/Driver/print-multi-selection-flags.c +++ b/clang/test/Driver/print-multi-selection-flags.c @@ -126,3 +126,20 @@ // CHECK-PIE1: -fpie // CHECK-ROPI: -fropi // CHECK-RWPI: -frwpi + +// RUN: %clang -multi-lib-config=%S/Inputs/multilib/empty.yaml -print-multi-flags-experimental --target=arm-none-eabi -march=armv7a -Os | FileCheck --check-prefix=CHECK-OPT-OS %s +// RUN: %clang -multi-lib-config=%S/Inputs/multilib/empty.yaml -print-multi-flags-experimental --target=arm-none-eabi -march=armv7a -Oz | FileCheck --check-prefix=CHECK-OPT-OZ %s +// RUN: %clang -multi-lib-config=%S/Inputs/multilib/empty.yaml -print-multi-flags-experimental --target=arm-none-eabi -march=armv7a | FileCheck --check-prefix=CHECK-OPT %s +// RUN: %clang -multi-lib-config=%S/Inputs/multilib/empty.yaml -print-multi-flags-experimental --target=arm-none-eabi -march=armv7a -O1 | FileCheck --check-prefix=CHECK-OPT %s +// RUN: %clang -multi-lib-config=%S/Inputs/multilib/empty.yaml -print-multi-flags-experimental --target=arm-none-eabi -march=armv7a -O2 | FileCheck --check-prefix=CHECK-OPT %s +// RUN: %clang -multi-lib-config=%S/Inputs/multilib/empty.yaml -print-multi-flags-experimental --target=arm-none-eabi -march=armv7a -O3 | FileCheck --check-prefix=CHECK-OPT %s +// RUN: %clang -multi-lib-config=%S/Inputs/multilib/empty.yaml -print-multi-flags-experimental --target=aarch64-none-eabi -Os | FileCheck --check-prefix=CHECK-OPT-OS %s +// RUN: %clang -multi-lib-config=%S/Inputs/multilib/empty.yaml -print-multi-flags-experimental --target=aarch64-none-eabi -Oz | FileCheck --check-prefix=CHECK-OPT-OZ %s +// RUN: %clang -multi-lib-config=%S/Inputs/multilib/empty.yaml -print-multi-flags-experimental --target=aarch64-none-eabi | FileCheck --check-prefix=CHECK-OPT %s +// RUN: %clang -multi-lib-config=%S/Inputs/multilib/empty.yaml -print-multi-flags-experimental --target=aarch64-none-eabi -O1 | FileCheck --check-prefix=CHECK-OPT %s +// RUN: %clang -multi-lib-config=%S/Inputs/multilib/empty.yaml -print-multi-flags-experimental --target=aarch64-none-eabi -O2 | FileCheck --check-prefix=CHECK-OPT %s +// RUN: %clang -multi-lib-config=%S/Inputs/multilib/empty.yaml -print-multi-flags-experimental --target=aarch64-none-eabi -O3 | FileCheck --check-prefix=CHECK-OPT %s +// CHECK-OPT-OZ: -Oz +// CHECK-OPT-OS: -Os +// CHECK-OPT-NOT: -Oz +// CHECK-OPT-NOT: -Os diff --git a/clang/test/Driver/sparc-target-features.c b/clang/test/Driver/sparc-target-features.c index a839604..bd17da1 100644 --- a/clang/test/Driver/sparc-target-features.c +++ b/clang/test/Driver/sparc-target-features.c @@ -20,6 +20,11 @@ // RUN: %clang --target=sparc -mvis2 %s -### 2>&1 | FileCheck -check-prefix=VIS2 %s // RUN: %clang --target=sparc -mno-vis2 %s -### 2>&1 | FileCheck -check-prefix=NO-VIS2 %s +/// Solaris/SPARC defaults to -mvis2 +// RUN: %clang --target=sparc-sun-solaris2.11 %s -### 2>&1 | FileCheck -check-prefix=VIS2 %s +// RUN: %clang --target=sparc-sun-solaris2.11 -mno-vis2 %s -### 2>&1 | FileCheck -check-prefix=NO-VIS2 %s +// RUN: %clang --target=sparcv9-sun-solaris2.11 %s -### 2>&1 | FileCheck -check-prefix=VIS2 %s +// RUN: %clang --target=sparcv9-sun-solaris2.11 -mno-vis2 %s -### 2>&1 | FileCheck -check-prefix=NO-VIS2 %s // VIS2: "-target-feature" "+vis2" // NO-VIS2: "-target-feature" "-vis2" @@ -34,4 +39,8 @@ // SOFT-QUAD-FLOAT: "-target-feature" "-hard-quad-float" // RUN: %clang --target=sparc -mv8plus %s -### 2>&1 | FileCheck -check-prefix=V8PLUS %s +/// 32-bit Solaris/SPARC defaults to -mv8plus +// RUN: %clang --target=sparc-sun-solaris2.11 %s -### 2>&1 | FileCheck -check-prefix=V8PLUS %s +// RUN: %clang --target=sparc-sun-solaris2.11 -mno-v8plus %s -### 2>&1 | FileCheck -check-prefix=NO-V8PLUS %s // V8PLUS: "-target-feature" "+v8plus" +// NO-V8PLUS-NOT: "-target-feature" "+v8plus" diff --git a/clang/test/Interpreter/fail.cpp b/clang/test/Interpreter/fail.cpp index 633d927..4963df8 100644 --- a/clang/test/Interpreter/fail.cpp +++ b/clang/test/Interpreter/fail.cpp @@ -18,4 +18,11 @@ extern "C" int printf(const char *, ...); int i = 42; auto r1 = printf("i = %d\n", i); // CHECK: i = 42 + +1aap = 42; // expected-error {{invalid digit 'a' in decimal constant}} +1aap = 42; i = 5; // expected-error {{invalid digit 'a' in decimal constant}} + +printf("i = %d\n", i); +// CHECK: i = 42 + %quit diff --git a/clang/test/Interpreter/pretty-print.c b/clang/test/Interpreter/pretty-print.c index 56488a1..e1408c0 100644 --- a/clang/test/Interpreter/pretty-print.c +++ b/clang/test/Interpreter/pretty-print.c @@ -3,7 +3,7 @@ // RUN: cat %s | clang-repl -Xcc -xc | FileCheck %s // RUN: cat %s | clang-repl -Xcc -std=c++11 | FileCheck %s -// UNSUPPORTED: hwasan +// UNSUPPORTED: hwasan, msan char c = 'a'; c diff --git a/clang/test/Interpreter/pretty-print.cpp b/clang/test/Interpreter/pretty-print.cpp index fd79d31..e1036ab 100644 --- a/clang/test/Interpreter/pretty-print.cpp +++ b/clang/test/Interpreter/pretty-print.cpp @@ -1,6 +1,7 @@ // RUN: clang-repl "int i = 10;" 'extern "C" int printf(const char*,...);' \ // RUN: 'auto r1 = printf("i = %d\n", i);' | FileCheck --check-prefix=CHECK-DRIVER %s -// UNSUPPORTED: system-aix +// The test is flaky with asan https://github.com/llvm/llvm-project/pull/148701. +// UNSUPPORTED: system-aix, asan // CHECK-DRIVER: i = 10 // RUN: cat %s | clang-repl -Xcc -std=c++11 -Xcc -fno-delayed-template-parsing | FileCheck %s extern "C" int printf(const char*,...); diff --git a/clang/test/Layout/ms-no-unique-address.cpp b/clang/test/Layout/ms-no-unique-address.cpp index 51cfd9a..fd92056 100644 --- a/clang/test/Layout/ms-no-unique-address.cpp +++ b/clang/test/Layout/ms-no-unique-address.cpp @@ -1,4 +1,5 @@ // RUN: %clang_cc1 -std=c++2a -fsyntax-only -triple x86_64-windows-msvc -fms-compatibility -fdump-record-layouts %s | FileCheck %s +// RUN: %clang_cc1 -std=c++2a -fsyntax-only -triple x86_64-uefi -fms-compatibility -fdump-record-layouts %s | FileCheck %s namespace Empty { struct A {}; diff --git a/clang/test/Misc/time-passes.c b/clang/test/Misc/time-passes.c index 370f52e..6afbcd4 100644 --- a/clang/test/Misc/time-passes.c +++ b/clang/test/Misc/time-passes.c @@ -19,6 +19,12 @@ // RUN: -ftime-report-json %s -o /dev/null \ // RUN: -mllvm -info-output-file=%t // RUN: cat %t | FileCheck %s --check-prefixes=JSON +// Check that -stats-file-timers only outputs pass time info in the stats file +// and not stderr. +// RUN: %clang_cc1 -emit-obj -O1 \ +// RUN: %s -o /dev/null \ +// RUN: -stats-file=%t -stats-file-timers 2>&1 | count 0 +// RUN: FileCheck %s -input-file=%t -check-prefixes=JSON // TIME: Pass execution timing report // TIME: Total Execution Time: diff --git a/clang/test/OpenMP/declare_variant_clauses_ast_print.cpp b/clang/test/OpenMP/declare_variant_clauses_ast_print.cpp index c14e19c..e98a23e 100644 --- a/clang/test/OpenMP/declare_variant_clauses_ast_print.cpp +++ b/clang/test/OpenMP/declare_variant_clauses_ast_print.cpp @@ -38,11 +38,11 @@ #ifndef HEADER #define HEADER -void foo_v1(float *AAA, float *BBB, int *I) {return;} -void foo_v2(float *AAA, float *BBB, int *I) {return;} -void foo_v3(float *AAA, float *BBB, int *I) {return;} +void foo_v1(float *AAA, float *BBB, int &CCC, int *I) {return;} +void foo_v2(float *AAA, float *BBB, int &CCC, int *I) {return;} +void foo_v3(float *AAA, float *BBB, int &CCC, int *I) {return;} -//DUMP: FunctionDecl{{.*}} foo 'void (float *, float *, int *)' +//DUMP: FunctionDecl{{.*}} foo 'void (float *, float *, int &, int *)' //DUMP: OMPDeclareVariantAttr{{.*}}device={arch(x86, x86_64)} //DUMP: DeclRefExpr{{.*}}Function{{.*}}foo_v3 //DUMP: DeclRefExpr{{.*}}ParmVar{{.*}}'I' @@ -54,9 +54,9 @@ void foo_v3(float *AAA, float *BBB, int *I) {return;} //DUMP: DeclRefExpr{{.*}}Function{{.*}}foo_v1 //DUMP: DeclRefExpr{{.*}}ParmVar{{.*}}'AAA' //DUMP: DeclRefExpr{{.*}}ParmVar{{.*}}'BBB' -//PRINT: #pragma omp declare variant(foo_v3) match(construct={dispatch}, device={arch(x86, x86_64)}) adjust_args(nothing:I) adjust_args(need_device_ptr:BBB) adjust_args(need_device_addr:AAA) +//PRINT: #pragma omp declare variant(foo_v3) match(construct={dispatch}, device={arch(x86, x86_64)}) adjust_args(nothing:I) adjust_args(need_device_ptr:BBB) adjust_args(need_device_addr:CCC) -//PRINT: #pragma omp declare variant(foo_v2) match(construct={dispatch}, device={arch(ppc)}) adjust_args(need_device_ptr:AAA) adjust_args(need_device_addr:BBB) +//PRINT: #pragma omp declare variant(foo_v2) match(construct={dispatch}, device={arch(ppc)}) adjust_args(need_device_ptr:AAA) adjust_args(need_device_addr:CCC) //PRINT: omp declare variant(foo_v1) match(construct={dispatch}, device={arch(arm)}) adjust_args(need_device_ptr:AAA,BBB) @@ -67,33 +67,33 @@ void foo_v3(float *AAA, float *BBB, int *I) {return;} #pragma omp declare variant(foo_v2) \ match(construct={dispatch}, device={arch(ppc)}), \ adjust_args(need_device_ptr:AAA) \ - adjust_args(need_device_addr:BBB) + adjust_args(need_device_addr:CCC) #pragma omp declare variant(foo_v3) \ adjust_args(need_device_ptr:BBB) adjust_args(nothing:I) \ - adjust_args(need_device_addr:AAA) \ + adjust_args(need_device_addr:CCC) \ match(construct={dispatch}, device={arch(x86,x86_64)}) -void foo(float *AAA, float *BBB, int *I) {return;} +void foo(float *AAA, float *BBB, int &CCC, int *I) {return;} -void Foo_Var(float *AAA, float *BBB, float *CCC) {return;} +void Foo_Var(float *AAA, float *BBB, float *&CCC) {return;} #pragma omp declare variant(Foo_Var) \ match(construct={dispatch}, device={arch(x86_64)}) \ adjust_args(need_device_ptr:AAA) adjust_args(nothing:BBB) \ adjust_args(need_device_addr:CCC) template<typename T> -void Foo(T *AAA, T *BBB, T *CCC) {return;} +void Foo(T *AAA, T *BBB, T *&CCC) {return;} //PRINT: #pragma omp declare variant(Foo_Var) match(construct={dispatch}, device={arch(x86_64)}) adjust_args(nothing:BBB) adjust_args(need_device_ptr:AAA) adjust_args(need_device_addr:CCC) -//DUMP: FunctionDecl{{.*}} Foo 'void (T *, T *, T *)' +//DUMP: FunctionDecl{{.*}} Foo 'void (T *, T *, T *&)' //DUMP: OMPDeclareVariantAttr{{.*}}device={arch(x86_64)} //DUMP: DeclRefExpr{{.*}}Function{{.*}}Foo_Var //DUMP: DeclRefExpr{{.*}}ParmVar{{.*}}'BBB' //DUMP: DeclRefExpr{{.*}}ParmVar{{.*}}'AAA' //DUMP: DeclRefExpr{{.*}}ParmVar{{.*}}'CCC' // -//DUMP: FunctionDecl{{.*}} Foo 'void (float *, float *, float *)' +//DUMP: FunctionDecl{{.*}} Foo 'void (float *, float *, float *&)' //DUMP: OMPDeclareVariantAttr{{.*}}device={arch(x86_64)} //DUMP: DeclRefExpr{{.*}}Function{{.*}}Foo_Var //DUMP: DeclRefExpr{{.*}}ParmVar{{.*}}'BBB' diff --git a/clang/test/OpenMP/declare_variant_clauses_messages.cpp b/clang/test/OpenMP/declare_variant_clauses_messages.cpp index bca9148..916d15f 100644 --- a/clang/test/OpenMP/declare_variant_clauses_messages.cpp +++ b/clang/test/OpenMP/declare_variant_clauses_messages.cpp @@ -91,6 +91,7 @@ void foo_v1(float *AAA, float *BBB, int *I) { return; } void foo_v2(float *AAA, float *BBB, int *I) { return; } void foo_v3(float *AAA, float *BBB, int *I) { return; } void foo_v4(float *AAA, float *BBB, int *I, omp_interop_t IOp) { return; } +void foo_v5(float *AAA, float *BBB, int I) { return; } #if _OPENMP >= 202011 // At least OpenMP 5.1 void vararg_foo(const char *fmt, omp_interop_t it, ...); @@ -129,6 +130,11 @@ void vararg_bar2(const char *fmt) { return; } adjust_args(nothing:J) \ match(construct={dispatch}, device={arch(x86,x86_64)}) +// expected-error@+2 {{expected reference type argument on 'adjust_args' clause with 'need_device_addr' modifier}} +#pragma omp declare variant(foo_v1) \ + adjust_args(need_device_addr:AAA) \ + match(construct={dispatch}, device={arch(x86,x86_64)}) + // expected-error@+2 {{expected reference to one of the parameters of function 'foo'}} #pragma omp declare variant(foo_v3) \ adjust_args(nothing:Other) \ @@ -218,6 +224,12 @@ void vararg_bar2(const char *fmt) { return; } void foo(float *AAA, float *BBB, int *I) { return; } +// expected-error@+2 {{expected reference type argument on 'adjust_args' clause with 'need_device_addr' modifier}} +#pragma omp declare variant(foo_v5) \ + adjust_args(need_device_addr:I) \ + match(construct={dispatch}, device={arch(x86,x86_64)}) +void foo5(float *AAA, float *BBB, int I) { return; } + #endif // NO_INTEROP_T_DEF #ifdef C diff --git a/clang/test/OpenMP/target_map_array_section_no_length_codegen.cpp b/clang/test/OpenMP/target_map_array_section_no_length_codegen.cpp new file mode 100644 index 0000000..43fd509 --- /dev/null +++ b/clang/test/OpenMP/target_map_array_section_no_length_codegen.cpp @@ -0,0 +1,361 @@ +// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --include-generated-funcs --replace-value-regex "__omp_offloading_[0-9a-z]+_[0-9a-z]+" "reduction_size[.].+[.]" "pl_cond[.].+[.|,]" --prefix-filecheck-ir-name _ --version 5 +// RUN: %clang_cc1 -verify -triple i386-unknown-unknown -fopenmp -fopenmp-version=60 -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s + +// RUN: %clang_cc1 -verify -triple i386-unknown-unknown -fopenmp -fopenmp-version=60 -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s +// RUN: %clang_cc1 -verify -triple i386-unknown-unknown -fopenmp -fopenmp-version=60 -fopenmp-targets=i386-pc-linux-gnu -include-pch %t -emit-llvm %s -o - | FileCheck %s + +// expected-no-diagnostics + +#ifndef HEADER +#define HEADER +void array_section_no_length_map_clause(float *d, int index) { + float **f; + + #pragma omp target map(tofrom : d[:]) + { + d[3] += 2; + } + + #pragma omp target map(to : d[2:]) + { + d[3] += 3; + } + + #pragma omp target map(alloc : f[index][:]) + { + f[index][2] += 4; + } + + #pragma omp target map(tofrom : f[index][index+1:]) + { + f[index][index] += 5; + } +} +#endif +// CHECK-LABEL: define dso_local void @_Z34array_section_no_length_map_clausePfi( +// CHECK-SAME: ptr noundef [[D:%.