diff options
Diffstat (limited to 'clang')
99 files changed, 10922 insertions, 741 deletions
diff --git a/clang/docs/analyzer/developer-docs/DebugChecks.rst b/clang/docs/analyzer/developer-docs/DebugChecks.rst index 767ef65..b3b9089 100644 --- a/clang/docs/analyzer/developer-docs/DebugChecks.rst +++ b/clang/docs/analyzer/developer-docs/DebugChecks.rst @@ -9,6 +9,22 @@ The analyzer contains a number of checkers which can aid in debugging. Enable them by using the "-analyzer-checker=" flag, followed by the name of the checker. +These checkers are especially useful when analyzing a specific function, using +the `-analyze-function` flag. The flag accepts the function name for C code, +like `-analyze-function=myfunction`. +For C++ code, due to overloading, the function name must include the +parameter list, like `-analyze-function="myfunction(int, _Bool)"`. + +Note that `bool` must be spelled as `_Bool` in the parameter list. +Refer to the output of `-analyzer-display-progress` to find the fully qualified +function name. + +There are cases when this name can still collide. For example with template +function instances with non-deducible (aka. explicit) template parameters. +In such cases, prefer passing a USR instead of a function name can resolve this +ambiguity, like this: `-analyze-function="c:@S@Window@F@overloaded#I#"`. + +Use the `clang-extdef-mapping` tool to find the USR for different functions. General Analysis Dumpers ======================== diff --git a/clang/include/clang/Basic/BuiltinsX86.td b/clang/include/clang/Basic/BuiltinsX86.td index e98bee2..a0181b7 100644 --- a/clang/include/clang/Basic/BuiltinsX86.td +++ b/clang/include/clang/Basic/BuiltinsX86.td @@ -52,7 +52,7 @@ def emms : X86Builtin<"void()"> { let Features = "mmx"; } -let Attributes = [NoThrow, Const, RequiredVectorWidth<64>], Features = "sse" in { +let Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<64>], Features = "sse" in { def vec_ext_v4hi : X86Builtin<"short(_Vector<4, short>, _Constant int)">; def vec_set_v4hi : X86Builtin<"_Vector<4, short>(_Vector<4, short>, short, _Constant int)">; } @@ -92,13 +92,6 @@ let Attributes = [Const, NoThrow, RequiredVectorWidth<128>] in { def cmpsd : X86Builtin<"_Vector<2, double>(_Vector<2, double>, _Vector<2, double>, _Constant char)">; } - let Features = "sse2" in { - def vec_ext_v2di : X86Builtin<"long long int(_Vector<2, long long int>, _Constant int)">; - def vec_ext_v4si : X86Builtin<"int(_Vector<4, int>, _Constant int)">; - def vec_ext_v4sf : X86Builtin<"float(_Vector<4, float>, _Constant int)">; - def vec_ext_v8hi : X86Builtin<"short(_Vector<8, short>, _Constant int)">; - def vec_set_v8hi : X86Builtin<"_Vector<8, short>(_Vector<8, short>, short, _Constant int)">; - } let Features = "sse2", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { def pavgb128 : X86Builtin<"_Vector<16, unsigned char>(_Vector<16, unsigned char>, _Vector<16, unsigned char>)">; @@ -108,6 +101,12 @@ let Attributes = [Const, NoThrow, RequiredVectorWidth<128>] in { def packsswb128 : X86Builtin<"_Vector<16, char>(_Vector<8, short>, _Vector<8, short>)">; def packssdw128 : X86Builtin<"_Vector<8, short>(_Vector<4, int>, _Vector<4, int>)">; def packuswb128 : X86Builtin<"_Vector<16, char>(_Vector<8, short>, _Vector<8, short>)">; + + def vec_ext_v2di : X86Builtin<"long long int(_Vector<2, long long int>, _Constant int)">; + def vec_ext_v4si : X86Builtin<"int(_Vector<4, int>, _Constant int)">; + def vec_ext_v4sf : X86Builtin<"float(_Vector<4, float>, _Constant int)">; + def vec_ext_v8hi : X86Builtin<"short(_Vector<8, short>, _Constant int)">; + def vec_set_v8hi : X86Builtin<"_Vector<8, short>(_Vector<8, short>, short, _Constant int)">; } let Features = "sse3" in { @@ -217,10 +216,13 @@ let Features = "sse2", Attributes = [NoThrow] in { def movnti : X86Builtin<"void(int *, int)">; } -let Features = "sse2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in { - def pshufd : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Constant int)">; +let Features = "sse2", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { def pshuflw : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Constant int)">; + def pshufd : X86Builtin<"_Vector<4, int>(_Vector<4, int>, _Constant int)">; def pshufhw : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Constant int)">; +} + +let Features = "sse2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in { def psadbw128 : X86Builtin<"_Vector<2, long long int>(_Vector<16, char>, _Vector<16, char>)">; def sqrtpd : X86Builtin<"_Vector<2, double>(_Vector<2, double>)">; def sqrtsd : X86Builtin<"_Vector<2, double>(_Vector<2, double>)">; @@ -323,9 +325,6 @@ let Features = "sse4.1", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] def ptestnzc128 : X86Builtin<"int(_Vector<2, long long int>, _Vector<2, long long int>)">; def mpsadbw128 : X86Builtin<"_Vector<16, char>(_Vector<16, char>, _Vector<16, char>, _Constant char)">; def phminposuw128 : X86Builtin<"_Vector<8, short>(_Vector<8, short>)">; - def vec_ext_v16qi : X86Builtin<"char(_Vector<16, char>, _Constant int)">; - def vec_set_v16qi : X86Builtin<"_Vector<16, char>(_Vector<16, char>, char, _Constant int)">; - def vec_set_v4si : X86Builtin<"_Vector<4, int>(_Vector<4, int>, int, _Constant int)">; } let Features = "sse4.1", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { @@ -338,6 +337,10 @@ let Features = "sse4.1", Attributes = [NoThrow, Const, Constexpr, RequiredVector def pmuldq128 : X86Builtin<"_Vector<2, long long int>(_Vector<4, int>, _Vector<4, int>)">; def packusdw128 : X86Builtin<"_Vector<8, short>(_Vector<4, int>, _Vector<4, int>)">; + + def vec_ext_v16qi : X86Builtin<"char(_Vector<16, char>, _Constant int)">; + def vec_set_v16qi : X86Builtin<"_Vector<16, char>(_Vector<16, char>, char, _Constant int)">; + def vec_set_v4si : X86Builtin<"_Vector<4, int>(_Vector<4, int>, int, _Constant int)">; } let Features = "sse4.2", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in { @@ -560,7 +563,7 @@ let Features = "avx", Attributes = [NoThrow, RequiredVectorWidth<128>] in { def maskstoreps : X86Builtin<"void(_Vector<4, float *>, _Vector<4, int>, _Vector<4, float>)">; } -let Features = "avx", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in { +let Features = "avx", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<256>] in { def vec_ext_v32qi : X86Builtin<"char(_Vector<32, char>, _Constant int)">; def vec_ext_v16hi : X86Builtin<"short(_Vector<16, short>, _Constant int)">; def vec_ext_v8si : X86Builtin<"int(_Vector<8, int>, _Constant int)">; @@ -584,9 +587,6 @@ let Features = "avx2", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] i def pmulhrsw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">; def psadbw256 : X86Builtin<"_Vector<4, long long int>(_Vector<32, char>, _Vector<32, char>)">; def pshufb256 : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>)">; - def pshufd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Constant int)">; - def pshuflw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Constant int)">; - def pshufhw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Constant int)">; def psignb256 : X86Builtin<"_Vector<32, char>(_Vector<32, char>, _Vector<32, char>)">; def psignw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Vector<16, short>)">; def psignd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Vector<8, int>)">; @@ -647,6 +647,10 @@ let Features = "avx2", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWi def packsswb256 : X86Builtin<"_Vector<32, char>(_Vector<16, short>, _Vector<16, short>)">; def packssdw256 : X86Builtin<"_Vector<16, short>(_Vector<8, int>, _Vector<8, int>)">; def packuswb256 : X86Builtin<"_Vector<32, char>(_Vector<16, short>, _Vector<16, short>)">; + + def pshuflw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Constant int)">; + def pshufhw256 : X86Builtin<"_Vector<16, short>(_Vector<16, short>, _Constant int)">; + def pshufd256 : X86Builtin<"_Vector<8, int>(_Vector<8, int>, _Constant int)">; } let Features = "avx2", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { @@ -1017,6 +1021,7 @@ let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<512> let Features = "avx512f", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<512>] in { def pmuldq512 : X86Builtin<"_Vector<8, long long int>(_Vector<16, int>, _Vector<16, int>)">; def pmuludq512 : X86Builtin<"_Vector<8, long long int>(_Vector<16, int>, _Vector<16, int>)">; + def pshufd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Constant int)">; } let Features = "avx512f", Attributes = [NoThrow, RequiredVectorWidth<512>] in { @@ -1990,13 +1995,13 @@ let Features = "avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVect } let Features = "avx512bw", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in { - def pshufhw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Constant int)">; - def pshuflw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Constant int)">; def psllw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<8, short>)">; } let Features = "avx512bw", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<512>] in { def psllv32hi : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Vector<32, short>)">; + def pshufhw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Constant int)">; + def pshuflw512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, _Constant int)">; } let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<256>] in { @@ -2026,8 +2031,7 @@ let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, Constexpr, Req def psrlv8hi : X86Builtin<"_Vector<8, short>(_Vector<8, short>, _Vector<8, short>)">; } -let Features = "avx512f", - Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<512>] in { +let Features = "avx512f", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<512>] in { def psrlwi512 : X86Builtin<"_Vector<32, short>(_Vector<32, short>, int)">; def psrldi512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, int)">; def psrlqi512 : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, int)">; @@ -3266,7 +3270,6 @@ let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<128> } let Features = "avx512f", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in { - def pshufd512 : X86Builtin<"_Vector<16, int>(_Vector<16, int>, _Constant int)">; def expanddf512_mask : X86Builtin<"_Vector<8, double>(_Vector<8, double>, _Vector<8, double>, unsigned char)">; def expanddi512_mask : X86Builtin<"_Vector<8, long long int>(_Vector<8, long long int>, _Vector<8, long long int>, unsigned char)">; } diff --git a/clang/include/clang/Basic/BuiltinsX86_64.td b/clang/include/clang/Basic/BuiltinsX86_64.td index 214b175..275278c 100644 --- a/clang/include/clang/Basic/BuiltinsX86_64.td +++ b/clang/include/clang/Basic/BuiltinsX86_64.td @@ -56,7 +56,7 @@ let Features = "sse2", Attributes = [NoThrow] in { def movnti64 : X86Builtin<"void(long long int *, long long int)">; } -let Features = "sse4.1", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in { +let Features = "sse4.1", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { def vec_set_v2di : X86Builtin<"_Vector<2, long long int>(_Vector<2, long long int>, long long int, _Constant int)">; } @@ -64,7 +64,7 @@ let Features = "crc32", Attributes = [NoThrow, Const] in { def crc32di : X86Builtin<"unsigned long long int(unsigned long long int, unsigned long long int)">; } -let Features = "avx", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in { +let Features = "avx", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<256>] in { def vec_ext_v4di : X86Builtin<"long long int(_Vector<4, long long int>, _Constant int)">; def vec_set_v4di : X86Builtin<"_Vector<4, long long int>(_Vector<4, long long int>, long long int, _Constant int)">; } diff --git a/clang/include/clang/CIR/Dialect/Builder/CIRBaseBuilder.h b/clang/include/clang/CIR/Dialect/Builder/CIRBaseBuilder.h index 8a5bf03..93d81e3 100644 --- a/clang/include/clang/CIR/Dialect/Builder/CIRBaseBuilder.h +++ b/clang/include/clang/CIR/Dialect/Builder/CIRBaseBuilder.h @@ -10,6 +10,7 @@ #define LLVM_CLANG_CIR_DIALECT_BUILDER_CIRBASEBUILDER_H #include "clang/AST/CharUnits.h" +#include "clang/Basic/AddressSpaces.h" #include "clang/CIR/Dialect/IR/CIRAttrs.h" #include "clang/CIR/Dialect/IR/CIRDialect.h" #include "clang/CIR/Dialect/IR/CIRTypes.h" @@ -129,8 +130,30 @@ public: return cir::PointerType::get(ty); } - cir::PointerType getVoidPtrTy() { - return getPointerTo(cir::VoidType::get(getContext())); + cir::PointerType getPointerTo(mlir::Type ty, cir::TargetAddressSpaceAttr as) { + return cir::PointerType::get(ty, as); + } + + cir::PointerType getPointerTo(mlir::Type ty, clang::LangAS langAS) { + if (langAS == clang::LangAS::Default) // Default address space. + return getPointerTo(ty); + + if (clang::isTargetAddressSpace(langAS)) { + unsigned addrSpace = clang::toTargetAddressSpace(langAS); + auto asAttr = cir::TargetAddressSpaceAttr::get( + getContext(), getUI32IntegerAttr(addrSpace)); + return getPointerTo(ty, asAttr); + } + + llvm_unreachable("language-specific address spaces NYI"); + } + + cir::PointerType getVoidPtrTy(clang::LangAS langAS = clang::LangAS::Default) { + return getPointerTo(cir::VoidType::get(getContext()), langAS); + } + + cir::PointerType getVoidPtrTy(cir::TargetAddressSpaceAttr as) { + return getPointerTo(cir::VoidType::get(getContext()), as); } cir::BoolAttr getCIRBoolAttr(bool state) { diff --git a/clang/include/clang/CIR/Dialect/IR/CIRAttrs.h b/clang/include/clang/CIR/Dialect/IR/CIRAttrs.h index 925a9a8..03a6a97 100644 --- a/clang/include/clang/CIR/Dialect/IR/CIRAttrs.h +++ b/clang/include/clang/CIR/Dialect/IR/CIRAttrs.h @@ -15,6 +15,7 @@ #include "mlir/IR/Attributes.h" #include "mlir/IR/BuiltinAttributeInterfaces.h" +#include "clang/Basic/AddressSpaces.h" #include "clang/CIR/Dialect/IR/CIROpsEnums.h" diff --git a/clang/include/clang/CIR/Dialect/IR/CIRAttrs.td b/clang/include/clang/CIR/Dialect/IR/CIRAttrs.td index f8358de..7714750 100644 --- a/clang/include/clang/CIR/Dialect/IR/CIRAttrs.td +++ b/clang/include/clang/CIR/Dialect/IR/CIRAttrs.td @@ -602,6 +602,33 @@ def CIR_VTableAttr : CIR_Attr<"VTable", "vtable", [TypedAttrInterface]> { } //===----------------------------------------------------------------------===// +// TargetAddressSpaceAttr +//===----------------------------------------------------------------------===// + +def CIR_TargetAddressSpaceAttr : CIR_Attr< "TargetAddressSpace", + "target_address_space"> { + let summary = "Represents a target-specific numeric address space"; + let description = [{ + The TargetAddressSpaceAttr represents a target-specific numeric address space, + corresponding to the LLVM IR `addressspace` qualifier and the clang + `address_space` attribute. + + A value of zero represents the default address space. The semantics of non-zero + address spaces are target-specific. + + Example: + ```mlir + // Target-specific numeric address spaces + !cir.ptr<!s32i, addrspace(target<1>)> + !cir.ptr<!s32i, addrspace(target<10>)> + ``` + }]; + + let parameters = (ins "mlir::IntegerAttr":$value); + let assemblyFormat = "`<` `target` `<` $value `>` `>`"; +} + +//===----------------------------------------------------------------------===// // ConstComplexAttr //===----------------------------------------------------------------------===// diff --git a/clang/include/clang/CIR/Dialect/IR/CIROps.td b/clang/include/clang/CIR/Dialect/IR/CIROps.td index 0a78492..7f2e55d 100644 --- a/clang/include/clang/CIR/Dialect/IR/CIROps.td +++ b/clang/include/clang/CIR/Dialect/IR/CIROps.td @@ -2341,6 +2341,12 @@ def CIR_FuncOp : CIR_Op<"func", [ The function linkage information is specified by `linkage`, as defined by `GlobalLinkageKind` attribute. + A compiler builtin function must be marked as `builtin` for further + processing when lowering from CIR. + + The `coroutine` keyword is used to mark a coroutine function, which requires + at least one `cir.await` instruction to be used in its body. + The `lambda` translates to a C++ `operator()` that implements a lambda, this allow callsites to make certain assumptions about the real function nature when writing analysis. @@ -2362,11 +2368,22 @@ def CIR_FuncOp : CIR_Op<"func", [ // Linkage information cir.func linkonce_odr @some_method(...) ``` + // Builtin function + cir.func builtin @__builtin_coro_end(!cir.ptr<i8>, !cir.bool) -> !cir.bool + // Coroutine + cir.func coroutine @_Z10silly_taskv() -> !CoroTask { + ... + cir.await(...) + ... + } + ``` }]; let arguments = (ins SymbolNameAttr:$sym_name, CIR_VisibilityAttr:$global_visibility, TypeAttrOf<CIR_FuncType>:$function_type, + UnitAttr:$builtin, + UnitAttr:$coroutine, UnitAttr:$lambda, UnitAttr:$no_proto, UnitAttr:$dso_local, diff --git a/clang/include/clang/CIR/Dialect/IR/CIRTypes.h b/clang/include/clang/CIR/Dialect/IR/CIRTypes.h index bfa165c..45f646f 100644 --- a/clang/include/clang/CIR/Dialect/IR/CIRTypes.h +++ b/clang/include/clang/CIR/Dialect/IR/CIRTypes.h @@ -16,6 +16,9 @@ #include "mlir/IR/BuiltinAttributes.h" #include "mlir/IR/Types.h" #include "mlir/Interfaces/DataLayoutInterfaces.h" +#include "clang/Basic/AddressSpaces.h" +#include "clang/CIR/Dialect/IR/CIRAttrs.h" +#include "clang/CIR/Dialect/IR/CIROpsEnums.h" #include "clang/CIR/Interfaces/CIRTypeInterfaces.h" namespace cir { diff --git a/clang/include/clang/CIR/Dialect/IR/CIRTypes.td b/clang/include/clang/CIR/Dialect/IR/CIRTypes.td index 4eec34c..3131847 100644 --- a/clang/include/clang/CIR/Dialect/IR/CIRTypes.td +++ b/clang/include/clang/CIR/Dialect/IR/CIRTypes.td @@ -14,10 +14,12 @@ #define CLANG_CIR_DIALECT_IR_CIRTYPES_TD include "clang/CIR/Dialect/IR/CIRDialect.td" +include "clang/CIR/Dialect/IR/CIREnumAttr.td" include "clang/CIR/Dialect/IR/CIRTypeConstraints.td" include "clang/CIR/Interfaces/CIRTypeInterfaces.td" include "mlir/Interfaces/DataLayoutInterfaces.td" include "mlir/IR/AttrTypeBase.td" +include "mlir/IR/EnumAttr.td" //===----------------------------------------------------------------------===// // CIR Types @@ -226,32 +228,54 @@ def CIR_PointerType : CIR_Type<"Pointer", "ptr", [ ]> { let summary = "CIR pointer type"; let description = [{ - The `!cir.ptr` type represents C and C++ pointer types and C++ reference - types, other than pointers-to-members. The `pointee` type is the type - pointed to. + The `!cir.ptr` type is a typed pointer type. It is used to represent + pointers to objects in C/C++. The type of the pointed-to object is given by + the `pointee` parameter. The `addrSpace` parameter is an optional address + space attribute that specifies the address space of the pointer. If not + specified, the pointer is assumed to be in the default address space. - TODO(CIR): The address space attribute is not yet implemented. + The `!cir.ptr` type can point to any type, including fundamental types, + records, arrays, vectors, functions, and other pointers. It can also point + to incomplete types, such as incomplete records. + + Examples: + + ```mlir + !cir.ptr<!cir.int<u, 8>> + !cir.ptr<!cir.float> + !cir.ptr<!cir.record<struct "MyStruct">> + !cir.ptr<!cir.int<u, 8>, target_address_space(1)> + !cir.ptr<!cir.record<struct "MyStruct">, target_address_space(5)> + ``` }]; - let parameters = (ins "mlir::Type":$pointee); + let parameters = (ins + "mlir::Type":$pointee, + OptionalParameter< + "cir::TargetAddressSpaceAttr">:$addrSpace + ); + let skipDefaultBuilders = 1; let builders = [ - TypeBuilderWithInferredContext<(ins "mlir::Type":$pointee), [{ - return $_get(pointee.getContext(), pointee); + TypeBuilderWithInferredContext<(ins + "mlir::Type":$pointee, + CArg<"cir::TargetAddressSpaceAttr", "nullptr">:$addrSpace), [{ + return $_get(pointee.getContext(), pointee, addrSpace); }]>, - TypeBuilder<(ins "mlir::Type":$pointee), [{ - return $_get($_ctxt, pointee); + TypeBuilder<(ins + "mlir::Type":$pointee, + CArg<"cir::TargetAddressSpaceAttr", "nullptr">:$addrSpace), [{ + return $_get($_ctxt, pointee, addrSpace); }]> ]; let assemblyFormat = [{ - `<` $pointee `>` + `<` + $pointee + ( `,` ` ` custom<TargetAddressSpace>($addrSpace)^ )? + `>` }]; - let genVerifyDecl = 1; - - let skipDefaultBuilders = 1; - let extraClassDeclaration = [{ template <typename ...Types> bool isPtrTo() const { diff --git a/clang/include/clang/CIR/MissingFeatures.h b/clang/include/clang/CIR/MissingFeatures.h index 3dfcafc..f7ca276 100644 --- a/clang/include/clang/CIR/MissingFeatures.h +++ b/clang/include/clang/CIR/MissingFeatures.h @@ -136,6 +136,13 @@ struct MissingFeatures { static bool recordZeroInitPadding() { return false; } static bool zeroSizeRecordMembers() { return false; } + // Coroutines + static bool coroAllocBuiltinCall() { return false; } + static bool coroBeginBuiltinCall() { return false; } + static bool coroEndBuiltinCall() { return false; } + static bool coroSizeBuiltinCall() { return false; } + static bool coroutineFrame() { return false; } + // Various handling of deferred processing in CIRGenModule. static bool cgmRelease() { return false; } static bool deferredVtables() { return false; } @@ -209,6 +216,7 @@ struct MissingFeatures { static bool dataLayoutTypeIsSized() { return false; } static bool dataLayoutTypeAllocSize() { return false; } static bool dataLayoutTypeStoreSize() { return false; } + static bool dataLayoutPtrHandlingBasedOnLangAS() { return false; } static bool deferredCXXGlobalInit() { return false; } static bool deleteArray() { return false; } static bool devirtualizeMemberFunction() { return false; } diff --git a/clang/include/clang/CrossTU/CrossTranslationUnit.h b/clang/include/clang/CrossTU/CrossTranslationUnit.h index e6b608a1..9e0721e 100644 --- a/clang/include/clang/CrossTU/CrossTranslationUnit.h +++ b/clang/include/clang/CrossTU/CrossTranslationUnit.h @@ -180,8 +180,8 @@ public: llvm::Expected<const VarDecl *> importDefinition(const VarDecl *VD, ASTUnit *Unit); - /// Get a name to identify a named decl. - static std::optional<std::string> getLookupName(const NamedDecl *ND); + /// Get a name to identify a decl. + static std::optional<std::string> getLookupName(const Decl *D); /// Emit diagnostics for the user for potential configuration errors. void emitCrossTUDiagnostics(const IndexError &IE); diff --git a/clang/include/clang/Driver/Options.td b/clang/include/clang/Driver/Options.td index 2ef6098..5a48f0b 100644 --- a/clang/include/clang/Driver/Options.td +++ b/clang/include/clang/Driver/Options.td @@ -1258,8 +1258,9 @@ def offload_compression_level_EQ : Joined<["--"], "offload-compression-level=">, HelpText<"Compression level for offload device binaries (HIP only)">; def offload_jobs_EQ : Joined<["--"], "offload-jobs=">, - HelpText<"Specify the number of threads to use for device offloading tasks" - " during compilation.">; + HelpText<"Specify the number of threads to use for device offloading tasks " + "during compilation. Can be a positive integer or the string " + "'jobserver' to use the make-style jobserver from the environment.">; defm offload_via_llvm : BoolFOption<"offload-via-llvm", LangOpts<"OffloadViaLLVM">, DefaultFalse, diff --git a/clang/include/clang/StaticAnalyzer/Core/PathSensitive/EntryPointStats.h b/clang/include/clang/StaticAnalyzer/Core/PathSensitive/EntryPointStats.h index 633fb7a..448e402 100644 --- a/clang/include/clang/StaticAnalyzer/Core/PathSensitive/EntryPointStats.h +++ b/clang/include/clang/StaticAnalyzer/Core/PathSensitive/EntryPointStats.h @@ -25,7 +25,7 @@ class EntryPointStat { public: llvm::StringLiteral name() const { return Name; } - static void lockRegistry(); + static void lockRegistry(llvm::StringRef CPPFileName); static void takeSnapshot(const Decl *EntryPoint); static void dumpStatsAsCSV(llvm::raw_ostream &OS); diff --git a/clang/lib/AST/ByteCode/InterpBuiltin.cpp b/clang/lib/AST/ByteCode/InterpBuiltin.cpp index a2e97fc..6053237 100644 --- a/clang/lib/AST/ByteCode/InterpBuiltin.cpp +++ b/clang/lib/AST/ByteCode/InterpBuiltin.cpp @@ -2773,6 +2773,50 @@ static bool interp__builtin_blend(InterpState &S, CodePtr OpPC, return true; } +static bool interp__builtin_ia32_pshuf(InterpState &S, CodePtr OpPC, + const CallExpr *Call, bool IsShufHW) { + assert(Call->getNumArgs() == 2 && "masked forms handled via select*"); + APSInt ControlImm = popToAPSInt(S, Call->getArg(1)); + const Pointer &Src = S.Stk.pop<Pointer>(); + const Pointer &Dst = S.Stk.peek<Pointer>(); + + unsigned NumElems = Dst.getNumElems(); + PrimType ElemT = Dst.getFieldDesc()->getPrimType(); + + unsigned ElemBits = static_cast<unsigned>(primSize(ElemT) * 8); + if (ElemBits != 16 && ElemBits != 32) + return false; + + unsigned LaneElts = 128u / ElemBits; + assert(LaneElts && (NumElems % LaneElts == 0)); + + uint8_t Ctl = static_cast<uint8_t>(ControlImm.getZExtValue()); + + for (unsigned Idx = 0; Idx != NumElems; Idx++) { + unsigned LaneBase = (Idx / LaneElts) * LaneElts; + unsigned LaneIdx = Idx % LaneElts; + unsigned SrcIdx = Idx; + unsigned Sel = (Ctl >> (2 * LaneIdx)) & 0x3; + if (ElemBits == 32) { + SrcIdx = LaneBase + Sel; + } else { + constexpr unsigned HalfSize = 4; + bool InHigh = LaneIdx >= HalfSize; + if (!IsShufHW && !InHigh) { + SrcIdx = LaneBase + Sel; + } else if (IsShufHW && InHigh) { + unsigned Rel = LaneIdx - HalfSize; + Sel = (Ctl >> (2 * Rel)) & 0x3; + SrcIdx = LaneBase + HalfSize + Sel; + } + } + + INT_TYPE_SWITCH_NO_BOOL(ElemT, { Dst.elem<T>(Idx) = Src.elem<T>(SrcIdx); }); + } + Dst.initializeAllElements(); + return true; +} + static bool interp__builtin_elementwise_triop( InterpState &S, CodePtr OpPC, const CallExpr *Call, llvm::function_ref<APInt(const APSInt &, const APSInt &, const APSInt &)> @@ -2878,6 +2922,61 @@ static bool interp__builtin_x86_insert_subvector(InterpState &S, CodePtr OpPC, return true; } +static bool interp__builtin_vec_ext(InterpState &S, CodePtr OpPC, + const CallExpr *Call, unsigned ID) { + assert(Call->getNumArgs() == 2); + + APSInt ImmAPS = popToAPSInt(S, Call->getArg(1)); + const Pointer &Vec = S.Stk.pop<Pointer>(); + if (!Vec.getFieldDesc()->isPrimitiveArray()) + return false; + + unsigned NumElems = Vec.getNumElems(); + unsigned Index = + static_cast<unsigned>(ImmAPS.getZExtValue() & (NumElems - 1)); + + PrimType ElemPT = Vec.getFieldDesc()->getPrimType(); + // FIXME(#161685): Replace float+int split with a numeric-only type switch + if (ElemPT == PT_Float) { + S.Stk.push<Floating>(Vec.elem<Floating>(Index)); + return true; + } + INT_TYPE_SWITCH_NO_BOOL(ElemPT, { + APSInt V = Vec.elem<T>(Index).toAPSInt(); + pushInteger(S, V, Call->getType()); + }); + + return true; +} + +static bool interp__builtin_vec_set(InterpState &S, CodePtr OpPC, + const CallExpr *Call, unsigned ID) { + assert(Call->getNumArgs() == 3); + + APSInt ImmAPS = popToAPSInt(S, Call->getArg(2)); + APSInt ValAPS = popToAPSInt(S, Call->getArg(1)); + + const Pointer &Base = S.Stk.pop<Pointer>(); + if (!Base.getFieldDesc()->isPrimitiveArray()) + return false; + + const Pointer &Dst = S.Stk.peek<Pointer>(); + + unsigned NumElems = Base.getNumElems(); + unsigned Index = + static_cast<unsigned>(ImmAPS.getZExtValue() & (NumElems - 1)); + + PrimType ElemPT = Base.getFieldDesc()->getPrimType(); + INT_TYPE_SWITCH_NO_BOOL(ElemPT, { + for (unsigned I = 0; I != NumElems; ++I) + Dst.elem<T>(I) = Base.elem<T>(I); + Dst.elem<T>(Index) = static_cast<T>(ValAPS); + }); + + Dst.initializeAllElements(); + return true; +} + bool InterpretBuiltin(InterpState &S, CodePtr OpPC, const CallExpr *Call, uint32_t BuiltinID) { if (!S.getASTContext().BuiltinInfo.isConstantEvaluated(BuiltinID)) @@ -3606,6 +3705,21 @@ bool InterpretBuiltin(InterpState &S, CodePtr OpPC, const CallExpr *Call, case X86::BI__builtin_ia32_selectpd_512: return interp__builtin_select(S, OpPC, Call); + case X86::BI__builtin_ia32_pshuflw: + case X86::BI__builtin_ia32_pshuflw256: + case X86::BI__builtin_ia32_pshuflw512: + return interp__builtin_ia32_pshuf(S, OpPC, Call, false); + + case X86::BI__builtin_ia32_pshufhw: + case X86::BI__builtin_ia32_pshufhw256: + case X86::BI__builtin_ia32_pshufhw512: + return interp__builtin_ia32_pshuf(S, OpPC, Call, true); + + case X86::BI__builtin_ia32_pshufd: + case X86::BI__builtin_ia32_pshufd256: + case X86::BI__builtin_ia32_pshufd512: + return interp__builtin_ia32_pshuf(S, OpPC, Call, false); + case X86::BI__builtin_ia32_kandqi: case X86::BI__builtin_ia32_kandhi: case X86::BI__builtin_ia32_kandsi: @@ -3686,6 +3800,29 @@ bool InterpretBuiltin(InterpState &S, CodePtr OpPC, const CallExpr *Call, case X86::BI__builtin_ia32_insert128i256: return interp__builtin_x86_insert_subvector(S, OpPC, Call, BuiltinID); + case X86::BI__builtin_ia32_vec_ext_v4hi: + case X86::BI__builtin_ia32_vec_ext_v16qi: + case X86::BI__builtin_ia32_vec_ext_v8hi: + case X86::BI__builtin_ia32_vec_ext_v4si: + case X86::BI__builtin_ia32_vec_ext_v2di: + case X86::BI__builtin_ia32_vec_ext_v32qi: + case X86::BI__builtin_ia32_vec_ext_v16hi: + case X86::BI__builtin_ia32_vec_ext_v8si: + case X86::BI__builtin_ia32_vec_ext_v4di: + case X86::BI__builtin_ia32_vec_ext_v4sf: + return interp__builtin_vec_ext(S, OpPC, Call, BuiltinID); + + case X86::BI__builtin_ia32_vec_set_v4hi: + case X86::BI__builtin_ia32_vec_set_v16qi: + case X86::BI__builtin_ia32_vec_set_v8hi: + case X86::BI__builtin_ia32_vec_set_v4si: + case X86::BI__builtin_ia32_vec_set_v2di: + case X86::BI__builtin_ia32_vec_set_v32qi: + case X86::BI__builtin_ia32_vec_set_v16hi: + case X86::BI__builtin_ia32_vec_set_v8si: + case X86::BI__builtin_ia32_vec_set_v4di: + return interp__builtin_vec_set(S, OpPC, Call, BuiltinID); + default: S.FFDiag(S.Current->getLocation(OpPC), diag::note_invalid_subexpr_in_const_expr) diff --git a/clang/lib/AST/ExprConstant.cpp b/clang/lib/AST/ExprConstant.cpp index b706b14..7bf28d9 100644 --- a/clang/lib/AST/ExprConstant.cpp +++ b/clang/lib/AST/ExprConstant.cpp @@ -11615,6 +11615,60 @@ static bool evalPackBuiltin(const CallExpr *E, EvalInfo &Info, APValue &Result, return true; } +static bool evalPshufBuiltin(EvalInfo &Info, const CallExpr *Call, + bool IsShufHW, APValue &Out) { + APValue Vec; + APSInt Imm; + if (!EvaluateAsRValue(Info, Call->getArg(0), Vec)) + return false; + if (!EvaluateInteger(Call->getArg(1), Imm, Info)) + return false; + + const auto *VT = Call->getType()->getAs<VectorType>(); + if (!VT) + return false; + + QualType ElemT = VT->getElementType(); + unsigned ElemBits = Info.Ctx.getTypeSize(ElemT); + unsigned NumElts = VT->getNumElements(); + + unsigned LaneBits = 128u; + unsigned LaneElts = LaneBits / ElemBits; + if (!LaneElts || (NumElts % LaneElts) != 0) + return false; + + uint8_t Ctl = static_cast<uint8_t>(Imm.getZExtValue()); + + SmallVector<APValue, 32> ResultElements; + ResultElements.reserve(NumElts); + + for (unsigned Idx = 0; Idx != NumElts; Idx++) { + unsigned LaneBase = (Idx / LaneElts) * LaneElts; + unsigned LaneIdx = Idx % LaneElts; + unsigned SrcIdx = Idx; + unsigned Sel = (Ctl >> (2 * LaneIdx)) & 0x3; + + if (ElemBits == 32) { + SrcIdx = LaneBase + Sel; + } else { + constexpr unsigned HalfSize = 4; + bool InHigh = LaneIdx >= HalfSize; + if (!IsShufHW && !InHigh) { + SrcIdx = LaneBase + Sel; + } else if (IsShufHW && InHigh) { + unsigned Rel = LaneIdx - HalfSize; + Sel = (Ctl >> (2 * Rel)) & 0x3; + SrcIdx = LaneBase + HalfSize + Sel; + } + } + + ResultElements.push_back(Vec.getVectorElt(SrcIdx)); + } + + Out = APValue(ResultElements.data(), ResultElements.size()); + return true; +} + bool VectorExprEvaluator::VisitCallExpr(const CallExpr *E) { if (!IsConstantEvaluatedBuiltinCall(E)) return ExprEvaluatorBaseTy::VisitCallExpr(E); @@ -11868,7 +11922,6 @@ bool VectorExprEvaluator::VisitCallExpr(const CallExpr *E) { return Success(APValue(ResultElements.data(), ResultElements.size()), E); } - case clang::X86::BI__builtin_ia32_vprotbi: case clang::X86::BI__builtin_ia32_vprotdi: case clang::X86::BI__builtin_ia32_vprotqi: @@ -12087,6 +12140,34 @@ bool VectorExprEvaluator::VisitCallExpr(const CallExpr *E) { return Success(APValue(ResultElements.data(), ResultElements.size()), E); } + + case X86::BI__builtin_ia32_pshuflw: + case X86::BI__builtin_ia32_pshuflw256: + case X86::BI__builtin_ia32_pshuflw512: { + APValue R; + if (!evalPshufBuiltin(Info, E, false, R)) + return false; + return Success(R, E); + } + + case X86::BI__builtin_ia32_pshufhw: + case X86::BI__builtin_ia32_pshufhw256: + case X86::BI__builtin_ia32_pshufhw512: { + APValue R; + if (!evalPshufBuiltin(Info, E, true, R)) + return false; + return Success(R, E); + } + + case X86::BI__builtin_ia32_pshufd: + case X86::BI__builtin_ia32_pshufd256: + case X86::BI__builtin_ia32_pshufd512: { + APValue R; + if (!evalPshufBuiltin(Info, E, false, R)) + return false; + return Success(R, E); + } + case Builtin::BI__builtin_elementwise_clzg: case Builtin::BI__builtin_elementwise_ctzg: { APValue SourceLHS; @@ -12235,6 +12316,41 @@ bool VectorExprEvaluator::VisitCallExpr(const CallExpr *E) { return Success(APValue(ResultElements.data(), ResultElements.size()), E); } + + case clang::X86::BI__builtin_ia32_vec_set_v4hi: + case clang::X86::BI__builtin_ia32_vec_set_v16qi: + case clang::X86::BI__builtin_ia32_vec_set_v8hi: + case clang::X86::BI__builtin_ia32_vec_set_v4si: + case clang::X86::BI__builtin_ia32_vec_set_v2di: + case clang::X86::BI__builtin_ia32_vec_set_v32qi: + case clang::X86::BI__builtin_ia32_vec_set_v16hi: + case clang::X86::BI__builtin_ia32_vec_set_v8si: + case clang::X86::BI__builtin_ia32_vec_set_v4di: { + APValue VecVal; + APSInt Scalar, IndexAPS; + if (!EvaluateVector(E->getArg(0), VecVal, Info) || + !EvaluateInteger(E->getArg(1), Scalar, Info) || + !EvaluateInteger(E->getArg(2), IndexAPS, Info)) + return false; + + QualType ElemTy = E->getType()->castAs<VectorType>()->getElementType(); + unsigned ElemWidth = Info.Ctx.getIntWidth(ElemTy); + bool ElemUnsigned = ElemTy->isUnsignedIntegerOrEnumerationType(); + Scalar.setIsUnsigned(ElemUnsigned); + APSInt ElemAPS = Scalar.extOrTrunc(ElemWidth); + APValue ElemAV(ElemAPS); + + unsigned NumElems = VecVal.getVectorLength(); + unsigned Index = + static_cast<unsigned>(IndexAPS.getZExtValue() & (NumElems - 1)); + + SmallVector<APValue, 4> Elems; + Elems.reserve(NumElems); + for (unsigned ElemNum = 0; ElemNum != NumElems; ++ElemNum) + Elems.push_back(ElemNum == Index ? ElemAV : VecVal.getVectorElt(ElemNum)); + + return Success(APValue(Elems.data(), NumElems), E); + } } } @@ -14822,6 +14938,25 @@ bool IntExprEvaluator::VisitBuiltinCallExpr(const CallExpr *E, return HandleMaskBinOp( [](const APSInt &LHS, const APSInt &RHS) { return LHS + RHS; }); } + + case clang::X86::BI__builtin_ia32_vec_ext_v4hi: + case clang::X86::BI__builtin_ia32_vec_ext_v16qi: + case clang::X86::BI__builtin_ia32_vec_ext_v8hi: + case clang::X86::BI__builtin_ia32_vec_ext_v4si: + case clang::X86::BI__builtin_ia32_vec_ext_v2di: + case clang::X86::BI__builtin_ia32_vec_ext_v32qi: + case clang::X86::BI__builtin_ia32_vec_ext_v16hi: + case clang::X86::BI__builtin_ia32_vec_ext_v8si: + case clang::X86::BI__builtin_ia32_vec_ext_v4di: { + APValue Vec; + APSInt IdxAPS; + if (!EvaluateVector(E->getArg(0), Vec, Info) || + !EvaluateInteger(E->getArg(1), IdxAPS, Info)) + return false; + unsigned N = Vec.getVectorLength(); + unsigned Idx = static_cast<unsigned>(IdxAPS.getZExtValue() & (N - 1)); + return Success(Vec.getVectorElt(Idx).getInt(), E); + } } } @@ -16638,6 +16773,17 @@ bool FloatExprEvaluator::VisitCallExpr(const CallExpr *E) { (void)Result.fusedMultiplyAdd(SourceY, SourceZ, RM); return true; } + + case clang::X86::BI__builtin_ia32_vec_ext_v4sf: { + APValue Vec; + APSInt IdxAPS; + if (!EvaluateVector(E->getArg(0), Vec, Info) || + !EvaluateInteger(E->getArg(1), IdxAPS, Info)) + return false; + unsigned N = Vec.getVectorLength(); + unsigned Idx = static_cast<unsigned>(IdxAPS.getZExtValue() & (N - 1)); + return Success(Vec.getVectorElt(Idx), E); + } } } diff --git a/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp b/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp index cf17de1..4cfa91e 100644 --- a/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp +++ b/clang/lib/CIR/CodeGen/CIRGenBuiltin.cpp @@ -428,6 +428,32 @@ RValue CIRGenFunction::emitBuiltinExpr(const GlobalDecl &gd, unsigned