*]], i32 noundef [[INDEX:%.*]]) #[[ATTR0:[0-9]+]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[D_ADDR:%.*]] = alloca ptr, align 4 +// CHECK-NEXT: [[INDEX_ADDR:%.*]] = alloca i32, align 4 +// CHECK-NEXT: [[F:%.*]] = alloca ptr, align 4 +// CHECK-NEXT: [[DOTOFFLOAD_BASEPTRS:%.*]] = alloca [1 x ptr], align 4 +// CHECK-NEXT: [[DOTOFFLOAD_PTRS:%.*]] = alloca [1 x ptr], align 4 +// CHECK-NEXT: [[DOTOFFLOAD_MAPPERS:%.*]] = alloca [1 x ptr], align 4 +// CHECK-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8 +// CHECK-NEXT: [[DOTOFFLOAD_BASEPTRS2:%.*]] = alloca [1 x ptr], align 4 +// CHECK-NEXT: [[DOTOFFLOAD_PTRS3:%.*]] = alloca [1 x ptr], align 4 +// CHECK-NEXT: [[DOTOFFLOAD_MAPPERS4:%.*]] = alloca [1 x ptr], align 4 +// CHECK-NEXT: [[KERNEL_ARGS5:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8 +// CHECK-NEXT: [[INDEX_CASTED:%.*]] = alloca i32, align 4 +// CHECK-NEXT: [[DOTOFFLOAD_BASEPTRS11:%.*]] = alloca [3 x ptr], align 4 +// CHECK-NEXT: [[DOTOFFLOAD_PTRS12:%.*]] = alloca [3 x ptr], align 4 +// CHECK-NEXT: [[DOTOFFLOAD_MAPPERS13:%.*]] = alloca [3 x ptr], align 4 +// CHECK-NEXT: [[KERNEL_ARGS14:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8 +// CHECK-NEXT: [[INDEX_CASTED17:%.*]] = alloca i32, align 4 +// CHECK-NEXT: [[DOTOFFLOAD_BASEPTRS22:%.*]] = alloca [3 x ptr], align 4 +// CHECK-NEXT: [[DOTOFFLOAD_PTRS23:%.*]] = alloca [3 x ptr], align 4 +// CHECK-NEXT: [[DOTOFFLOAD_MAPPERS24:%.*]] = alloca [3 x ptr], align 4 +// CHECK-NEXT: [[DOTOFFLOAD_SIZES:%.*]] = alloca [3 x i64], align 4 +// CHECK-NEXT: [[KERNEL_ARGS25:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8 +// CHECK-NEXT: store ptr [[D]], ptr [[D_ADDR]], align 4 +// CHECK-NEXT: store i32 [[INDEX]], ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: [[TMP0:%.*]] = load ptr, ptr [[D_ADDR]], align 4 +// CHECK-NEXT: [[TMP1:%.*]] = load ptr, ptr [[D_ADDR]], align 4 +// CHECK-NEXT: [[TMP2:%.*]] = load ptr, ptr [[D_ADDR]], align 4 +// CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds nuw float, ptr [[TMP2]], i32 0 +// CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0 +// CHECK-NEXT: store ptr [[TMP1]], ptr [[TMP3]], align 4 +// CHECK-NEXT: [[TMP4:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0 +// CHECK-NEXT: store ptr [[ARRAYIDX]], ptr [[TMP4]], align 4 +// CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_MAPPERS]], i32 0, i32 0 +// CHECK-NEXT: store ptr null, ptr [[TMP5]], align 4 +// CHECK-NEXT: [[TMP6:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0 +// CHECK-NEXT: [[TMP7:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS]], i32 0, i32 0 +// CHECK-NEXT: [[TMP8:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 0 +// CHECK-NEXT: store i32 3, ptr [[TMP8]], align 4 +// CHECK-NEXT: [[TMP9:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 1 +// CHECK-NEXT: store i32 1, ptr [[TMP9]], align 4 +// CHECK-NEXT: [[TMP10:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 2 +// CHECK-NEXT: store ptr [[TMP6]], ptr [[TMP10]], align 4 +// CHECK-NEXT: [[TMP11:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 3 +// CHECK-NEXT: store ptr [[TMP7]], ptr [[TMP11]], align 4 +// CHECK-NEXT: [[TMP12:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 4 +// CHECK-NEXT: store ptr @.offload_sizes, ptr [[TMP12]], align 4 +// CHECK-NEXT: [[TMP13:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 5 +// CHECK-NEXT: store ptr @.offload_maptypes, ptr [[TMP13]], align 4 +// CHECK-NEXT: [[TMP14:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 6 +// CHECK-NEXT: store ptr null, ptr [[TMP14]], align 4 +// CHECK-NEXT: [[TMP15:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 7 +// CHECK-NEXT: store ptr null, ptr [[TMP15]], align 4 +// CHECK-NEXT: [[TMP16:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 8 +// CHECK-NEXT: store i64 0, ptr [[TMP16]], align 8 +// CHECK-NEXT: [[TMP17:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 9 +// CHECK-NEXT: store i64 0, ptr [[TMP17]], align 8 +// CHECK-NEXT: [[TMP18:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 10 +// CHECK-NEXT: store [3 x i32] [i32 -1, i32 0, i32 0], ptr [[TMP18]], align 4 +// CHECK-NEXT: [[TMP19:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 11 +// CHECK-NEXT: store [3 x i32] zeroinitializer, ptr [[TMP19]], align 4 +// CHECK-NEXT: [[TMP20:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS]], i32 0, i32 12 +// CHECK-NEXT: store i32 0, ptr [[TMP20]], align 4 +// CHECK-NEXT: [[TMP21:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1:[0-9]+]], i64 -1, i32 -1, i32 0, ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z34array_section_no_length_map_clausePfi_l14.region_id, ptr [[KERNEL_ARGS]]) +// CHECK-NEXT: [[TMP22:%.*]] = icmp ne i32 [[TMP21]], 0 +// CHECK-NEXT: br i1 [[TMP22]], label %[[OMP_OFFLOAD_FAILED:.*]], label %[[OMP_OFFLOAD_CONT:.*]] +// CHECK: [[OMP_OFFLOAD_FAILED]]: +// CHECK-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z34array_section_no_length_map_clausePfi_l14(ptr [[TMP0]]) #[[ATTR2:[0-9]+]] +// CHECK-NEXT: br label %[[OMP_OFFLOAD_CONT]] +// CHECK: [[OMP_OFFLOAD_CONT]]: +// CHECK-NEXT: [[TMP23:%.*]] = load ptr, ptr [[D_ADDR]], align 4 +// CHECK-NEXT: [[TMP24:%.*]] = load ptr, ptr [[D_ADDR]], align 4 +// CHECK-NEXT: [[TMP25:%.*]] = load ptr, ptr [[D_ADDR]], align 4 +// CHECK-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds nuw float, ptr [[TMP25]], i32 2 +// CHECK-NEXT: [[TMP26:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS2]], i32 0, i32 0 +// CHECK-NEXT: store ptr [[TMP24]], ptr [[TMP26]], align 4 +// CHECK-NEXT: [[TMP27:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS3]], i32 0, i32 0 +// CHECK-NEXT: store ptr [[ARRAYIDX1]], ptr [[TMP27]], align 4 +// CHECK-NEXT: [[TMP28:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_MAPPERS4]], i32 0, i32 0 +// CHECK-NEXT: store ptr null, ptr [[TMP28]], align 4 +// CHECK-NEXT: [[TMP29:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_BASEPTRS2]], i32 0, i32 0 +// CHECK-NEXT: [[TMP30:%.*]] = getelementptr inbounds [1 x ptr], ptr [[DOTOFFLOAD_PTRS3]], i32 0, i32 0 +// CHECK-NEXT: [[TMP31:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 0 +// CHECK-NEXT: store i32 3, ptr [[TMP31]], align 4 +// CHECK-NEXT: [[TMP32:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 1 +// CHECK-NEXT: store i32 1, ptr [[TMP32]], align 4 +// CHECK-NEXT: [[TMP33:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 2 +// CHECK-NEXT: store ptr [[TMP29]], ptr [[TMP33]], align 4 +// CHECK-NEXT: [[TMP34:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 3 +// CHECK-NEXT: store ptr [[TMP30]], ptr [[TMP34]], align 4 +// CHECK-NEXT: [[TMP35:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 4 +// CHECK-NEXT: store ptr @.offload_sizes.1, ptr [[TMP35]], align 4 +// CHECK-NEXT: [[TMP36:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 5 +// CHECK-NEXT: store ptr @.offload_maptypes.2, ptr [[TMP36]], align 4 +// CHECK-NEXT: [[TMP37:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 6 +// CHECK-NEXT: store ptr null, ptr [[TMP37]], align 4 +// CHECK-NEXT: [[TMP38:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 7 +// CHECK-NEXT: store ptr null, ptr [[TMP38]], align 4 +// CHECK-NEXT: [[TMP39:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 8 +// CHECK-NEXT: store i64 0, ptr [[TMP39]], align 8 +// CHECK-NEXT: [[TMP40:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 9 +// CHECK-NEXT: store i64 0, ptr [[TMP40]], align 8 +// CHECK-NEXT: [[TMP41:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 10 +// CHECK-NEXT: store [3 x i32] [i32 -1, i32 0, i32 0], ptr [[TMP41]], align 4 +// CHECK-NEXT: [[TMP42:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 11 +// CHECK-NEXT: store [3 x i32] zeroinitializer, ptr [[TMP42]], align 4 +// CHECK-NEXT: [[TMP43:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS5]], i32 0, i32 12 +// CHECK-NEXT: store i32 0, ptr [[TMP43]], align 4 +// CHECK-NEXT: [[TMP44:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 -1, i32 0, ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z34array_section_no_length_map_clausePfi_l19.region_id, ptr [[KERNEL_ARGS5]]) +// CHECK-NEXT: [[TMP45:%.*]] = icmp ne i32 [[TMP44]], 0 +// CHECK-NEXT: br i1 [[TMP45]], label %[[OMP_OFFLOAD_FAILED6:.*]], label %[[OMP_OFFLOAD_CONT7:.*]] +// CHECK: [[OMP_OFFLOAD_FAILED6]]: +// CHECK-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z34array_section_no_length_map_clausePfi_l19(ptr [[TMP23]]) #[[ATTR2]] +// CHECK-NEXT: br label %[[OMP_OFFLOAD_CONT7]] +// CHECK: [[OMP_OFFLOAD_CONT7]]: +// CHECK-NEXT: [[TMP46:%.*]] = load ptr, ptr [[F]], align 4 +// CHECK-NEXT: [[TMP47:%.*]] = load i32, ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: store i32 [[TMP47]], ptr [[INDEX_CASTED]], align 4 +// CHECK-NEXT: [[TMP48:%.*]] = load i32, ptr [[INDEX_CASTED]], align 4 +// CHECK-NEXT: [[TMP49:%.*]] = load ptr, ptr [[F]], align 4 +// CHECK-NEXT: [[TMP50:%.*]] = load ptr, ptr [[F]], align 4 +// CHECK-NEXT: [[TMP51:%.*]] = load i32, ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: [[ARRAYIDX8:%.*]] = getelementptr inbounds ptr, ptr [[TMP50]], i32 [[TMP51]] +// CHECK-NEXT: [[TMP52:%.*]] = load ptr, ptr [[F]], align 4 +// CHECK-NEXT: [[TMP53:%.*]] = load i32, ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: [[ARRAYIDX9:%.*]] = getelementptr inbounds ptr, ptr [[TMP52]], i32 [[TMP53]] +// CHECK-NEXT: [[TMP54:%.*]] = load ptr, ptr [[ARRAYIDX9]], align 4 +// CHECK-NEXT: [[ARRAYIDX10:%.*]] = getelementptr inbounds nuw float, ptr [[TMP54]], i32 0 +// CHECK-NEXT: [[TMP55:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS11]], i32 0, i32 0 +// CHECK-NEXT: store ptr [[TMP49]], ptr [[TMP55]], align 4 +// CHECK-NEXT: [[TMP56:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS12]], i32 0, i32 0 +// CHECK-NEXT: store ptr [[ARRAYIDX8]], ptr [[TMP56]], align 4 +// CHECK-NEXT: [[TMP57:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS13]], i32 0, i32 0 +// CHECK-NEXT: store ptr null, ptr [[TMP57]], align 4 +// CHECK-NEXT: [[TMP58:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS11]], i32 0, i32 1 +// CHECK-NEXT: store ptr [[ARRAYIDX8]], ptr [[TMP58]], align 4 +// CHECK-NEXT: [[TMP59:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS12]], i32 0, i32 1 +// CHECK-NEXT: store ptr [[ARRAYIDX10]], ptr [[TMP59]], align 4 +// CHECK-NEXT: [[TMP60:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS13]], i32 0, i32 1 +// CHECK-NEXT: store ptr null, ptr [[TMP60]], align 4 +// CHECK-NEXT: [[TMP61:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS11]], i32 0, i32 2 +// CHECK-NEXT: store i32 [[TMP48]], ptr [[TMP61]], align 4 +// CHECK-NEXT: [[TMP62:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS12]], i32 0, i32 2 +// CHECK-NEXT: store i32 [[TMP48]], ptr [[TMP62]], align 4 +// CHECK-NEXT: [[TMP63:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS13]], i32 0, i32 2 +// CHECK-NEXT: store ptr null, ptr [[TMP63]], align 4 +// CHECK-NEXT: [[TMP64:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS11]], i32 0, i32 0 +// CHECK-NEXT: [[TMP65:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS12]], i32 0, i32 0 +// CHECK-NEXT: [[TMP66:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 0 +// CHECK-NEXT: store i32 3, ptr [[TMP66]], align 4 +// CHECK-NEXT: [[TMP67:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 1 +// CHECK-NEXT: store i32 3, ptr [[TMP67]], align 4 +// CHECK-NEXT: [[TMP68:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 2 +// CHECK-NEXT: store ptr [[TMP64]], ptr [[TMP68]], align 4 +// CHECK-NEXT: [[TMP69:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 3 +// CHECK-NEXT: store ptr [[TMP65]], ptr [[TMP69]], align 4 +// CHECK-NEXT: [[TMP70:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 4 +// CHECK-NEXT: store ptr @.offload_sizes.3, ptr [[TMP70]], align 4 +// CHECK-NEXT: [[TMP71:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 5 +// CHECK-NEXT: store ptr @.offload_maptypes.4, ptr [[TMP71]], align 4 +// CHECK-NEXT: [[TMP72:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 6 +// CHECK-NEXT: store ptr null, ptr [[TMP72]], align 4 +// CHECK-NEXT: [[TMP73:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 7 +// CHECK-NEXT: store ptr null, ptr [[TMP73]], align 4 +// CHECK-NEXT: [[TMP74:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 8 +// CHECK-NEXT: store i64 0, ptr [[TMP74]], align 8 +// CHECK-NEXT: [[TMP75:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 9 +// CHECK-NEXT: store i64 0, ptr [[TMP75]], align 8 +// CHECK-NEXT: [[TMP76:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 10 +// CHECK-NEXT: store [3 x i32] [i32 -1, i32 0, i32 0], ptr [[TMP76]], align 4 +// CHECK-NEXT: [[TMP77:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 11 +// CHECK-NEXT: store [3 x i32] zeroinitializer, ptr [[TMP77]], align 4 +// CHECK-NEXT: [[TMP78:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS14]], i32 0, i32 12 +// CHECK-NEXT: store i32 0, ptr [[TMP78]], align 4 +// CHECK-NEXT: [[TMP79:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 -1, i32 0, ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z34array_section_no_length_map_clausePfi_l24.region_id, ptr [[KERNEL_ARGS14]]) +// CHECK-NEXT: [[TMP80:%.*]] = icmp ne i32 [[TMP79]], 0 +// CHECK-NEXT: br i1 [[TMP80]], label %[[OMP_OFFLOAD_FAILED15:.*]], label %[[OMP_OFFLOAD_CONT16:.*]] +// CHECK: [[OMP_OFFLOAD_FAILED15]]: +// CHECK-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z34array_section_no_length_map_clausePfi_l24(ptr [[TMP46]], i32 [[TMP48]]) #[[ATTR2]] +// CHECK-NEXT: br label %[[OMP_OFFLOAD_CONT16]] +// CHECK: [[OMP_OFFLOAD_CONT16]]: +// CHECK-NEXT: [[TMP81:%.*]] = load ptr, ptr [[F]], align 4 +// CHECK-NEXT: [[TMP82:%.*]] = load i32, ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: store i32 [[TMP82]], ptr [[INDEX_CASTED17]], align 4 +// CHECK-NEXT: [[TMP83:%.*]] = load i32, ptr [[INDEX_CASTED17]], align 4 +// CHECK-NEXT: [[TMP84:%.*]] = load ptr, ptr [[F]], align 4 +// CHECK-NEXT: [[TMP85:%.*]] = load ptr, ptr [[F]], align 4 +// CHECK-NEXT: [[TMP86:%.*]] = load i32, ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: [[ARRAYIDX18:%.*]] = getelementptr inbounds ptr, ptr [[TMP85]], i32 [[TMP86]] +// CHECK-NEXT: [[TMP87:%.*]] = load i32, ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP87]], 1 +// CHECK-NEXT: [[TMP88:%.*]] = load ptr, ptr [[F]], align 4 +// CHECK-NEXT: [[TMP89:%.*]] = load i32, ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: [[ARRAYIDX19:%.*]] = getelementptr inbounds ptr, ptr [[TMP88]], i32 [[TMP89]] +// CHECK-NEXT: [[TMP90:%.*]] = load ptr, ptr [[ARRAYIDX19]], align 4 +// CHECK-NEXT: [[ARRAYIDX20:%.*]] = getelementptr inbounds nuw float, ptr [[TMP90]], i32 [[ADD]] +// CHECK-NEXT: [[TMP91:%.*]] = load i32, ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: [[ADD21:%.*]] = add nsw i32 [[TMP91]], 1 +// CHECK-NEXT: [[TMP92:%.*]] = mul nuw i32 [[ADD21]], 4 +// CHECK-NEXT: [[TMP93:%.*]] = icmp ugt i32 4, [[TMP92]] +// CHECK-NEXT: [[TMP94:%.*]] = sub nuw i32 4, [[TMP92]] +// CHECK-NEXT: [[TMP95:%.*]] = select i1 [[TMP93]], i32 [[TMP94]], i32 0 +// CHECK-NEXT: [[TMP96:%.*]] = sext i32 [[TMP95]] to i64 +// CHECK-NEXT: call void @llvm.memcpy.p0.p0.i32(ptr align 4 [[DOTOFFLOAD_SIZES]], ptr align 4 @.offload_sizes.5, i32 24, i1 false) +// CHECK-NEXT: [[TMP97:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS22]], i32 0, i32 0 +// CHECK-NEXT: store ptr [[TMP84]], ptr [[TMP97]], align 4 +// CHECK-NEXT: [[TMP98:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS23]], i32 0, i32 0 +// CHECK-NEXT: store ptr [[ARRAYIDX18]], ptr [[TMP98]], align 4 +// CHECK-NEXT: [[TMP99:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS24]], i32 0, i32 0 +// CHECK-NEXT: store ptr null, ptr [[TMP99]], align 4 +// CHECK-NEXT: [[TMP100:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS22]], i32 0, i32 1 +// CHECK-NEXT: store ptr [[ARRAYIDX18]], ptr [[TMP100]], align 4 +// CHECK-NEXT: [[TMP101:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS23]], i32 0, i32 1 +// CHECK-NEXT: store ptr [[ARRAYIDX20]], ptr [[TMP101]], align 4 +// CHECK-NEXT: [[TMP102:%.*]] = getelementptr inbounds [3 x i64], ptr [[DOTOFFLOAD_SIZES]], i32 0, i32 1 +// CHECK-NEXT: store i64 [[TMP96]], ptr [[TMP102]], align 4 +// CHECK-NEXT: [[TMP103:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS24]], i32 0, i32 1 +// CHECK-NEXT: store ptr null, ptr [[TMP103]], align 4 +// CHECK-NEXT: [[TMP104:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS22]], i32 0, i32 2 +// CHECK-NEXT: store i32 [[TMP83]], ptr [[TMP104]], align 4 +// CHECK-NEXT: [[TMP105:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS23]], i32 0, i32 2 +// CHECK-NEXT: store i32 [[TMP83]], ptr [[TMP105]], align 4 +// CHECK-NEXT: [[TMP106:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_MAPPERS24]], i32 0, i32 2 +// CHECK-NEXT: store ptr null, ptr [[TMP106]], align 4 +// CHECK-NEXT: [[TMP107:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_BASEPTRS22]], i32 0, i32 0 +// CHECK-NEXT: [[TMP108:%.*]] = getelementptr inbounds [3 x ptr], ptr [[DOTOFFLOAD_PTRS23]], i32 0, i32 0 +// CHECK-NEXT: [[TMP109:%.*]] = getelementptr inbounds [3 x i64], ptr [[DOTOFFLOAD_SIZES]], i32 0, i32 0 +// CHECK-NEXT: [[TMP110:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 0 +// CHECK-NEXT: store i32 3, ptr [[TMP110]], align 4 +// CHECK-NEXT: [[TMP111:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 1 +// CHECK-NEXT: store i32 3, ptr [[TMP111]], align 4 +// CHECK-NEXT: [[TMP112:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 2 +// CHECK-NEXT: store ptr [[TMP107]], ptr [[TMP112]], align 4 +// CHECK-NEXT: [[TMP113:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 3 +// CHECK-NEXT: store ptr [[TMP108]], ptr [[TMP113]], align 4 +// CHECK-NEXT: [[TMP114:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 4 +// CHECK-NEXT: store ptr [[TMP109]], ptr [[TMP114]], align 4 +// CHECK-NEXT: [[TMP115:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 5 +// CHECK-NEXT: store ptr @.offload_maptypes.6, ptr [[TMP115]], align 4 +// CHECK-NEXT: [[TMP116:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 6 +// CHECK-NEXT: store ptr null, ptr [[TMP116]], align 4 +// CHECK-NEXT: [[TMP117:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 7 +// CHECK-NEXT: store ptr null, ptr [[TMP117]], align 4 +// CHECK-NEXT: [[TMP118:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 8 +// CHECK-NEXT: store i64 0, ptr [[TMP118]], align 8 +// CHECK-NEXT: [[TMP119:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 9 +// CHECK-NEXT: store i64 0, ptr [[TMP119]], align 8 +// CHECK-NEXT: [[TMP120:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 10 +// CHECK-NEXT: store [3 x i32] [i32 -1, i32 0, i32 0], ptr [[TMP120]], align 4 +// CHECK-NEXT: [[TMP121:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 11 +// CHECK-NEXT: store [3 x i32] zeroinitializer, ptr [[TMP121]], align 4 +// CHECK-NEXT: [[TMP122:%.*]] = getelementptr inbounds nuw [[STRUCT___TGT_KERNEL_ARGUMENTS]], ptr [[KERNEL_ARGS25]], i32 0, i32 12 +// CHECK-NEXT: store i32 0, ptr [[TMP122]], align 4 +// CHECK-NEXT: [[TMP123:%.*]] = call i32 @__tgt_target_kernel(ptr @[[GLOB1]], i64 -1, i32 -1, i32 0, ptr @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z34array_section_no_length_map_clausePfi_l29.region_id, ptr [[KERNEL_ARGS25]]) +// CHECK-NEXT: [[TMP124:%.*]] = icmp ne i32 [[TMP123]], 0 +// CHECK-NEXT: br i1 [[TMP124]], label %[[OMP_OFFLOAD_FAILED26:.*]], label %[[OMP_OFFLOAD_CONT27:.*]] +// CHECK: [[OMP_OFFLOAD_FAILED26]]: +// CHECK-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z34array_section_no_length_map_clausePfi_l29(ptr [[TMP81]], i32 [[TMP83]]) #[[ATTR2]] +// CHECK-NEXT: br label %[[OMP_OFFLOAD_CONT27]] +// CHECK: [[OMP_OFFLOAD_CONT27]]: +// CHECK-NEXT: ret void +// +// +// CHECK-LABEL: define internal void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z34array_section_no_length_map_clausePfi_l14( +// CHECK-SAME: ptr noundef [[D:%.*]]) #[[ATTR1:[0-9]+]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[D_ADDR:%.*]] = alloca ptr, align 4 +// CHECK-NEXT: store ptr [[D]], ptr [[D_ADDR]], align 4 +// CHECK-NEXT: [[TMP0:%.*]] = load ptr, ptr [[D_ADDR]], align 4 +// CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP0]], i32 3 +// CHECK-NEXT: [[TMP1:%.*]] = load float, ptr [[ARRAYIDX]], align 4 +// CHECK-NEXT: [[ADD:%.*]] = fadd float [[TMP1]], 2.000000e+00 +// CHECK-NEXT: store float [[ADD]], ptr [[ARRAYIDX]], align 4 +// CHECK-NEXT: ret void +// +// +// CHECK-LABEL: define internal void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z34array_section_no_length_map_clausePfi_l19( +// CHECK-SAME: ptr noundef [[D:%.*]]) #[[ATTR1]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[D_ADDR:%.*]] = alloca ptr, align 4 +// CHECK-NEXT: store ptr [[D]], ptr [[D_ADDR]], align 4 +// CHECK-NEXT: [[TMP0:%.*]] = load ptr, ptr [[D_ADDR]], align 4 +// CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[TMP0]], i32 3 +// CHECK-NEXT: [[TMP1:%.*]] = load float, ptr [[ARRAYIDX]], align 4 +// CHECK-NEXT: [[ADD:%.*]] = fadd float [[TMP1]], 3.000000e+00 +// CHECK-NEXT: store float [[ADD]], ptr [[ARRAYIDX]], align 4 +// CHECK-NEXT: ret void +// +// +// CHECK-LABEL: define internal void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z34array_section_no_length_map_clausePfi_l24( +// CHECK-SAME: ptr noundef [[F:%.*]], i32 noundef [[INDEX:%.*]]) #[[ATTR1]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[F_ADDR:%.*]] = alloca ptr, align 4 +// CHECK-NEXT: [[INDEX_ADDR:%.*]] = alloca i32, align 4 +// CHECK-NEXT: store ptr [[F]], ptr [[F_ADDR]], align 4 +// CHECK-NEXT: store i32 [[INDEX]], ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: [[TMP0:%.*]] = load ptr, ptr [[F_ADDR]], align 4 +// CHECK-NEXT: [[TMP1:%.*]] = load i32, ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP0]], i32 [[TMP1]] +// CHECK-NEXT: [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 4 +// CHECK-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds float, ptr [[TMP2]], i32 2 +// CHECK-NEXT: [[TMP3:%.*]] = load float, ptr [[ARRAYIDX1]], align 4 +// CHECK-NEXT: [[ADD:%.*]] = fadd float [[TMP3]], 4.000000e+00 +// CHECK-NEXT: store float [[ADD]], ptr [[ARRAYIDX1]], align 4 +// CHECK-NEXT: ret void +// +// +// CHECK-LABEL: define internal void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z34array_section_no_length_map_clausePfi_l29( +// CHECK-SAME: ptr noundef [[F:%.*]], i32 noundef [[INDEX:%.*]]) #[[ATTR1]] { +// CHECK-NEXT: [[ENTRY:.*:]] +// CHECK-NEXT: [[F_ADDR:%.*]] = alloca ptr, align 4 +// CHECK-NEXT: [[INDEX_ADDR:%.*]] = alloca i32, align 4 +// CHECK-NEXT: store ptr [[F]], ptr [[F_ADDR]], align 4 +// CHECK-NEXT: store i32 [[INDEX]], ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: [[TMP0:%.*]] = load ptr, ptr [[F_ADDR]], align 4 +// CHECK-NEXT: [[TMP1:%.*]] = load i32, ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds ptr, ptr [[TMP0]], i32 [[TMP1]] +// CHECK-NEXT: [[TMP2:%.*]] = load ptr, ptr [[ARRAYIDX]], align 4 +// CHECK-NEXT: [[TMP3:%.*]] = load i32, ptr [[INDEX_ADDR]], align 4 +// CHECK-NEXT: [[ARRAYIDX1:%.*]] = getelementptr inbounds float, ptr [[TMP2]], i32 [[TMP3]] +// CHECK-NEXT: [[TMP4:%.*]] = load float, ptr [[ARRAYIDX1]], align 4 +// CHECK-NEXT: [[ADD:%.*]] = fadd float [[TMP4]], 5.000000e+00 +// CHECK-NEXT: store float [[ADD]], ptr [[ARRAYIDX1]], align 4 +// CHECK-NEXT: ret void +// diff --git a/clang/test/OpenMP/target_map_messages.cpp b/clang/test/OpenMP/target_map_messages.cpp index 4a026584..0ee70be 100644 --- a/clang/test/OpenMP/target_map_messages.cpp +++ b/clang/test/OpenMP/target_map_messages.cpp @@ -122,9 +122,9 @@ struct SA { {} #pragma omp target map(always, tofrom: c,f[1:2]) {} - #pragma omp target map(always, tofrom: c[:],f) // expected-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} + #pragma omp target map(always, tofrom: c[:],f) // lt60-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} {} - #pragma omp target map(always, tofrom: c,f[:]) // expected-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} + #pragma omp target map(always, tofrom: c,f[:]) // lt60-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} {} #pragma omp target map(always) // expected-error {{use of undeclared identifier 'always'}} {} @@ -134,9 +134,9 @@ struct SA { {} #pragma omp target map(self, tofrom: c,f[1:2]) // lt60-error {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper'}} {} - #pragma omp target map(self, tofrom: c[:],f) // lt60-error {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper'}} // expected-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} + #pragma omp target map(self, tofrom: c[:],f) // lt60-error {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper'}} // lt60-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} {} - #pragma omp target map(self, tofrom: c,f[:]) // lt60-error {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper'}} // expected-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} + #pragma omp target map(self, tofrom: c,f[:]) // lt60-error {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper'}} // lt60-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} {} #pragma omp target map(close, tofrom: c,f) {} @@ -144,9 +144,9 @@ struct SA { {} #pragma omp target map(close, tofrom: c,f[1:2]) {} - #pragma omp target map(close, tofrom: c[:],f) // expected-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} + #pragma omp target map(close, tofrom: c[:],f) // lt60-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} {} - #pragma omp target map(close, tofrom: c,f[:]) // expected-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} + #pragma omp target map(close, tofrom: c,f[:]) // lt60-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} {} #pragma omp target map(close) // expected-error {{use of undeclared identifier 'close'}} {} @@ -159,11 +159,11 @@ struct SA { // lt51-error@+1 {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper'}} #pragma omp target map(present, tofrom: c,f[1:2]) {} - // expected-error@+2 {{section length is unspecified and cannot be inferred because subscripted value is not an array}} + // lt60-error@+2 {{section length is unspecified and cannot be inferred because subscripted value is not an array}} // lt51-error@+1 {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper'}} #pragma omp target map(present, tofrom: c[:],f) {} - // expected-error@+2 {{section length is unspecified and cannot be inferred because subscripted value is not an array}} + // lt60-error@+2 {{section length is unspecified and cannot be inferred because subscripted value is not an array}} // lt51-error@+1 {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper'}} #pragma omp target map(present, tofrom: c,f[:]) {} @@ -190,14 +190,14 @@ struct SA { {} // ge60-error@+5 {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper', 'present', 'iterator', 'self}} // ge52-error@+4 {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper', 'present', 'iterator'}} - // expected-error@+3 {{section length is unspecified and cannot be inferred because subscripted value is not an array}} + // lt60-error@+3 {{section length is unspecified and cannot be inferred because subscripted value is not an array}} // ge51-omp-error@+2 {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper', 'present'}} // lt51-omp-error@+1 {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper'}} #pragma omp target map(ompx_hold, tofrom: c[:],f) {} // ge60-error@+5 {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper', 'present', 'iterator', 'self}} // ge52-error@+4 {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper', 'present', 'iterator'}} - // expected-error@+3 {{section length is unspecified and cannot be inferred because subscripted value is not an array}} + // lt60-error@+3 {{section length is unspecified and cannot be inferred because subscripted value is not an array}} // ge51-omp-error@+2 {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper', 'present'}} // lt51-omp-error@+1 {{incorrect map type modifier, expected one of: 'always', 'close', 'mapper'}} #pragma omp target map(ompx_hold, tofrom: c,f[:]) @@ -448,7 +448,7 @@ void SAclient(int arg) { {} #pragma omp target map(mptr[:1][:2] [0:2]) // expected-error {{array section does not specify contiguous storage}} {} -#pragma omp target map(mptr[:1][:] [0:2]) // expected-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} +#pragma omp target map(mptr[:1][:] [0:2]) // lt60-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} {} #pragma omp target map(mptr[:2][:1] [0:2]) // expected-error {{array section does not specify contiguous storage}} {} @@ -517,7 +517,7 @@ void SAclient(int arg) { {} #pragma omp target map(r.S.Ptr [4:5]) {} -#pragma omp target map(r.S.Ptr[:]) // expected-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} +#pragma omp target map(r.S.Ptr[:]) // lt60-error {{section length is unspecified and cannot be inferred because subscripted value is not an array}} {} #pragma omp target map((p + 1)->A) // lt50-error {{expected expression containing only member accesses and/or array sections based on named variables}} {} diff --git a/clang/test/Preprocessor/arm-acle-6.4.c b/clang/test/Preprocessor/arm-acle-6.4.c index 2c8f486..48deba7 100644 --- a/clang/test/Preprocessor/arm-acle-6.4.c +++ b/clang/test/Preprocessor/arm-acle-6.4.c @@ -188,6 +188,37 @@ // RUN: %clang --target=arm-arm-none-eabi -mcpu=cortex-m33 -x c -E -dM %s -o - | FileCheck %s -check-prefix CHECK-M-DSP // RUN: %clang --target=arm-arm-none-eabi -march=armv8m.main+dsp -x c -E -dM %s -o - | FileCheck %s -check-prefix CHECK-M-DSP +// RUN: %clang -target arm-none-linux-eabi -march=armv8m.base -x c -E -dM %s -o - | FileCheck %s -check-prefix CHECK-V8M-BASE + +// CHECK-V8M-BASE-NOT: __ARM_ARCH_ISA_ARM +// CHECK-V8M-BASE-NOT: __ARM_FEATURE_DSP +// CHECK-V8M-BASE-NOT: __ARM_FEATURE_SIMD32 +// CHECK-V8M-BASE: __ARM_ARCH 8 +// CHECK-V8M-BASE: __ARM_ARCH_ISA_THUMB 1 +// CHECK-V8M-BASE: __ARM_ARCH_PROFILE 'M' +// CHECK-V8M-BASE: __ARM_FEATURE_CLZ 1 +// CHECK-V8M-BASE: __ARM_FEATURE_IDIV 1 +// CHECK-V8M-BASE: __ARM_FEATURE_LDREX 0x7 +// CHECK-V8M-BASE: __ARM_FEATURE_QBIT 1 +// CHECK-V8M-BASE: __ARM_FEATURE_SAT +// CHECK-V8M-BASE-NOT: __ARM_FEATURE_UNALIGNED + +// RUN: %clang -target arm-none-linux-eabi -march=armv8m.main -x c -E -dM %s -o - | FileCheck %s -check-prefix CHECK-V8M-MAIN +// RUN: %clang -target arm-none-linux-eabi -march=armv8.1m.main -x c -E -dM %s -o - | FileCheck %s -check-prefix CHECK-V8M-MAIN + +// CHECK-V8M-MAIN-NOT: __ARM_ARCH_ISA_ARM +// CHECK-V8M-MAIN-NOT: __ARM_FEATURE_DSP +// CHECK-V8M-MAIN-NOT: __ARM_FEATURE_SIMD32 +// CHECK-V8M-MAIN: __ARM_ARCH 8 +// CHECK-V8M-MAIN: __ARM_ARCH_ISA_THUMB 2 +// CHECK-V8M-MAIN: __ARM_ARCH_PROFILE 'M' +// CHECK-V8M-MAIN: __ARM_FEATURE_CLZ 1 +// CHECK-V8M-MAIN: __ARM_FEATURE_IDIV 1 +// CHECK-V8M-MAIN: __ARM_FEATURE_LDREX 0x7 +// CHECK-V8M-MAIN: __ARM_FEATURE_QBIT 1 +// CHECK-V8M-MAIN: __ARM_FEATURE_SAT 1 +// CHECK-V8M-MAIN: __ARM_FEATURE_UNALIGNED 1 + // CHECK-M-DSP: __ARM_FEATURE_DSP 1 // CHECK-M-DSP: __ARM_FEATURE_SIMD32 1 diff --git a/clang/test/Preprocessor/pragma-pushpop-macro-diag.c b/clang/test/Preprocessor/pragma-pushpop-macro-diag.c new file mode 100644 index 0000000..293cb82 --- /dev/null +++ b/clang/test/Preprocessor/pragma-pushpop-macro-diag.c @@ -0,0 +1,4 @@ +// RUN: %clang_cc1 -fms-extensions %s -fsyntax-only -verify + +#pragma push_macro("") // expected-warning {{'#pragma push_macro' expected a non-empty string}} +#pragma pop_macro("") // expected-warning {{'#pragma pop_macro' expected a non-empty string}} diff --git a/clang/test/Preprocessor/pragma-pushpop-macro.c b/clang/test/Preprocessor/pragma-pushpop-macro.c index 0aee074..238e3ed 100644 --- a/clang/test/Preprocessor/pragma-pushpop-macro.c +++ b/clang/test/Preprocessor/pragma-pushpop-macro.c @@ -56,3 +56,6 @@ int P; // CHECK: int pmy2 = 4 // CHECK: int Q; // CHECK: int P; + +#pragma push_macro("") +#pragma pop_macro("") diff --git a/clang/test/Sema/builtins-elementwise-math.c b/clang/test/Sema/builtins-elementwise-math.c index 01057b3..8548d3b 100644 --- a/clang/test/Sema/builtins-elementwise-math.c +++ b/clang/test/Sema/builtins-elementwise-math.c @@ -386,6 +386,96 @@ void test_builtin_elementwise_minimum(int i, short s, float f, double d, float4 // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was '_Complex float')}} } +void test_builtin_elementwise_maximumnum(int i, short s, float f, double d, float4 fv, double4 dv, int3 iv, unsigned3 uv, int *p) { + i = __builtin_elementwise_maximumnum(p, d); + // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was 'int *')}} + + struct Foo foo = __builtin_elementwise_maximumnum(d, d); + // expected-error@-1 {{initializing 'struct Foo' with an expression of incompatible type 'double'}} + + i = __builtin_elementwise_maximumnum(i); + // expected-error@-1 {{too few arguments to function call, expected 2, have 1}} + + i = __builtin_elementwise_maximumnum(); + // expected-error@-1 {{too few arguments to function call, expected 2, have 0}} + + i = __builtin_elementwise_maximumnum(i, i, i); + // expected-error@-1 {{too many arguments to function call, expected 2, have 3}} + + i = __builtin_elementwise_maximumnum(fv, iv); + // expected-error@-1 {{arguments are of different types ('float4' (vector of 4 'float' values) vs 'int3' (vector of 3 'int' values))}} + + i = __builtin_elementwise_maximumnum(uv, iv); + // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was 'unsigned3' (vector of 3 'unsigned int' values))}} + + dv = __builtin_elementwise_maximumnum(fv, dv); + // expected-error@-1 {{arguments are of different types ('float4' (vector of 4 'float' values) vs 'double4' (vector of 4 'double' values))}} + + d = __builtin_elementwise_maximumnum(f, d); + // expected-error@-1 {{arguments are of different types ('float' vs 'double')}} + + fv = __builtin_elementwise_maximumnum(fv, fv); + + i = __builtin_elementwise_maximumnum(iv, iv); + // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was 'int3' (vector of 3 'int' values))}} + + i = __builtin_elementwise_maximumnum(i, i); + // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was 'int')}} + + int A[10]; + A = __builtin_elementwise_maximumnum(A, A); + // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was 'int *')}} + + _Complex float c1, c2; + c1 = __builtin_elementwise_maximumnum(c1, c2); + // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was '_Complex float')}} +} + +void test_builtin_elementwise_minimumnum(int i, short s, float f, double d, float4 fv, double4 dv, int3 iv, unsigned3 uv, int *p) { + i = __builtin_elementwise_minimumnum(p, d); + // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was 'int *')}} + + struct Foo foo = __builtin_elementwise_minimumnum(d, d); + // expected-error@-1 {{initializing 'struct Foo' with an expression of incompatible type 'double'}} + + i = __builtin_elementwise_minimumnum(i); + // expected-error@-1 {{too few arguments to function call, expected 2, have 1}} + + i = __builtin_elementwise_minimumnum(); + // expected-error@-1 {{too few arguments to function call, expected 2, have 0}} + + i = __builtin_elementwise_minimumnum(i, i, i); + // expected-error@-1 {{too many arguments to function call, expected 2, have 3}} + + i = __builtin_elementwise_minimumnum(fv, iv); + // expected-error@-1 {{arguments are of different types ('float4' (vector of 4 'float' values) vs 'int3' (vector of 3 'int' values))}} + + i = __builtin_elementwise_minimumnum(uv, iv); + // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was 'unsigned3' (vector of 3 'unsigned int' values))}} + + dv = __builtin_elementwise_minimumnum(fv, dv); + // expected-error@-1 {{arguments are of different types ('float4' (vector of 4 'float' values) vs 'double4' (vector of 4 'double' values))}} + + d = __builtin_elementwise_minimumnum(f, d); + // expected-error@-1 {{arguments are of different types ('float' vs 'double')}} + + fv = __builtin_elementwise_minimumnum(fv, fv); + + i = __builtin_elementwise_minimumnum(iv, iv); + // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was 'int3' (vector of 3 'int' values))}} + + i = __builtin_elementwise_minimumnum(i, i); + // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was 'int')}} + + int A[10]; + A = __builtin_elementwise_minimumnum(A, A); + // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was 'int *')}} + + _Complex float c1, c2; + c1 = __builtin_elementwise_minimumnum(c1, c2); + // expected-error@-1 {{1st argument must be a scalar or vector of floating-point types (was '_Complex float')}} +} + void test_builtin_elementwise_bitreverse(int i, float f, double d, float4 v, int3 iv, unsigned u, unsigned4 uv) { struct Foo s = __builtin_elementwise_bitreverse(i); diff --git a/clang/test/Sema/warn-lifetime-safety-dataflow.cpp b/clang/test/Sema/warn-lifetime-safety-dataflow.cpp index 0e98904ade..2b934ac 100644 --- a/clang/test/Sema/warn-lifetime-safety-dataflow.cpp +++ b/clang/test/Sema/warn-lifetime-safety-dataflow.cpp @@ -1,4 +1,4 @@ -// RUN: %clang_cc1 -mllvm -debug-only=LifetimeFacts,LifetimeLoanPropagation -Wexperimental-lifetime-safety %s 2>&1 | FileCheck %s +// RUN: %clang_cc1 -fexperimental-lifetime-safety -mllvm -debug-only=LifetimeFacts -Wexperimental-lifetime-safety %s 2>&1 | FileCheck %s // REQUIRES: asserts struct MyObj { @@ -19,10 +19,6 @@ MyObj* return_local_addr() { // CHECK: ReturnOfOrigin (OriginID: [[O_RET_VAL]]) // CHECK: Expire (LoanID: [[L_X]]) } -// CHECK: LoanPropagation results: -// CHECK-DAG: Origin [[O_ADDR_X]] contains Loan [[L_X]] -// CHECK-DAG: Origin [[O_P]] contains Loan [[L_X]] -// CHECK-DAG: Origin [[O_RET_VAL]] contains Loan [[L_X]] // Pointer Assignment and Return @@ -47,15 +43,6 @@ MyObj* assign_and_return_local_addr() { // CHECK: ReturnOfOrigin (OriginID: [[O_PTR2_RVAL_2]]) // CHECK: Expire (LoanID: [[L_Y]]) } -// CHECK: LoanPropagation results: -// CHECK-DAG: Origin [[O_ADDR_Y]] contains Loan [[L_Y]] -// CHECK-DAG: Origin [[O_PTR1]] contains Loan [[L_Y]] -// CHECK-DAG: Origin [[O_PTR2]] contains Loan [[L_Y]] -// CHECK-DAG: Origin [[O_PTR1_RVAL]] contains Loan [[L_Y]] -// CHECK-DAG: Origin [[O_PTR1_RVAL_2]] contains Loan [[L_Y]] -// CHECK-DAG: Origin [[O_PTR2_RVAL]] contains Loan [[L_Y]] -// CHECK-DAG: Origin [[O_PTR2_RVAL_2]] contains Loan [[L_Y]] - // Return of Non-Pointer Type // CHECK-LABEL: Function: return_int_val @@ -65,8 +52,6 @@ int return_int_val() { return x; } // CHECK-NEXT: End of Block -// CHECK: LoanPropagation results: -// CHECK: <empty> // Loan Expiration (Automatic Variable, C++) @@ -79,9 +64,6 @@ void loan_expires_cpp() { // CHECK: AssignOrigin (DestID: [[O_POBJ:[0-9]+]], SrcID: [[O_ADDR_OBJ]]) // CHECK: Expire (LoanID: [[L_OBJ]]) } -// CHECK: LoanPropagation results: -// CHECK-DAG: Origin [[O_ADDR_OBJ]] contains Loan [[L_OBJ]] -// CHECK-DAG: Origin [[O_POBJ]] contains Loan [[L_OBJ]] // FIXME: No expire for Trivial Destructors @@ -96,10 +78,6 @@ void loan_expires_trivial() { // CHECK-NEXT: End of Block // FIXME: Add check for Expire once trivial destructors are handled for expiration. } -// CHECK: LoanPropagation results: -// CHECK-DAG: Origin [[O_ADDR_TRIVIAL_OBJ]] contains Loan [[L_TRIVIAL_OBJ]] -// CHECK-DAG: Origin [[O_PTOBJ]] contains Loan [[L_TRIVIAL_OBJ]] - // CHECK-LABEL: Function: conditional void conditional(bool condition) { @@ -119,13 +97,6 @@ void conditional(bool condition) { // CHECK: AssignOrigin (DestID: [[O_P_RVAL:[0-9]+]], SrcID: [[O_P]]) // CHECK: AssignOrigin (DestID: [[O_Q:[0-9]+]], SrcID: [[O_P_RVAL]]) } -// CHECK: LoanPropagation results: -// CHECK-DAG: Origin [[O_ADDR_A]] contains Loan [[L_A]] -// CHECK-DAG: Origin [[O_ADDR_B]] contains Loan [[L_B]] -// CHECK-DAG: Origin [[O_P]] contains Loan [[L_A]] -// CHECK-DAG: Origin [[O_P]] contains Loan [[L_B]] -// CHECK-DAG: Origin [[O_Q]] contains Loan [[L_A]] -// CHECK-DAG: Origin [[O_Q]] contains Loan [[L_B]] // CHECK-LABEL: Function: pointers_in_a_cycle @@ -161,25 +132,6 @@ void pointers_in_a_cycle(bool condition) { // CHECK: AssignOrigin (DestID: [[O_P3]], SrcID: [[O_TEMP_RVAL]]) } } -// At the end of the analysis, the origins for the pointers involved in the cycle -// (p1, p2, p3, temp) should all contain the loans from v1, v2, and v3 at the fixed point. -// CHECK: LoanPropagation results: -// CHECK-DAG: Origin [[O_P1]] contains Loan [[L_V1]] -// CHECK-DAG: Origin [[O_P1]] contains Loan [[L_V2]] -// CHECK-DAG: Origin [[O_P1]] contains Loan [[L_V3]] -// CHECK-DAG: Origin [[O_P2]] contains Loan [[L_V1]] -// CHECK-DAG: Origin [[O_P2]] contains Loan [[L_V2]] -// CHECK-DAG: Origin [[O_P2]] contains Loan [[L_V3]] -// CHECK-DAG: Origin [[O_P3]] contains Loan [[L_V1]] -// CHECK-DAG: Origin [[O_P3]] contains Loan [[L_V2]] -// CHECK-DAG: Origin [[O_P3]] contains Loan [[L_V3]] -// CHECK-DAG: Origin [[O_TEMP]] contains Loan [[L_V1]] -// CHECK-DAG: Origin [[O_TEMP]] contains Loan [[L_V2]] -// CHECK-DAG: Origin [[O_TEMP]] contains Loan [[L_V3]] -// CHECK-DAG: Origin [[O_ADDR_V1]] contains Loan [[L_V1]] -// CHECK-DAG: Origin [[O_ADDR_V2]] contains Loan [[L_V2]] -// CHECK-DAG: Origin [[O_ADDR_V3]] contains Loan [[L_V3]] - // CHECK-LABEL: Function: overwrite_origin void overwrite_origin() { @@ -195,10 +147,6 @@ void overwrite_origin() { // CHECK: Expire (LoanID: [[L_S2]]) // CHECK: Expire (LoanID: [[L_S1]]) } -// CHECK: LoanPropagation results: -// CHECK: Origin [[O_P]] contains Loan [[L_S2]] -// CHECK-NOT: Origin [[O_P]] contains Loan [[L_S1]] - // CHECK-LABEL: Function: reassign_to_null void reassign_to_null() { @@ -213,8 +161,6 @@ void reassign_to_null() { } // FIXME: Have a better representation for nullptr than just an empty origin. // It should be a separate loan and origin kind. -// CHECK: LoanPropagation results: -// CHECK: Origin [[O_P]] contains no loans // CHECK-LABEL: Function: reassign_in_if @@ -235,11 +181,6 @@ void reassign_in_if(bool condition) { // CHECK: Expire (LoanID: [[L_S2]]) // CHECK: Expire (LoanID: [[L_S1]]) } -// CHECK: LoanPropagation results: -// CHECK-DAG: Origin [[O_P]] contains Loan [[L_S1]] -// CHECK-DAG: Origin [[O_P]] contains Loan [[L_S2]] -// CHECK-DAG: Origin [[O_ADDR_S1]] contains Loan [[L_S1]] -// CHECK-DAG: Origin [[O_ADDR_S2]] contains Loan [[L_S2]] // CHECK-LABEL: Function: assign_in_switch @@ -276,14 +217,6 @@ void assign_in_switch(int mode) { // CHECK-DAG: Expire (LoanID: [[L_S2]]) // CHECK-DAG: Expire (LoanID: [[L_S1]]) } -// CHECK: LoanPropagation results: -// CHECK-DAG: Origin [[O_P]] contains Loan [[L_S1]] -// CHECK-DAG: Origin [[O_P]] contains Loan [[L_S2]] -// CHECK-DAG: Origin [[O_P]] contains Loan [[L_S3]] -// CHECK-DAG: Origin [[O_ADDR_S1]] contains Loan [[L_S1]] -// CHECK-DAG: Origin [[O_ADDR_S2]] contains Loan [[L_S2]] -// CHECK-DAG: Origin [[O_ADDR_S3]] contains Loan [[L_S3]] - // CHECK-LABEL: Function: loan_in_loop void loan_in_loop(bool condition) { @@ -299,10 +232,6 @@ void loan_in_loop(bool condition) { // CHECK: Expire (LoanID: [[L_INNER]]) } } -// CHECK: LoanPropagation results: -// CHECK-DAG: Origin [[O_P]] contains Loan [[L_INNER]] -// CHECK-DAG: Origin [[O_ADDR_INNER]] contains Loan [[L_INNER]] - // CHECK-LABEL: Function: loop_with_break void loop_with_break(int count) { @@ -326,13 +255,6 @@ void loop_with_break(int count) { // CHECK: Expire (LoanID: [[L_S1]]) } -// CHECK-LABEL: LoanPropagation results: -// CHECK-DAG: Origin [[O_P]] contains Loan [[L_S1]] -// CHECK-DAG: Origin [[O_P]] contains Loan [[L_S2]] -// CHECK-DAG: Origin [[O_ADDR_S1]] contains Loan [[L_S1]] -// CHECK-DAG: Origin [[O_ADDR_S2]] contains Loan [[L_S2]] - - // CHECK-LABEL: Function: nested_scopes void nested_scopes() { MyObj* p = nullptr; @@ -355,13 +277,6 @@ void nested_scopes() { // CHECK: Expire (LoanID: [[L_OUTER]]) } -// CHECK-LABEL: LoanPropagation results: -// CHECK-DAG: Origin [[O_P]] contains Loan [[L_INNER]] -// CHECK-DAG: Origin [[O_ADDR_INNER]] contains Loan [[L_INNER]] -// CHECK-DAG: Origin [[O_ADDR_OUTER]] contains Loan [[L_OUTER]] -// CHECK-NOT: Origin [[O_P]] contains Loan [[L_OUTER]] - - // CHECK-LABEL: Function: pointer_indirection void pointer_indirection() { int a; diff --git a/clang/test/SemaCXX/attr-target-clones-riscv.cpp b/clang/test/SemaCXX/attr-target-clones-riscv.cpp index 102bb4b..7648284 100644 --- a/clang/test/SemaCXX/attr-target-clones-riscv.cpp +++ b/clang/test/SemaCXX/attr-target-clones-riscv.cpp @@ -9,6 +9,9 @@ void __attribute__((target_clones("default", "mtune=sifive-u74"))) mtune() {} // expected-warning@+1 {{version list contains duplicate entries}} void __attribute__((target_clones("default", "arch=+c", "arch=+c"))) dupVersion() {} +// expected-warning@+1 {{version list contains duplicate entries}} +void __attribute__((target_clones(" default", "default "))) dupDefault() {} + // expected-warning@+1 {{unsupported '' in the 'target_clones' attribute string; 'target_clones' attribute ignored}} void __attribute__((target_clones("default", ""))) emptyVersion() {} diff --git a/clang/test/SemaCXX/cxx2a-ms-no-unique-address.cpp b/clang/test/SemaCXX/cxx2a-ms-no-unique-address.cpp index 726cfa3..a064401 100644 --- a/clang/test/SemaCXX/cxx2a-ms-no-unique-address.cpp +++ b/clang/test/SemaCXX/cxx2a-ms-no-unique-address.cpp @@ -1,5 +1,6 @@ // RUN: %clang_cc1 -std=c++2a %s -verify=unsupported -triple x86_64-linux-gnu // RUN: %clang_cc1 -std=c++2a %s -verify -triple x86_64-windows -fms-compatibility +// RUN: %clang_cc1 -std=c++2a %s -verify -triple x86_64-uefi -fms-compatibility [[msvc::no_unique_address]] int a; // expected-error {{only applies to non-bit-field non-static data members}} unsupported-warning {{unknown}} [[msvc::no_unique_address]] void f(); // expected-error {{only applies to non-bit-field non-static data members}} unsupported-warning {{unknown}} diff --git a/clang/test/SemaTemplate/concepts.cpp b/clang/test/SemaTemplate/concepts.cpp index 62a4f95..663bc98 100644 --- a/clang/test/SemaTemplate/concepts.cpp +++ b/clang/test/SemaTemplate/concepts.cpp @@ -1250,3 +1250,11 @@ static_assert(!D<Priv>::has, "Private should be invisible."); static_assert(!D<Prot>::has, "Protected should be invisible."); } + + +namespace GH149986 { +template <typename T> concept PerfectSquare = [](){} // expected-note 2{{here}} +([](auto) { return true; }) < PerfectSquare <class T>; +// expected-error@-1 {{declaration of 'T' shadows template parameter}} \ +// expected-error@-1 {{a concept definition cannot refer to itself}} +} diff --git a/clang/tools/driver/cc1_main.cpp b/clang/tools/driver/cc1_main.cpp index 2c17f28..9b1e390 100644 --- a/clang/tools/driver/cc1_main.cpp +++ b/clang/tools/driver/cc1_main.cpp @@ -304,7 +304,7 @@ int cc1_main(ArrayRef<const char *> Argv, const char *Argv0, void *MainAddr) { *IOFile << "{\n"; llvm::TimerGroup::printAllJSONValues(*IOFile, ""); *IOFile << "\n}\n"; - } else { + } else if (!Clang->getCodeGenOpts().TimePassesStatsFile) { llvm::TimerGroup::printAll(*IOFile); } llvm::TimerGroup::clearAll(); diff --git a/clang/unittests/Analysis/CMakeLists.txt b/clang/unittests/Analysis/CMakeLists.txt index 059a748..52e7d28 100644 --- a/clang/unittests/Analysis/CMakeLists.txt +++ b/clang/unittests/Analysis/CMakeLists.txt @@ -4,6 +4,7 @@ add_clang_unittest(ClangAnalysisTests CloneDetectionTest.cpp ExprMutationAnalyzerTest.cpp IntervalPartitionTest.cpp + LifetimeSafetyTest.cpp MacroExpansionContextTest.cpp UnsafeBufferUsageTest.cpp CLANG_LIBS diff --git a/clang/unittests/Analysis/LifetimeSafetyTest.cpp b/clang/unittests/Analysis/LifetimeSafetyTest.cpp new file mode 100644 index 0000000..a48dc45 --- /dev/null +++ b/clang/unittests/Analysis/LifetimeSafetyTest.cpp @@ -0,0 +1,710 @@ +//===- LifetimeSafetyTest.cpp - Lifetime Safety Tests -*---------- C++-*-===// +// +// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. +// See https://llvm.org/LICENSE.txt for license information. +// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception +// +//===----------------------------------------------------------------------===// + +#include "clang/Analysis/Analyses/LifetimeSafety.h" +#include "clang/ASTMatchers/ASTMatchFinder.h" +#include "clang/ASTMatchers/ASTMatchers.h" +#include "clang/Testing/TestAST.h" +#include "llvm/ADT/StringMap.h" +#include "gmock/gmock.h" +#include "gtest/gtest.h" +#include <optional> +#include <vector> + +namespace clang::lifetimes::internal { +namespace { + +using namespace ast_matchers; +using ::testing::UnorderedElementsAreArray; + +// A helper class to run the full lifetime analysis on a piece of code +// and provide an interface for querying the results. +class LifetimeTestRunner { +public: + LifetimeTestRunner(llvm::StringRef Code) { + std::string FullCode = R"( + #define POINT(name) void("__lifetime_test_point_" #name) + struct MyObj { ~MyObj() {} int i; }; + )"; + FullCode += Code.str(); + + AST = std::make_unique<clang::TestAST>(FullCode); + ASTCtx = &AST->context(); + + // Find the target function using AST matchers. + auto MatchResult = + match(functionDecl(hasName("target")).bind("target"), *ASTCtx); + auto *FD = selectFirst<FunctionDecl>("target", MatchResult); + if (!FD) { + ADD_FAILURE() << "Test case must have a function named 'target'"; + return; + } + AnalysisCtx = std::make_unique<AnalysisDeclContext>(nullptr, FD); + CFG::BuildOptions &BuildOptions = AnalysisCtx->getCFGBuildOptions(); + BuildOptions.setAllAlwaysAdd(); + BuildOptions.AddImplicitDtors = true; + BuildOptions.AddTemporaryDtors = true; + + // Run the main analysis. + Analysis = std::make_unique<LifetimeSafetyAnalysis>(*AnalysisCtx); + Analysis->run(); + + AnnotationToPointMap = Analysis->getTestPoints(); + } + + LifetimeSafetyAnalysis &getAnalysis() { return *Analysis; } + ASTContext &getASTContext() { return *ASTCtx; } + + ProgramPoint getProgramPoint(llvm::StringRef Annotation) { + auto It = AnnotationToPointMap.find(Annotation); + if (It == AnnotationToPointMap.end()) { + ADD_FAILURE() << "Annotation '" << Annotation << "' not found."; + return nullptr; + } + return It->second; + } + +private: + std::unique_ptr<TestAST> AST; + ASTContext *ASTCtx = nullptr; + std::unique_ptr<AnalysisDeclContext> AnalysisCtx; + std::unique_ptr<LifetimeSafetyAnalysis> Analysis; + llvm::StringMap<ProgramPoint> AnnotationToPointMap; +}; + +// A convenience wrapper that uses the LifetimeSafetyAnalysis public API. +class LifetimeTestHelper { +public: + LifetimeTestHelper(LifetimeTestRunner &Runner) + : Runner(Runner), Analysis(Runner.getAnalysis()) {} + + std::optional<OriginID> getOriginForDecl(llvm::StringRef VarName) { + auto *VD = findDecl<ValueDecl>(VarName); + if (!VD) + return std::nullopt; + auto OID = Analysis.getOriginIDForDecl(VD); + if (!OID) + ADD_FAILURE() << "Origin for '" << VarName << "' not found."; + return OID; + } + + std::optional<LoanID> getLoanForVar(llvm::StringRef VarName) { + auto *VD = findDecl<VarDecl>(VarName); + if (!VD) + return std::nullopt; + std::vector<LoanID> LID = Analysis.getLoanIDForVar(VD); + if (LID.empty()) { + ADD_FAILURE() << "Loan for '" << VarName << "' not found."; + return std::nullopt; + } + // TODO: Support retrieving more than one loans to a var. + if (LID.size() > 1) { + ADD_FAILURE() << "More than 1 loans found for '" << VarName; + return std::nullopt; + } + return LID[0]; + } + + std::optional<LoanSet> getLoansAtPoint(OriginID OID, + llvm::StringRef Annotation) { + ProgramPoint PP = Runner.getProgramPoint(Annotation); + if (!PP) + return std::nullopt; + return Analysis.getLoansAtPoint(OID, PP); + } + + std::optional<LoanSet> getExpiredLoansAtPoint(llvm::StringRef Annotation) { + ProgramPoint PP = Runner.getProgramPoint(Annotation); + if (!PP) + return std::nullopt; + return Analysis.getExpiredLoansAtPoint(PP); + } + +private: + template <typename DeclT> DeclT *findDecl(llvm::StringRef Name) { + auto &Ctx = Runner.getASTContext(); + auto Results = match(valueDecl(hasName(Name)).bind("v"), Ctx); + if (Results.empty()) { + ADD_FAILURE() << "Declaration '" << Name << "' not found in AST."; + return nullptr; + } + return const_cast<DeclT *>(selectFirst<DeclT>("v", Results)); + } + + LifetimeTestRunner &Runner; + LifetimeSafetyAnalysis &Analysis; +}; + +// ========================================================================= // +// GTest Matchers & Fixture +// ========================================================================= // + +// A helper class to represent a set of loans, identified by variable names. +class LoanSetInfo { +public: + LoanSetInfo(const std::vector<std::string> &Vars, LifetimeTestHelper &H) + : LoanVars(Vars), Helper(H) {} + std::vector<std::string> LoanVars; + LifetimeTestHelper &Helper; +}; + +// It holds the name of the origin variable and a reference to the helper. +class OriginInfo { +public: + OriginInfo(llvm::StringRef OriginVar, LifetimeTestHelper &Helper) + : OriginVar(OriginVar), Helper(Helper) {} + llvm::StringRef OriginVar; + LifetimeTestHelper &Helper; +}; + +/// Matcher to verify the set of loans held by an origin at a specific +/// program point. +/// +/// This matcher is intended to be used with an \c OriginInfo object. +/// +/// \param LoanVars A vector of strings, where each string is the name of a +/// variable expected to be the source of a loan. +/// \param Annotation A string identifying the program point (created with +/// POINT()) where the check should be performed. +MATCHER_P2(HasLoansToImpl, LoanVars, Annotation, "") { + const OriginInfo &Info = arg; + std::optional<OriginID> OIDOpt = Info.Helper.getOriginForDecl(Info.OriginVar); + if (!OIDOpt) { + *result_listener << "could not find origin for '" << Info.OriginVar.str() + << "'"; + return false; + } + + std::optional<LoanSet> ActualLoansSetOpt = + Info.Helper.getLoansAtPoint(*OIDOpt, Annotation); + if (!ActualLoansSetOpt) { + *result_listener << "could not get a valid loan set at point '" + << Annotation << "'"; + return false; + } + std::vector<LoanID> ActualLoans(ActualLoansSetOpt->begin(), + ActualLoansSetOpt->end()); + + std::vector<LoanID> ExpectedLoans; + for (const auto &LoanVar : LoanVars) { + std::optional<LoanID> ExpectedLIDOpt = Info.Helper.getLoanForVar(LoanVar); + if (!ExpectedLIDOpt) { + *result_listener << "could not find loan for var '" << LoanVar << "'"; + return false; + } + ExpectedLoans.push_back(*ExpectedLIDOpt); + } + + return ExplainMatchResult(UnorderedElementsAreArray(ExpectedLoans), + ActualLoans, result_listener); +} + +/// Matcher to verify that the complete set of expired loans at a program point +/// matches the expected loan set. +MATCHER_P(AreExpiredAt, Annotation, "") { + const LoanSetInfo &Info = arg; + auto &Helper = Info.Helper; + + auto ActualExpiredSetOpt = Helper.getExpiredLoansAtPoint(Annotation); + if (!ActualExpiredSetOpt) { + *result_listener << "could not get a valid expired loan set at point '" + << Annotation << "'"; + return false; + } + std::vector<LoanID> ActualExpiredLoans(ActualExpiredSetOpt->begin(), + ActualExpiredSetOpt->end()); + std::vector<LoanID> ExpectedExpiredLoans; + for (const auto &VarName : Info.LoanVars) { + auto LoanIDOpt = Helper.getLoanForVar(VarName); + if (!LoanIDOpt) { + *result_listener << "could not find a loan for variable '" << VarName + << "'"; + return false; + } + ExpectedExpiredLoans.push_back(*LoanIDOpt); + } + return ExplainMatchResult(UnorderedElementsAreArray(ExpectedExpiredLoans), + ActualExpiredLoans, result_listener); +} + +// Base test fixture to manage the runner and helper. +class LifetimeAnalysisTest : public ::testing::Test { +protected: + void SetupTest(llvm::StringRef Code) { + Runner = std::make_unique<LifetimeTestRunner>(Code); + Helper = std::make_unique<LifetimeTestHelper>(*Runner); + } + + OriginInfo Origin(llvm::StringRef OriginVar) { + return OriginInfo(OriginVar, *Helper); + } + + /// Factory function that hides the std::vector creation. + LoanSetInfo LoansTo(std::initializer_list<std::string> LoanVars) { + return LoanSetInfo({LoanVars}, *Helper); + } + + /// A convenience helper for asserting that no loans are expired. + LoanSetInfo NoLoans() { return LoansTo({}); } + + // Factory function that hides the std::vector creation. + auto HasLoansTo(std::initializer_list<std::string> LoanVars, + const char *Annotation) { + return HasLoansToImpl(std::vector<std::string>(LoanVars), Annotation); + } + + std::unique_ptr<LifetimeTestRunner> Runner; + std::unique_ptr<LifetimeTestHelper> Helper; +}; + +// ========================================================================= // +// TESTS +// ========================================================================= // + +TEST_F(LifetimeAnalysisTest, SimpleLoanAndOrigin) { + SetupTest(R"( + void target() { + int x; + int* p = &x; + POINT(p1); + } + )"); + EXPECT_THAT(Origin("p"), HasLoansTo({"x"}, "p1")); +} + +TEST_F(LifetimeAnalysisTest, OverwriteOrigin) { + SetupTest(R"( + void target() { + MyObj s1, s2; + + MyObj* p = &s1; + POINT(after_s1); + + p = &s2; + POINT(after_s2); + } + )"); + EXPECT_THAT(Origin("p"), HasLoansTo({"s1"}, "after_s1")); + EXPECT_THAT(Origin("p"), HasLoansTo({"s2"}, "after_s2")); +} + +TEST_F(LifetimeAnalysisTest, ConditionalLoan) { + SetupTest(R"( + void target(bool cond) { + int a, b; + int *p = nullptr; + if (cond) { + p = &a; + POINT(after_then); + } else { + p = &b; + POINT(after_else); + } + POINT(after_if); + } + )"); + EXPECT_THAT(Origin("p"), HasLoansTo({"a"}, "after_then")); + EXPECT_THAT(Origin("p"), HasLoansTo({"b"}, "after_else")); + EXPECT_THAT(Origin("p"), HasLoansTo({"a", "b"}, "after_if")); +} + +TEST_F(LifetimeAnalysisTest, PointerChain) { + SetupTest(R"( + void target() { + MyObj y; + MyObj* ptr1 = &y; + POINT(p1); + + MyObj* ptr2 = ptr1; + POINT(p2); + + ptr2 = ptr1; + POINT(p3); + + ptr2 = ptr2; // Self assignment + POINT(p4); + } + )"); + EXPECT_THAT(Origin("ptr1"), HasLoansTo({"y"}, "p1")); + EXPECT_THAT(Origin("ptr2"), HasLoansTo({"y"}, "p2")); + EXPECT_THAT(Origin("ptr2"), HasLoansTo({"y"}, "p3")); + EXPECT_THAT(Origin("ptr2"), HasLoansTo({"y"}, "p4")); +} + +TEST_F(LifetimeAnalysisTest, ReassignToNull) { + SetupTest(R"( + void target() { + MyObj s1; + MyObj* p = &s1; + POINT(before_null); + p = nullptr; + POINT(after_null); + } + )"); + EXPECT_THAT(Origin("p"), HasLoansTo({"s1"}, "before_null")); + EXPECT_THAT(Origin("p"), HasLoansTo({}, "after_null")); +} + +TEST_F(LifetimeAnalysisTest, ReassignInIf) { + SetupTest(R"( + void target(bool condition) { + MyObj s1, s2; + MyObj* p = &s1; + POINT(before_if); + if (condition) { + p = &s2; + POINT(after_reassign); + } + POINT(after_if); + } + )"); + EXPECT_THAT(Origin("p"), HasLoansTo({"s1"}, "before_if")); + EXPECT_THAT(Origin("p"), HasLoansTo({"s2"}, "after_reassign")); + EXPECT_THAT(Origin("p"), HasLoansTo({"s1", "s2"}, "after_if")); +} + +TEST_F(LifetimeAnalysisTest, AssignInSwitch) { + SetupTest(R"( + void target(int mode) { + MyObj s1, s2, s3; + MyObj* p = nullptr; + switch (mode) { + case 1: + p = &s1; + POINT(case1); + break; + case 2: + p = &s2; + POINT(case2); + break; + default: + p = &s3; + POINT(case3); + break; + } + POINT(after_switch); + } + )"); + EXPECT_THAT(Origin("p"), HasLoansTo({"s1"}, "case1")); + EXPECT_THAT(Origin("p"), HasLoansTo({"s2"}, "case2")); + EXPECT_THAT(Origin("p"), HasLoansTo({"s3"}, "case3")); + EXPECT_THAT(Origin("p"), HasLoansTo({"s1", "s2", "s3"}, "after_switch")); +} + +TEST_F(LifetimeAnalysisTest, LoanInLoop) { + SetupTest(R"( + void target(bool condition) { + MyObj* p = nullptr; + while (condition) { + POINT(start_loop); + MyObj inner; + p = &inner; + POINT(end_loop); + } + POINT(after_loop); + } + )"); + EXPECT_THAT(Origin("p"), HasLoansTo({"inner"}, "start_loop")); + EXPECT_THAT(LoansTo({"inner"}), AreExpiredAt("start_loop")); + + EXPECT_THAT(Origin("p"), HasLoansTo({"inner"}, "end_loop")); + EXPECT_THAT(NoLoans(), AreExpiredAt("end_loop")); + + EXPECT_THAT(Origin("p"), HasLoansTo({"inner"}, "after_loop")); + EXPECT_THAT(LoansTo({"inner"}), AreExpiredAt("after_loop")); +} + +TEST_F(LifetimeAnalysisTest, LoopWithBreak) { + SetupTest(R"( + void target(int count) { + MyObj s1; + MyObj s2; + MyObj* p = &s1; + POINT(before_loop); + for (int i = 0; i < count; ++i) { + if (i == 5) { + p = &s2; + POINT(inside_if); + break; + } + POINT(after_if); + } + POINT(after_loop); + } + )"); + EXPECT_THAT(Origin("p"), HasLoansTo({"s1"}, "before_loop")); + EXPECT_THAT(Origin("p"), HasLoansTo({"s2"}, "inside_if")); + // At the join point after if, s2 cannot make it to p without the if. + EXPECT_THAT(Origin("p"), HasLoansTo({"s1"}, "after_if")); + // At the join point after the loop, p could hold a loan to s1 (if the loop + // completed normally) or to s2 (if the loop was broken). + EXPECT_THAT(Origin("p"), HasLoansTo({"s1", "s2"}, "after_loop")); +} + +TEST_F(LifetimeAnalysisTest, PointersInACycle) { + SetupTest(R"( + void target(bool condition) { + MyObj v1, v2, v3; + MyObj *p1 = &v1, *p2 = &v2, *p3 = &v3; + + POINT(before_while); + while (condition) { + MyObj* temp = p1; + p1 = p2; + p2 = p3; + p3 = temp; + } + POINT(after_loop); + } + )"); + EXPECT_THAT(Origin("p1"), HasLoansTo({"v1"}, "before_while")); + EXPECT_THAT(Origin("p2"), HasLoansTo({"v2"}, "before_while")); + EXPECT_THAT(Origin("p3"), HasLoansTo({"v3"}, "before_while")); + + // At the fixed point after the loop, all pointers could point to any of + // the three variables. + EXPECT_THAT(Origin("p1"), HasLoansTo({"v1", "v2", "v3"}, "after_loop")); + EXPECT_THAT(Origin("p2"), HasLoansTo({"v1", "v2", "v3"}, "after_loop")); + EXPECT_THAT(Origin("p3"), HasLoansTo({"v1", "v2", "v3"}, "after_loop")); + EXPECT_THAT(Origin("temp"), HasLoansTo({"v1", "v2", "v3"}, "after_loop")); +} + +TEST_F(LifetimeAnalysisTest, PointersAndExpirationInACycle) { + SetupTest(R"( + void target(bool condition) { + MyObj v1, v2; + MyObj *p1 = &v1, *p2 = &v2; + + POINT(before_while); + while (condition) { + POINT(in_loop_before_temp); + MyObj temp; + p1 = &temp; + POINT(in_loop_after_temp); + + MyObj* q = p1; + p1 = p2; + p2 = q; + } + POINT(after_loop); + } + )"); + EXPECT_THAT(Origin("p1"), HasLoansTo({"v1"}, "before_while")); + EXPECT_THAT(Origin("p2"), HasLoansTo({"v2"}, "before_while")); + EXPECT_THAT(NoLoans(), AreExpiredAt("before_while")); + + EXPECT_THAT(Origin("p1"), + HasLoansTo({"v1", "v2", "temp"}, "in_loop_before_temp")); + EXPECT_THAT(Origin("p2"), HasLoansTo({"v2", "temp"}, "in_loop_before_temp")); + EXPECT_THAT(LoansTo({"temp"}), AreExpiredAt("in_loop_before_temp")); + + EXPECT_THAT(Origin("p1"), HasLoansTo({"temp"}, "in_loop_after_temp")); + EXPECT_THAT(Origin("p2"), HasLoansTo({"v2", "temp"}, "in_loop_after_temp")); + EXPECT_THAT(NoLoans(), AreExpiredAt("in_loop_after_temp")); + + EXPECT_THAT(Origin("p1"), HasLoansTo({"v1", "v2", "temp"}, "after_loop")); + EXPECT_THAT(Origin("p2"), HasLoansTo({"v2", "temp"}, "after_loop")); + EXPECT_THAT(LoansTo({"temp"}), AreExpiredAt("after_loop")); +} + +TEST_F(LifetimeAnalysisTest, NestedScopes) { + SetupTest(R"( + void target() { + MyObj* p = nullptr; + { + MyObj outer; + p = &outer; + POINT(before_inner_scope); + { + MyObj inner; + p = &inner; + POINT(inside_inner_scope); + } // inner expires + POINT(after_inner_scope); + } // outer expires + } + )"); + EXPECT_THAT(Origin("p"), HasLoansTo({"outer"}, "before_inner_scope")); + EXPECT_THAT(Origin("p"), HasLoansTo({"inner"}, "inside_inner_scope")); + EXPECT_THAT(Origin("p"), HasLoansTo({"inner"}, "after_inner_scope")); +} + +TEST_F(LifetimeAnalysisTest, SimpleExpiry) { + SetupTest(R"( + void target() { + MyObj* p = nullptr; + { + MyObj s; + p = &s; + POINT(before_expiry); + } // s goes out of scope here + POINT(after_expiry); + } + )"); + EXPECT_THAT(NoLoans(), AreExpiredAt("before_expiry")); + EXPECT_THAT(LoansTo({"s"}), AreExpiredAt("after_expiry")); +} + +TEST_F(LifetimeAnalysisTest, NestedExpiry) { + SetupTest(R"( + void target() { + MyObj s1; + MyObj* p = &s1; + POINT(before_inner); + { + MyObj s2; + p = &s2; + POINT(in_inner); + } // s2 expires + POINT(after_inner); + } + )"); + EXPECT_THAT(NoLoans(), AreExpiredAt("before_inner")); + EXPECT_THAT(NoLoans(), AreExpiredAt("in_inner")); + EXPECT_THAT(LoansTo({"s2"}), AreExpiredAt("after_inner")); +} + +TEST_F(LifetimeAnalysisTest, ConditionalExpiry) { + SetupTest(R"( + void target(bool cond) { + MyObj s1; + MyObj* p = &s1; + POINT(before_if); + if (cond) { + MyObj s2; + p = &s2; + POINT(then_block); + } // s2 expires here + POINT(after_if); + } + )"); + EXPECT_THAT(NoLoans(), AreExpiredAt("before_if")); + EXPECT_THAT(NoLoans(), AreExpiredAt("then_block")); + EXPECT_THAT(LoansTo({"s2"}), AreExpiredAt("after_if")); +} + +TEST_F(LifetimeAnalysisTest, LoopExpiry) { + SetupTest(R"( + void target() { + MyObj *p = nullptr; + for (int i = 0; i < 2; ++i) { + POINT(start_loop); + MyObj s; + p = &s; + POINT(end_loop); + } // s expires here on each iteration + POINT(after_loop); + } + )"); + EXPECT_THAT(LoansTo({"s"}), AreExpiredAt("start_loop")); + EXPECT_THAT(NoLoans(), AreExpiredAt("end_loop")); + EXPECT_THAT(LoansTo({"s"}), AreExpiredAt("after_loop")); +} + +TEST_F(LifetimeAnalysisTest, MultipleExpiredLoans) { + SetupTest(R"( + void target() { + MyObj *p1, *p2, *p3; + { + MyObj s1; + p1 = &s1; + POINT(p1); + } // s1 expires + POINT(p2); + { + MyObj s2; + p2 = &s2; + MyObj s3; + p3 = &s3; + POINT(p3); + } // s2, s3 expire + POINT(p4); + } + )"); + EXPECT_THAT(NoLoans(), AreExpiredAt("p1")); + EXPECT_THAT(LoansTo({"s1"}), AreExpiredAt("p2")); + EXPECT_THAT(LoansTo({"s1"}), AreExpiredAt("p3")); + EXPECT_THAT(LoansTo({"s1", "s2", "s3"}), AreExpiredAt("p4")); +} + +TEST_F(LifetimeAnalysisTest, GotoJumpsOutOfScope) { + SetupTest(R"( + void target(bool cond) { + MyObj *p = nullptr; + { + MyObj s; + p = &s; + POINT(before_goto); + if (cond) { + goto end; + } + } // `s` expires here on the path that doesn't jump + POINT(after_scope); + end: + POINT(after_goto); + } + )"); + EXPECT_THAT(NoLoans(), AreExpiredAt("before_goto")); + EXPECT_THAT(LoansTo({"s"}), AreExpiredAt("after_scope")); + EXPECT_THAT(LoansTo({"s"}), AreExpiredAt("after_goto")); +} + +TEST_F(LifetimeAnalysisTest, ContinueInLoop) { + SetupTest(R"( + void target(int count) { + MyObj *p = nullptr; + MyObj outer; + p = &outer; + POINT(before_loop); + + for (int i = 0; i < count; ++i) { + if (i % 2 == 0) { + MyObj s_even; + p = &s_even; + POINT(in_even_iter); + continue; + } + MyObj s_odd; + p = &s_odd; + POINT(in_odd_iter); + } + POINT(after_loop); + } + )"); + EXPECT_THAT(NoLoans(), AreExpiredAt("before_loop")); + EXPECT_THAT(LoansTo({"s_odd"}), AreExpiredAt("in_even_iter")); + EXPECT_THAT(LoansTo({"s_even"}), AreExpiredAt("in_odd_iter")); + EXPECT_THAT(LoansTo({"s_even", "s_odd"}), AreExpiredAt("after_loop")); +} + +TEST_F(LifetimeAnalysisTest, ReassignedPointerThenOriginalExpires) { + SetupTest(R"( + void target() { + MyObj* p = nullptr; + { + MyObj s1; + p = &s1; + POINT(p_has_s1); + { + MyObj s2; + p = &s2; + POINT(p_has_s2); + } + POINT(p_after_s2_expires); + } // s1 expires here. + POINT(p_after_s1_expires); + } + )"); + EXPECT_THAT(NoLoans(), AreExpiredAt("p_has_s1")); + EXPECT_THAT(NoLoans(), AreExpiredAt("p_has_s2")); + EXPECT_THAT(LoansTo({"s2"}), AreExpiredAt("p_after_s2_expires")); + EXPECT_THAT(LoansTo({"s1", "s2"}), AreExpiredAt("p_after_s1_expires")); +} + +} // anonymous namespace +} // namespace clang::lifetimes::internal diff --git a/clang/unittests/Format/BracesInserterTest.cpp b/clang/unittests/Format/BracesInserterTest.cpp index e0c447d..572e53e 100644 --- a/clang/unittests/Format/BracesInserterTest.cpp +++ b/clang/unittests/Format/BracesInserterTest.cpp @@ -257,9 +257,9 @@ TEST_F(BracesInserterTest, InsertBracesRange) { FormatStyle Style = getLLVMStyle(); Style.InsertBraces = true; - const StringRef Code("while (a)\n" - " if (b)\n" - " return;"); + constexpr StringRef Code("while (a)\n" + " if (b)\n" + " return;"); verifyFormat("while (a) {\n" " if (b)\n" diff --git a/clang/unittests/Format/FormatTest.cpp b/clang/unittests/Format/FormatTest.cpp index 7e3d2a9..dbf6950 100644 --- a/clang/unittests/Format/FormatTest.cpp +++ b/clang/unittests/Format/FormatTest.cpp @@ -4803,12 +4803,13 @@ TEST_F(FormatTest, FormatsInlineASM) { "int i;"); auto Style = getLLVMStyleWithColumns(0); - const StringRef Code1{"asm(\"xyz\" : \"=a\"(a), \"=d\"(b) : \"a\"(data));"}; - const StringRef Code2{"asm(\"xyz\"\n" - " : \"=a\"(a), \"=d\"(b)\n" - " : \"a\"(data));"}; - const StringRef Code3{"asm(\"xyz\" : \"=a\"(a), \"=d\"(b)\n" - " : \"a\"(data));"}; + constexpr StringRef Code1( + "asm(\"xyz\" : \"=a\"(a), \"=d\"(b) : \"a\"(data));"); + constexpr StringRef Code2("asm(\"xyz\"\n" + " : \"=a\"(a), \"=d\"(b)\n" + " : \"a\"(data));"); + constexpr StringRef Code3("asm(\"xyz\" : \"=a\"(a), \"=d\"(b)\n" + " : \"a\"(data));"); Style.BreakBeforeInlineASMColon = FormatStyle::BBIAS_OnlyMultiline; verifyFormat(Code1, Style); @@ -6704,16 +6705,16 @@ TEST_F(FormatTest, EscapedNewlines) { " int x(int a);", AlignLeft); - constexpr StringRef Code{"#define A \\\n" + constexpr StringRef Code("#define A \\\n" " int a123; \\\n" " int a; \\\n" - " int a1234;"}; + " int a1234;"); verifyFormat(Code, AlignLeft); - constexpr StringRef Code2{"#define A \\\n" + constexpr StringRef Code2("#define A \\\n" " int a123; \\\n" " int a; \\\n" - " int a1234;"}; + " int a1234;"); auto LastLine = getLLVMStyle(); LastLine.AlignEscapedNewlines = FormatStyle::ENAS_LeftWithLastLine; verifyFormat(Code2, LastLine); @@ -12108,9 +12109,9 @@ TEST_F(FormatTest, PointerAlignmentFallback) { FormatStyle Style = getLLVMStyle(); Style.DerivePointerAlignment = true; - const StringRef Code("int* p;\n" - "int *q;\n" - "int * r;"); + constexpr StringRef Code("int* p;\n" + "int *q;\n" + "int * r;"); EXPECT_EQ(Style.PointerAlignment, FormatStyle::PAS_Right); verifyFormat("int *p;\n" @@ -15025,7 +15026,7 @@ TEST_F(FormatTest, PullTrivialFunctionDefinitionsIntoSingleLine) { " aaaaaaaaaaaaaaaaaa,\n" " aaaaaaaaaaaaaaaaaabbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb) {}"); - constexpr StringRef Code{"void foo() { /* Empty */ }"}; + constexpr StringRef Code("void foo() { /* Empty */ }"); verifyFormat(Code); verifyFormat(Code, "void foo() { /* Empty */\n" "}"); @@ -27255,7 +27256,7 @@ TEST_F(FormatTest, IndentAccessModifiers) { TEST_F(FormatTest, LimitlessStringsAndComments) { auto Style = getLLVMStyleWithColumns(0); - constexpr StringRef Code = + constexpr StringRef Code( "/**\n" " * This is a multiline comment with quite some long lines, at least for " "the LLVM Style.\n" @@ -27276,7 +27277,7 @@ TEST_F(FormatTest, LimitlessStringsAndComments) { " const std::string SmallString = \"Hello World\";\n" " // Small line comment\n" " return String.size() > SmallString.size();\n" - "}"; + "}"); verifyNoChange(Code, Style); } @@ -28400,9 +28401,9 @@ TEST_F(FormatTest, InsertNewlineAtEOF) { verifyNoChange("int i;\n", Style); verifyFormat("int i;\n", "int i;", Style); - constexpr StringRef Code{"namespace {\n" + constexpr StringRef Code("namespace {\n" "int i;\n" - "} // namespace"}; + "} // namespace"); verifyFormat(Code.str() + '\n', Code, Style, {tooling::Range(19, 13)}); // line 3 } @@ -28411,7 +28412,7 @@ TEST_F(FormatTest, KeepEmptyLinesAtEOF) { FormatStyle Style = getLLVMStyle(); Style.KeepEmptyLines.AtEndOfFile = true; - const StringRef Code{"int i;\n\n"}; + constexpr StringRef Code("int i;\n\n"); verifyNoChange(Code, Style); verifyFormat(Code, "int i;\n\n\n", Style); } @@ -28644,8 +28645,8 @@ TEST_F(FormatTest, PPDirectivesAndCommentsInBracedInit) { } TEST_F(FormatTest, BreakAdjacentStringLiterals) { - constexpr StringRef Code{ - "return \"Code\" \"\\0\\52\\26\\55\\55\\0\" \"x013\" \"\\02\\xBA\";"}; + constexpr StringRef Code( + "return \"Code\" \"\\0\\52\\26\\55\\55\\0\" \"x013\" \"\\02\\xBA\";"); verifyFormat("return \"Code\"\n" " \"\\0\\52\\26\\55\\55\\0\"\n" @@ -29056,9 +29057,9 @@ TEST_F(FormatTest, KeepFormFeed) { auto Style = getLLVMStyle(); Style.KeepFormFeed = true; - constexpr StringRef NoFormFeed{"int i;\n" + constexpr StringRef NoFormFeed("int i;\n" "\n" - "void f();"}; + "void f();"); verifyFormat(NoFormFeed, "int i;\n" " \f\n" @@ -29080,9 +29081,9 @@ TEST_F(FormatTest, KeepFormFeed) { "void f();\f", Style); - constexpr StringRef FormFeed{"int i;\n" + constexpr StringRef FormFeed("int i;\n" "\f\n" - "void f();"}; + "void f();"); verifyNoChange(FormFeed, Style); Style.LineEnding = FormatStyle::LE_LF; @@ -29092,10 +29093,10 @@ TEST_F(FormatTest, KeepFormFeed) { "void f();", Style); - constexpr StringRef FormFeedBeforeEmptyLine{"int i;\n" + constexpr StringRef FormFeedBeforeEmptyLine("int i;\n" "\f\n" "\n" - "void f();"}; + "void f();"); Style.MaxEmptyLinesToKeep = 2; verifyFormat(FormFeedBeforeEmptyLine, "int i;\n" diff --git a/clang/unittests/Format/FormatTestComments.cpp b/clang/unittests/Format/FormatTestComments.cpp index 8870755..69026bc 100644 --- a/clang/unittests/Format/FormatTestComments.cpp +++ b/clang/unittests/Format/FormatTestComments.cpp @@ -1120,11 +1120,11 @@ TEST_F(FormatTestComments, KeepsLevelOfCommentBeforePPDirective) { " }\n" "}")); - const StringRef Code("void func() {\n" - " // clang-format off\n" - " #define KV(value) #value, value\n" - " // clang-format on\n" - "}"); + constexpr StringRef Code("void func() {\n" + " // clang-format off\n" + " #define KV(value) #value, value\n" + " // clang-format on\n" + "}"); verifyNoChange(Code); auto Style = getLLVMStyle(); diff --git a/clang/unittests/Format/FormatTestJava.cpp b/clang/unittests/Format/FormatTestJava.cpp index ca5aba0..1275564 100644 --- a/clang/unittests/Format/FormatTestJava.cpp +++ b/clang/unittests/Format/FormatTestJava.cpp @@ -631,17 +631,17 @@ TEST_F(FormatTestJava, SwitchExpression) { "});", Style); - constexpr StringRef Code1{"i = switch (day) {\n" + constexpr StringRef Code1("i = switch (day) {\n" " case THURSDAY, SATURDAY -> 8;\n" " case WEDNESDAY -> 9;\n" " default -> 0;\n" - "};"}; + "};"); verifyFormat(Code1, Style); Style.IndentCaseLabels = true; verifyFormat(Code1, Style); - constexpr StringRef Code2{"i = switch (day) {\n" + constexpr StringRef Code2("i = switch (day) {\n" " case THURSDAY, SATURDAY -> {\n" " foo();\n" " yield 8;\n" @@ -653,17 +653,17 @@ TEST_F(FormatTestJava, SwitchExpression) { " default -> {\n" " yield 0;\n" " }\n" - "};"}; + "};"); verifyFormat(Code2, Style); Style.IndentCaseLabels = false; verifyFormat(Code2, Style); - constexpr StringRef Code3{"switch (day) {\n" + constexpr StringRef Code3("switch (day) {\n" "case THURSDAY, SATURDAY -> i = 8;\n" "case WEDNESDAY -> i = 9;\n" "default -> i = 0;\n" - "};"}; + "};"); verifyFormat(Code3, Style); Style.IndentCaseLabels = true; diff --git a/clang/unittests/Format/FormatTestSelective.cpp b/clang/unittests/Format/FormatTestSelective.cpp index fa7630a..1a01153 100644 --- a/clang/unittests/Format/FormatTestSelective.cpp +++ b/clang/unittests/Format/FormatTestSelective.cpp @@ -672,15 +672,14 @@ TEST_F(FormatTestSelective, FormatMacroRegardlessOfPreviousIndent) { // need to be adapted. Style = getLLVMStyle(); - const StringRef Code{" class Foo {\n" - " void test() {\n" - " #ifdef 1\n" - " #define some\n" // format this line - " #endif\n" - " }};"}; - - EXPECT_EQ(Style.IndentPPDirectives, - FormatStyle::PPDirectiveIndentStyle::PPDIS_None); + constexpr StringRef Code(" class Foo {\n" + " void test() {\n" + " #ifdef 1\n" + " #define some\n" // format this line + " #endif\n" + " }};"); + + EXPECT_EQ(Style.IndentPPDirectives, FormatStyle::PPDIS_None); EXPECT_EQ(" class Foo {\n" " void test() {\n" " #ifdef 1\n" diff --git a/clang/unittests/Format/IntegerLiteralSeparatorTest.cpp b/clang/unittests/Format/IntegerLiteralSeparatorTest.cpp index b1e42e9..8681c3d 100644 --- a/clang/unittests/Format/IntegerLiteralSeparatorTest.cpp +++ b/clang/unittests/Format/IntegerLiteralSeparatorTest.cpp @@ -24,7 +24,7 @@ TEST_F(IntegerLiteralSeparatorTest, SingleQuoteAsSeparator) { EXPECT_EQ(Style.IntegerLiteralSeparator.Decimal, 0); EXPECT_EQ(Style.IntegerLiteralSeparator.Hex, 0); - const StringRef Binary("b = 0b10011'11'0110'1u;"); + constexpr StringRef Binary("b = 0b10011'11'0110'1u;"); verifyFormat(Binary, Style); Style.IntegerLiteralSeparator.Binary = -1; verifyFormat("b = 0b100111101101u;", Binary, Style); @@ -33,14 +33,14 @@ TEST_F(IntegerLiteralSeparatorTest, SingleQuoteAsSeparator) { Style.IntegerLiteralSeparator.Binary = 4; verifyFormat("b = 0b1001'1110'1101u;", Binary, Style); - const StringRef Decimal("d = 184467'440737'0'95505'92Ull;"); + constexpr StringRef Decimal("d = 184467'440737'0'95505'92Ull;"); verifyFormat(Decimal, Style); Style.IntegerLiteralSeparator.Decimal = -1; verifyFormat("d = 18446744073709550592Ull;", Decimal, Style); Style.IntegerLiteralSeparator.Decimal = 3; verifyFormat("d = 18'446'744'073'709'550'592Ull;", Decimal, Style); - const StringRef Hex("h = 0xDEAD'BEEF'DE'AD'BEE'Fuz;"); + constexpr StringRef Hex("h = 0xDEAD'BEEF'DE'AD'BEE'Fuz;"); verifyFormat(Hex, Style); Style.IntegerLiteralSeparator.Hex = -1; verifyFormat("h = 0xDEADBEEFDEADBEEFuz;", Hex, Style); @@ -87,9 +87,9 @@ TEST_F(IntegerLiteralSeparatorTest, SingleQuoteAsSeparator) { TEST_F(IntegerLiteralSeparatorTest, UnderscoreAsSeparator) { FormatStyle Style = getLLVMStyle(); - const StringRef Binary("B = 0B10011_11_0110_1;"); - const StringRef Decimal("d = 184467_440737_0_95505_92;"); - const StringRef Hex("H = 0XDEAD_BEEF_DE_AD_BEE_F;"); + constexpr StringRef Binary("B = 0B10011_11_0110_1;"); + constexpr StringRef Decimal("d = 184467_440737_0_95505_92;"); + constexpr StringRef Hex("H = 0XDEAD_BEEF_DE_AD_BEE_F;"); auto TestUnderscore = [&](auto Language) { Style.Language = Language; @@ -173,16 +173,16 @@ TEST_F(IntegerLiteralSeparatorTest, FixRanges) { FormatStyle Style = getLLVMStyle(); Style.IntegerLiteralSeparator.Decimal = 3; - const StringRef Code("i = -12'34;\n" - "// clang-format off\n" - "j = 123'4;\n" - "// clang-format on\n" - "k = +1'23'4;"); - const StringRef Expected("i = -1'234;\n" + constexpr StringRef Code("i = -12'34;\n" "// clang-format off\n" "j = 123'4;\n" "// clang-format on\n" - "k = +1'234;"); + "k = +1'23'4;"); + constexpr StringRef Expected("i = -1'234;\n" + "// clang-format off\n" + "j = 123'4;\n" + "// clang-format on\n" + "k = +1'234;"); verifyFormat(Expected, Code, Style); diff --git a/clang/unittests/Format/SortIncludesTest.cpp b/clang/unittests/Format/SortIncludesTest.cpp index 5194d65..48ecd5d 100644 --- a/clang/unittests/Format/SortIncludesTest.cpp +++ b/clang/unittests/Format/SortIncludesTest.cpp @@ -1084,10 +1084,10 @@ TEST_F(SortIncludesTest, DoNotSortLikelyXml) { } TEST_F(SortIncludesTest, DoNotSortCSharp) { - constexpr StringRef Code{"const string expectedDataStruct = @\"\n" + constexpr StringRef Code("const string expectedDataStruct = @\"\n" " #include <b.h>\n" " #include <a.h>\n" - " \";"}; + " \";"); FmtStyle.Language = FormatStyle::LK_CSharp; EXPECT_TRUE(sortIncludes(FmtStyle, Code, GetCodeRange(Code), "a.cs").empty()); } diff --git a/clang/unittests/Format/TokenAnnotatorTest.cpp b/clang/unittests/Format/TokenAnnotatorTest.cpp index ce7787e..7f99655 100644 --- a/clang/unittests/Format/TokenAnnotatorTest.cpp +++ b/clang/unittests/Format/TokenAnnotatorTest.cpp @@ -618,7 +618,7 @@ TEST_F(TokenAnnotatorTest, UnderstandsStructs) { EXPECT_TOKEN(Tokens[19], tok::l_brace, TT_StructLBrace); EXPECT_TOKEN(Tokens[20], tok::r_brace, TT_StructRBrace); - constexpr StringRef Code{"struct EXPORT StructName {};"}; + constexpr StringRef Code("struct EXPORT StructName {};"); Tokens = annotate(Code); ASSERT_EQ(Tokens.size(), 7u) << Tokens; @@ -3958,7 +3958,7 @@ TEST_F(TokenAnnotatorTest, SplitPenalty) { } TEST_F(TokenAnnotatorTest, TemplateName) { - constexpr StringRef Code{"return Foo < A || B > (C ^ D);"}; + constexpr StringRef Code("return Foo < A || B > (C ^ D);"); auto Tokens = annotate(Code); ASSERT_EQ(Tokens.size(), 14u) << Tokens; diff --git a/clang/unittests/Interpreter/InterpreterTest.cpp b/clang/unittests/Interpreter/InterpreterTest.cpp index 2ba15cb..768058b 100644 --- a/clang/unittests/Interpreter/InterpreterTest.cpp +++ b/clang/unittests/Interpreter/InterpreterTest.cpp @@ -160,12 +160,12 @@ TEST_F(InterpreterTest, UndoCommand) { // Fail to undo. auto Err1 = Interp->Undo(); - EXPECT_EQ("Operation failed. Too many undos", + EXPECT_EQ("Operation failed. No input left to undo", llvm::toString(std::move(Err1))); auto Err2 = Interp->Parse("int foo = 42;"); EXPECT_TRUE(!!Err2); auto Err3 = Interp->Undo(2); - EXPECT_EQ("Operation failed. Too many undos", + EXPECT_EQ("Operation failed. Wanted to undo 2 inputs, only have 1.", llvm::toString(std::move(Err3))); // Succeed to undo. diff --git a/clang/unittests/StaticAnalyzer/BlockEntranceCallbackTest.cpp b/clang/unittests/StaticAnalyzer/BlockEntranceCallbackTest.cpp index 0f05c39..d15bec0 100644 --- a/clang/unittests/StaticAnalyzer/BlockEntranceCallbackTest.cpp +++ b/clang/unittests/StaticAnalyzer/BlockEntranceCallbackTest.cpp @@ -91,8 +91,7 @@ void addBlockEntranceTester(AnalysisASTConsumer &AnalysisConsumer, AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { Registry.addChecker(®isterChecker<BlockEntranceCallbackTester>, &shouldAlwaysRegister, "test.BlockEntranceTester", - "EmptyDescription", "EmptyDocsUri", - /*IsHidden=*/false); + "EmptyDescription"); }); } @@ -102,8 +101,7 @@ void addBranchConditionTester(AnalysisASTConsumer &AnalysisConsumer, AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { Registry.addChecker(®isterChecker<BranchConditionCallbackTester>, &shouldAlwaysRegister, "test.BranchConditionTester", - "EmptyDescription", "EmptyDocsUri", - /*IsHidden=*/false); + "EmptyDescription"); }); } diff --git a/clang/unittests/StaticAnalyzer/BugReportInterestingnessTest.cpp b/clang/unittests/StaticAnalyzer/BugReportInterestingnessTest.cpp index 0ef63b0..fc50f00 100644 --- a/clang/unittests/StaticAnalyzer/BugReportInterestingnessTest.cpp +++ b/clang/unittests/StaticAnalyzer/BugReportInterestingnessTest.cpp @@ -120,7 +120,7 @@ public: std::move(ExpectedDiags), Compiler.getSourceManager())); AnalysisConsumer->AddCheckerRegistrationFn([](CheckerRegistry &Registry) { Registry.addChecker<InterestingnessTestChecker>("test.Interestingness", - "Description", ""); + "MockDescription"); }); Compiler.getAnalyzerOpts().CheckersAndPackages = { {"test.Interestingness", true}}; diff --git a/clang/unittests/StaticAnalyzer/CallDescriptionTest.cpp b/clang/unittests/StaticAnalyzer/CallDescriptionTest.cpp index 4cb6bd3..e2007a9 100644 --- a/clang/unittests/StaticAnalyzer/CallDescriptionTest.cpp +++ b/clang/unittests/StaticAnalyzer/CallDescriptionTest.cpp @@ -616,8 +616,8 @@ void addCallDescChecker(AnalysisASTConsumer &AnalysisConsumer, AnalyzerOptions &AnOpts) { AnOpts.CheckersAndPackages = {{"test.CallDescChecker", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { - Registry.addChecker<CallDescChecker>("test.CallDescChecker", "Description", - ""); + Registry.addChecker<CallDescChecker>("test.CallDescChecker", + "MockDescription"); }); } diff --git a/clang/unittests/StaticAnalyzer/CallEventTest.cpp b/clang/unittests/StaticAnalyzer/CallEventTest.cpp index 2843572e..8b5289e 100644 --- a/clang/unittests/StaticAnalyzer/CallEventTest.cpp +++ b/clang/unittests/StaticAnalyzer/CallEventTest.cpp @@ -56,7 +56,7 @@ void addCXXDeallocatorChecker(AnalysisASTConsumer &AnalysisConsumer, AnOpts.CheckersAndPackages = {{"test.CXXDeallocator", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { Registry.addChecker<CXXDeallocatorChecker>("test.CXXDeallocator", - "Description", ""); + "MockDescription"); }); } diff --git a/clang/unittests/StaticAnalyzer/ConflictingEvalCallsTest.cpp b/clang/unittests/StaticAnalyzer/ConflictingEvalCallsTest.cpp index e410cca..cffdbf1 100644 --- a/clang/unittests/StaticAnalyzer/ConflictingEvalCallsTest.cpp +++ b/clang/unittests/StaticAnalyzer/ConflictingEvalCallsTest.cpp @@ -33,10 +33,8 @@ void addEvalFooCheckers(AnalysisASTConsumer &AnalysisConsumer, AnOpts.CheckersAndPackages = {{"test.EvalFoo1", true}, {"test.EvalFoo2", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { - Registry.addChecker<EvalCallFoo1>("test.EvalFoo1", "EmptyDescription", - "EmptyDocsUri"); - Registry.addChecker<EvalCallFoo2>("test.EvalFoo2", "EmptyDescription", - "EmptyDocsUri"); + Registry.addChecker<EvalCallFoo1>("test.EvalFoo1", "MockDescription"); + Registry.addChecker<EvalCallFoo2>("test.EvalFoo2", "MockDescription"); }); } } // namespace diff --git a/clang/unittests/StaticAnalyzer/ExprEngineVisitTest.cpp b/clang/unittests/StaticAnalyzer/ExprEngineVisitTest.cpp index b6eeb9c..12be228 100644 --- a/clang/unittests/StaticAnalyzer/ExprEngineVisitTest.cpp +++ b/clang/unittests/StaticAnalyzer/ExprEngineVisitTest.cpp @@ -78,7 +78,7 @@ void addExprEngineVisitPreChecker(AnalysisASTConsumer &AnalysisConsumer, AnOpts.CheckersAndPackages = {{"ExprEngineVisitPreChecker", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { Registry.addChecker<ExprEngineVisitPreChecker>("ExprEngineVisitPreChecker", - "Desc", "DocsURI"); + "MockDescription"); }); } @@ -87,7 +87,7 @@ void addExprEngineVisitPostChecker(AnalysisASTConsumer &AnalysisConsumer, AnOpts.CheckersAndPackages = {{"ExprEngineVisitPostChecker", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { Registry.addChecker<ExprEngineVisitPostChecker>( - "ExprEngineVisitPostChecker", "Desc", "DocsURI"); + "ExprEngineVisitPostChecker", "MockDescription"); }); } @@ -95,8 +95,8 @@ void addMemAccessChecker(AnalysisASTConsumer &AnalysisConsumer, AnalyzerOptions &AnOpts) { AnOpts.CheckersAndPackages = {{"MemAccessChecker", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { - Registry.addChecker<MemAccessChecker>("MemAccessChecker", "Desc", - "DocsURI"); + Registry.addChecker<MemAccessChecker>("MemAccessChecker", + "MockDescription"); }); } diff --git a/clang/unittests/StaticAnalyzer/FalsePositiveRefutationBRVisitorTest.cpp b/clang/unittests/StaticAnalyzer/FalsePositiveRefutationBRVisitorTest.cpp index 8f0a96d..146797f 100644 --- a/clang/unittests/StaticAnalyzer/FalsePositiveRefutationBRVisitorTest.cpp +++ b/clang/unittests/StaticAnalyzer/FalsePositiveRefutationBRVisitorTest.cpp @@ -92,8 +92,8 @@ void addFalsePositiveGenerator(AnalysisASTConsumer &AnalysisConsumer, AnOpts.CheckersAndPackages = {{"test.FalsePositiveGenerator", true}, {"debug.ViewExplodedGraph", false}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { - Registry.addChecker<FalsePositiveGenerator>( - "test.FalsePositiveGenerator", "EmptyDescription", "EmptyDocsUri"); + Registry.addChecker<FalsePositiveGenerator>("test.FalsePositiveGenerator", + "MockDescription"); }); } diff --git a/clang/unittests/StaticAnalyzer/MemRegionDescriptiveNameTest.cpp b/clang/unittests/StaticAnalyzer/MemRegionDescriptiveNameTest.cpp index 0f6e49b..7b837f3 100644 --- a/clang/unittests/StaticAnalyzer/MemRegionDescriptiveNameTest.cpp +++ b/clang/unittests/StaticAnalyzer/MemRegionDescriptiveNameTest.cpp @@ -46,7 +46,7 @@ void addDescriptiveNameChecker(AnalysisASTConsumer &AnalysisConsumer, AnOpts.CheckersAndPackages = {{"DescriptiveNameChecker", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { Registry.addChecker<DescriptiveNameChecker>("DescriptiveNameChecker", - "Desc", "DocsURI"); + "MockDescription"); }); } diff --git a/clang/unittests/StaticAnalyzer/NoStateChangeFuncVisitorTest.cpp b/clang/unittests/StaticAnalyzer/NoStateChangeFuncVisitorTest.cpp index a903342..68d2678 100644 --- a/clang/unittests/StaticAnalyzer/NoStateChangeFuncVisitorTest.cpp +++ b/clang/unittests/StaticAnalyzer/NoStateChangeFuncVisitorTest.cpp @@ -140,7 +140,7 @@ void addNonThoroughStatefulChecker(AnalysisASTConsumer &AnalysisConsumer, AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { Registry .addChecker<StatefulChecker<NonThoroughErrorNotPreventedFuncVisitor>>( - "test.StatefulChecker", "Description", ""); + "test.StatefulChecker", "MockDescription"); }); } @@ -233,7 +233,7 @@ void addThoroughStatefulChecker(AnalysisASTConsumer &AnalysisConsumer, AnOpts.CheckersAndPackages = {{"test.StatefulChecker", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { Registry.addChecker<StatefulChecker<ThoroughErrorNotPreventedFuncVisitor>>( - "test.StatefulChecker", "Description", ""); + "test.StatefulChecker", "MockDescription"); }); } diff --git a/clang/unittests/StaticAnalyzer/ObjcBug-124477.cpp b/clang/unittests/StaticAnalyzer/ObjcBug-124477.cpp index 51bd332..ab78090 100644 --- a/clang/unittests/StaticAnalyzer/ObjcBug-124477.cpp +++ b/clang/unittests/StaticAnalyzer/ObjcBug-124477.cpp @@ -37,7 +37,7 @@ void addFlagFlipperChecker(AnalysisASTConsumer &AnalysisConsumer, AnOpts.CheckersAndPackages = {{"test.FlipFlagOnCheckLocation", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { Registry.addChecker<FlipFlagOnCheckLocation>("test.FlipFlagOnCheckLocation", - "Description", ""); + "MockDescription"); }); } diff --git a/clang/unittests/StaticAnalyzer/RegisterCustomCheckersTest.cpp b/clang/unittests/StaticAnalyzer/RegisterCustomCheckersTest.cpp index 454eee9..e17d107 100644 --- a/clang/unittests/StaticAnalyzer/RegisterCustomCheckersTest.cpp +++ b/clang/unittests/StaticAnalyzer/RegisterCustomCheckersTest.cpp @@ -44,7 +44,7 @@ void addCustomChecker(AnalysisASTConsumer &AnalysisConsumer, AnalyzerOptions &AnOpts) { AnOpts.CheckersAndPackages = {{"test.CustomChecker", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { - Registry.addChecker<CustomChecker>("test.CustomChecker", "Description", ""); + Registry.addChecker<CustomChecker>("test.CustomChecker", "MockDescription"); }); } @@ -73,8 +73,8 @@ void addLocIncDecChecker(AnalysisASTConsumer &AnalysisConsumer, AnalyzerOptions &AnOpts) { AnOpts.CheckersAndPackages = {{"test.LocIncDecChecker", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { - Registry.addChecker<CustomChecker>("test.LocIncDecChecker", "Description", - ""); + Registry.addChecker<CustomChecker>("test.LocIncDecChecker", + "MockDescription"); }); } @@ -119,10 +119,10 @@ bool shouldRegisterCheckerRegistrationOrderPrinter(const CheckerManager &mgr) { void addCheckerRegistrationOrderPrinter(CheckerRegistry &Registry) { Registry.addChecker(registerCheckerRegistrationOrderPrinter, shouldRegisterCheckerRegistrationOrderPrinter, - "test.RegistrationOrder", "Description", "", false); + "test.RegistrationOrder", "Description"); } -#define UNITTEST_CHECKER(CHECKER_NAME, DIAG_MSG) \ +#define UNITTEST_CHECKER(CHECKER_NAME) \ class CHECKER_NAME : public Checker<check::PreStmt<DeclStmt>> { \ public: \ void checkPreStmt(const DeclStmt *DS, CheckerContext &C) const {} \ @@ -137,11 +137,11 @@ void addCheckerRegistrationOrderPrinter(CheckerRegistry &Registry) { } \ void add##CHECKER_NAME(CheckerRegistry &Registry) { \ Registry.addChecker(register##CHECKER_NAME, shouldRegister##CHECKER_NAME, \ - "test." #CHECKER_NAME, "Description", "", false); \ + "test." #CHECKER_NAME, "Description"); \ } -UNITTEST_CHECKER(StrongDep, "Strong") -UNITTEST_CHECKER(Dep, "Dep") +UNITTEST_CHECKER(StrongDep) +UNITTEST_CHECKER(Dep) bool shouldRegisterStrongFALSE(const CheckerManager &mgr) { return false; @@ -154,7 +154,7 @@ void addDep(AnalysisASTConsumer &AnalysisConsumer, {"test.RegistrationOrder", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { Registry.addChecker(registerStrongDep, shouldRegisterStrongFALSE, - "test.Strong", "Description", "", false); + "test.Strong", "Description"); addStrongDep(Registry); addDep(Registry); addCheckerRegistrationOrderPrinter(Registry); @@ -172,7 +172,7 @@ TEST(RegisterDeps, UnsatisfiedDependency) { // Weak checker dependencies. //===----------------------------------------------------------------------===// -UNITTEST_CHECKER(WeakDep, "Weak") +UNITTEST_CHECKER(WeakDep) void addWeakDepCheckerBothEnabled(AnalysisASTConsumer &AnalysisConsumer, AnalyzerOptions &AnOpts) { @@ -225,8 +225,8 @@ void addWeakDepCheckerDepUnspecified(AnalysisASTConsumer &AnalysisConsumer, }); } -UNITTEST_CHECKER(WeakDep2, "Weak2") -UNITTEST_CHECKER(Dep2, "Dep2") +UNITTEST_CHECKER(WeakDep2) +UNITTEST_CHECKER(Dep2) void addWeakDepHasWeakDep(AnalysisASTConsumer &AnalysisConsumer, AnalyzerOptions &AnOpts) { diff --git a/clang/unittests/StaticAnalyzer/SValSimplifyerTest.cpp b/clang/unittests/StaticAnalyzer/SValSimplifyerTest.cpp index 85cfe2c..4331ffc 100644 --- a/clang/unittests/StaticAnalyzer/SValSimplifyerTest.cpp +++ b/clang/unittests/StaticAnalyzer/SValSimplifyerTest.cpp @@ -68,8 +68,7 @@ static void addSimplifyChecker(AnalysisASTConsumer &AnalysisConsumer, AnalyzerOptions &AnOpts) { AnOpts.CheckersAndPackages = {{"SimplifyChecker", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { - Registry.addChecker<SimplifyChecker>("SimplifyChecker", "EmptyDescription", - "EmptyDocsUri"); + Registry.addChecker<SimplifyChecker>("SimplifyChecker", "MockDescription"); }); } diff --git a/clang/unittests/StaticAnalyzer/SValTest.cpp b/clang/unittests/StaticAnalyzer/SValTest.cpp index d8897b0..58e9a8d 100644 --- a/clang/unittests/StaticAnalyzer/SValTest.cpp +++ b/clang/unittests/StaticAnalyzer/SValTest.cpp @@ -139,10 +139,10 @@ class SValTest : public testing::TestWithParam<TestClangConfig> {}; \ void add##NAME##SValCollector(AnalysisASTConsumer &AnalysisConsumer, \ AnalyzerOptions &AnOpts) { \ - AnOpts.CheckersAndPackages = {{"test.##NAME##SValCollector", true}}; \ + AnOpts.CheckersAndPackages = {{"test." #NAME "SValColl", true}}; \ AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { \ - Registry.addChecker<NAME##SValCollector>("test.##NAME##SValCollector", \ - "Description", ""); \ + Registry.addChecker<NAME##SValCollector>("test." #NAME "SValColl", \ + "MockDescription"); \ }); \ } \ \ diff --git a/clang/unittests/StaticAnalyzer/TestReturnValueUnderConstruction.cpp b/clang/unittests/StaticAnalyzer/TestReturnValueUnderConstruction.cpp index 5fc084a..0cb3c59 100644 --- a/clang/unittests/StaticAnalyzer/TestReturnValueUnderConstruction.cpp +++ b/clang/unittests/StaticAnalyzer/TestReturnValueUnderConstruction.cpp @@ -49,9 +49,9 @@ void addTestReturnValueUnderConstructionChecker( AnOpts.CheckersAndPackages = {{"test.TestReturnValueUnderConstruction", true}}; AnalysisConsumer.AddCheckerRegistrationFn([](CheckerRegistry &Registry) { - Registry.addChecker<TestReturnValueUnderConstructionChecker>( - "test.TestReturnValueUnderConstruction", "", ""); - }); + Registry.addChecker<TestReturnValueUnderConstructionChecker>( + "test.TestReturnValueUnderConstruction", "MockDescription"); + }); } TEST(TestReturnValueUnderConstructionChecker, |