builtinID, return emitUnaryFPBuiltin<cir::ATanOp>(*this, *e); case Builtin::BI__builtin_elementwise_cos: return emitUnaryFPBuiltin<cir::CosOp>(*this, *e); + case Builtin::BI__builtin_coro_id: + case Builtin::BI__builtin_coro_promise: + case Builtin::BI__builtin_coro_resume: + case Builtin::BI__builtin_coro_noop: + case Builtin::BI__builtin_coro_destroy: + case Builtin::BI__builtin_coro_done: + case Builtin::BI__builtin_coro_alloc: + case Builtin::BI__builtin_coro_begin: + case Builtin::BI__builtin_coro_end: + case Builtin::BI__builtin_coro_suspend: + case Builtin::BI__builtin_coro_align: + cgm.errorNYI(e->getSourceRange(), "BI__builtin_coro_id like NYI"); + return getUndefRValue(e->getType()); + + case Builtin::BI__builtin_coro_frame: { + cgm.errorNYI(e->getSourceRange(), "BI__builtin_coro_frame NYI"); + assert(!cir::MissingFeatures::coroutineFrame()); + return getUndefRValue(e->getType()); + } + case Builtin::BI__builtin_coro_free: + case Builtin::BI__builtin_coro_size: { + cgm.errorNYI(e->getSourceRange(), + "BI__builtin_coro_free, BI__builtin_coro_size NYI"); + assert(!cir::MissingFeatures::coroSizeBuiltinCall()); + return getUndefRValue(e->getType()); + } } // If this is an alias for a lib function (e.g. __builtin_sin), emit diff --git a/clang/lib/CIR/CodeGen/CIRGenCoroutine.cpp b/clang/lib/CIR/CodeGen/CIRGenCoroutine.cpp new file mode 100644 index 0000000..c25cce4 --- /dev/null +++ b/clang/lib/CIR/CodeGen/CIRGenCoroutine.cpp @@ -0,0 +1,82 @@ +//===----- CGCoroutine.cpp - Emit CIR Code for C++ coroutines -------------===// +// +// 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 +// +//===----------------------------------------------------------------------===// +// +// This contains code dealing with C++ code generation of coroutines. +// +//===----------------------------------------------------------------------===// + +#include "CIRGenFunction.h" +#include "mlir/Support/LLVM.h" +#include "clang/AST/StmtCXX.h" +#include "clang/Basic/TargetInfo.h" +#include "clang/CIR/Dialect/IR/CIRTypes.h" + +using namespace clang; +using namespace clang::CIRGen; + +struct clang::CIRGen::CGCoroData { + // Stores the __builtin_coro_id emitted in the function so that we can supply + // it as the first argument to other builtins. + cir::CallOp coroId = nullptr; +}; + +// Defining these here allows to keep CGCoroData private to this file. +CIRGenFunction::CGCoroInfo::CGCoroInfo() {} +CIRGenFunction::CGCoroInfo::~CGCoroInfo() {} + +static void createCoroData(CIRGenFunction &cgf, + CIRGenFunction::CGCoroInfo &curCoro, + cir::CallOp coroId) { + assert(!curCoro.data && "EmitCoroutineBodyStatement called twice?"); + + curCoro.data = std::make_unique<CGCoroData>(); + curCoro.data->coroId = coroId; +} + +cir::CallOp CIRGenFunction::emitCoroIDBuiltinCall(mlir::Location loc, + mlir::Value nullPtr) { + cir::IntType int32Ty = builder.getUInt32Ty(); + + const TargetInfo &ti = cgm.getASTContext().getTargetInfo(); + unsigned newAlign = ti.getNewAlign() / ti.getCharWidth(); + + mlir::Operation *builtin = cgm.getGlobalValue(cgm.builtinCoroId); + + cir::FuncOp fnOp; + if (!builtin) { + fnOp = cgm.createCIRBuiltinFunction( + loc, cgm.builtinCoroId, + cir::FuncType::get({int32Ty, VoidPtrTy, VoidPtrTy, VoidPtrTy}, int32Ty), + /*FD=*/nullptr); + assert(fnOp && "should always succeed"); + } else { + fnOp = cast<cir::FuncOp>(builtin); + } + + return builder.createCallOp(loc, fnOp, + mlir::ValueRange{builder.getUInt32(newAlign, loc), + nullPtr, nullPtr, nullPtr}); +} + +mlir::LogicalResult +CIRGenFunction::emitCoroutineBody(const CoroutineBodyStmt &s) { + mlir::Location openCurlyLoc = getLoc(s.getBeginLoc()); + cir::ConstantOp nullPtrCst = builder.getNullPtr(VoidPtrTy, openCurlyLoc); + + auto fn = mlir::cast<cir::FuncOp>(curFn); + fn.setCoroutine(true); + cir::CallOp coroId = emitCoroIDBuiltinCall(openCurlyLoc, nullPtrCst); + createCoroData(*this, curCoro, coroId); + + assert(!cir::MissingFeatures::coroAllocBuiltinCall()); + + assert(!cir::MissingFeatures::coroBeginBuiltinCall()); + + assert(!cir::MissingFeatures::generateDebugInfo()); + return mlir::success(); +} diff --git a/clang/lib/CIR/CodeGen/CIRGenExpr.cpp b/clang/lib/CIR/CodeGen/CIRGenExpr.cpp index fa68ad9..be94890 100644 --- a/clang/lib/CIR/CodeGen/CIRGenExpr.cpp +++ b/clang/lib/CIR/CodeGen/CIRGenExpr.cpp @@ -1108,8 +1108,9 @@ CIRGenFunction::emitArraySubscriptExpr(const clang::ArraySubscriptExpr *e) { return lv; } -LValue CIRGenFunction::emitStringLiteralLValue(const StringLiteral *e) { - cir::GlobalOp globalOp = cgm.getGlobalForStringLiteral(e); +LValue CIRGenFunction::emitStringLiteralLValue(const StringLiteral *e, + llvm::StringRef name) { + cir::GlobalOp globalOp = cgm.getGlobalForStringLiteral(e, name); assert(globalOp.getAlignment() && "expected alignment for string literal"); unsigned align = *(globalOp.getAlignment()); mlir::Value addr = @@ -2052,8 +2053,8 @@ mlir::Value CIRGenFunction::emitAlloca(StringRef name, mlir::Type ty, // CIR uses its own alloca address space rather than follow the target data // layout like original CodeGen. The data layout awareness should be done in // the lowering pass instead. - assert(!cir::MissingFeatures::addressSpace()); - cir::PointerType localVarPtrTy = builder.getPointerTo(ty); + cir::PointerType localVarPtrTy = + builder.getPointerTo(ty, getCIRAllocaAddressSpace()); mlir::IntegerAttr alignIntAttr = cgm.getSize(alignment); mlir::Value addr; @@ -2372,6 +2373,21 @@ mlir::Value CIRGenFunction::emitScalarConstant( return builder.getConstant(getLoc(e->getSourceRange()), constant.getValue()); } +LValue CIRGenFunction::emitPredefinedLValue(const PredefinedExpr *e) { + const StringLiteral *sl = e->getFunctionName(); + assert(sl != nullptr && "No StringLiteral name in PredefinedExpr"); + auto fn = cast<cir::FuncOp>(curFn); + StringRef fnName = fn.getName(); + fnName.consume_front("\01"); + std::array<StringRef, 2> nameItems = { + PredefinedExpr::getIdentKindName(e->getIdentKind()), fnName}; + std::string gvName = llvm::join(nameItems, "."); + if (isa_and_nonnull<BlockDecl>(curCodeDecl)) + cgm.errorNYI(e->getSourceRange(), "predefined lvalue in block"); + + return emitStringLiteralLValue(sl, gvName); +} + /// An LValue is a candidate for having its loads and stores be made atomic if /// we are operating under /volatile:ms *and* the LValue itself is volatile and /// performing such an operation can be performed without a libcall. diff --git a/clang/lib/CIR/CodeGen/CIRGenFunction.cpp b/clang/lib/CIR/CodeGen/CIRGenFunction.cpp index b26b4f2..52fb0d7 100644 --- a/clang/lib/CIR/CodeGen/CIRGenFunction.cpp +++ b/clang/lib/CIR/CodeGen/CIRGenFunction.cpp @@ -342,6 +342,9 @@ void CIRGenFunction::LexicalScope::cleanup() { cir::ReturnOp CIRGenFunction::LexicalScope::emitReturn(mlir::Location loc) { CIRGenBuilderTy &builder = cgf.getBuilder(); + // If we are on a coroutine, add the coro_end builtin call. + assert(!cir::MissingFeatures::coroEndBuiltinCall()); + auto fn = dyn_cast<cir::FuncOp>(cgf.curFn); assert(fn && "emitReturn from non-function"); if (!fn.getFunctionType().hasVoidReturn()) { @@ -815,6 +818,8 @@ LValue CIRGenFunction::emitLValue(const Expr *e) { return emitMemberExpr(cast<MemberExpr>(e)); case Expr::CompoundLiteralExprClass: return emitCompoundLiteralLValue(cast<CompoundLiteralExpr>(e)); + case Expr::PredefinedExprClass: + return emitPredefinedLValue(cast<PredefinedExpr>(e)); case Expr::BinaryOperatorClass: return emitBinaryOperatorLValue(cast<BinaryOperator>(e)); case Expr::CompoundAssignOperatorClass: { diff --git a/clang/lib/CIR/CodeGen/CIRGenFunction.h b/clang/lib/CIR/CodeGen/CIRGenFunction.h index cb7cf98..dfd9d2c 100644 --- a/clang/lib/CIR/CodeGen/CIRGenFunction.h +++ b/clang/lib/CIR/CodeGen/CIRGenFunction.h @@ -47,6 +47,8 @@ class LoopOp; namespace clang::CIRGen { +struct CGCoroData; + class CIRGenFunction : public CIRGenTypeCache { public: CIRGenModule &cgm; @@ -66,6 +68,18 @@ public: /// The compiler-generated variable that holds the return value. std::optional<mlir::Value> fnRetAlloca; + // Holds coroutine data if the current function is a coroutine. We use a + // wrapper to manage its lifetime, so that we don't have to define CGCoroData + // in this header. + struct CGCoroInfo { + std::unique_ptr<CGCoroData> data; + CGCoroInfo(); + ~CGCoroInfo(); + }; + CGCoroInfo curCoro; + + bool isCoroutine() const { return curCoro.data != nullptr; } + /// The temporary alloca to hold the return value. This is /// invalid iff the function has no return value. Address returnValue = Address::invalid(); @@ -1174,6 +1188,10 @@ public: void emitConstructorBody(FunctionArgList &args); + mlir::LogicalResult emitCoroutineBody(const CoroutineBodyStmt &s); + cir::CallOp emitCoroEndBuiltinCall(mlir::Location loc, mlir::Value nullPtr); + cir::CallOp emitCoroIDBuiltinCall(mlir::Location loc, mlir::Value nullPtr); + void emitDestroy(Address addr, QualType type, Destroyer *destroyer); void emitDestructorBody(FunctionArgList &args); @@ -1279,6 +1297,8 @@ public: void emitInitializerForField(clang::FieldDecl *field, LValue lhs, clang::Expr *init); + LValue emitPredefinedLValue(const PredefinedExpr *e); + mlir::Value emitPromotedComplexExpr(const Expr *e, QualType promotionType); mlir::Value emitPromotedScalarExpr(const Expr *e, QualType promotionType); @@ -1473,7 +1493,8 @@ public: mlir::Value emitStoreThroughBitfieldLValue(RValue src, LValue dstresult); - LValue emitStringLiteralLValue(const StringLiteral *e); + LValue emitStringLiteralLValue(const StringLiteral *e, + llvm::StringRef name = ".str"); mlir::LogicalResult emitSwitchBody(const clang::Stmt *s); mlir::LogicalResult emitSwitchCase(const clang::SwitchCase &s, diff --git a/clang/lib/CIR/CodeGen/CIRGenModule.cpp b/clang/lib/CIR/CodeGen/CIRGenModule.cpp index 2bd2729..910c8a9 100644 --- a/clang/lib/CIR/CodeGen/CIRGenModule.cpp +++ b/clang/lib/CIR/CodeGen/CIRGenModule.cpp @@ -76,6 +76,7 @@ CIRGenModule::CIRGenModule(mlir::MLIRContext &mlirContext, SInt128Ty = cir::IntType::get(&getMLIRContext(), 128, /*isSigned=*/true); UInt8Ty = cir::IntType::get(&getMLIRContext(), 8, /*isSigned=*/false); UInt8PtrTy = cir::PointerType::get(UInt8Ty); + cirAllocaAddressSpace = getTargetCIRGenInfo().getCIRAllocaAddressSpace(); UInt16Ty = cir::IntType::get(&getMLIRContext(), 16, /*isSigned=*/false); UInt32Ty = cir::IntType::get(&getMLIRContext(), 32, /*isSigned=*/false); UInt64Ty = cir::IntType::get(&getMLIRContext(), 64, /*isSigned=*/false); @@ -1343,32 +1344,36 @@ cir::GlobalOp CIRGenModule::getGlobalForStringLiteral(const StringLiteral *s, mlir::Attribute c = getConstantArrayFromStringLiteral(s); - if (getLangOpts().WritableStrings) { - errorNYI(s->getSourceRange(), - "getGlobalForStringLiteral: Writable strings"); - } - - // Mangle the string literal if that's how the ABI merges duplicate strings. - // Don't do it if they are writable, since we don't want writes in one TU to - // affect strings in another. - if (getCXXABI().getMangleContext().shouldMangleStringLiteral(s) && - !getLangOpts().WritableStrings) { - errorNYI(s->getSourceRange(), - "getGlobalForStringLiteral: mangle string literals"); - } - - // Unlike LLVM IR, CIR doesn't automatically unique names for globals, so - // we need to do that explicitly. - std::string uniqueName = getUniqueGlobalName(name.str()); - mlir::Location loc = getLoc(s->getSourceRange()); - auto typedC = llvm::cast<mlir::TypedAttr>(c); - cir::GlobalOp gv = - generateStringLiteral(loc, typedC, cir::GlobalLinkageKind::PrivateLinkage, - *this, uniqueName, alignment); - setDSOLocal(static_cast<mlir::Operation *>(gv)); + cir::GlobalOp gv; + if (!getLangOpts().WritableStrings && constantStringMap.count(c)) { + gv = constantStringMap[c]; + // The bigger alignment always wins. + if (!gv.getAlignment() || + uint64_t(alignment.getQuantity()) > *gv.getAlignment()) + gv.setAlignmentAttr(getSize(alignment)); + } else { + // Mangle the string literal if that's how the ABI merges duplicate strings. + // Don't do it if they are writable, since we don't want writes in one TU to + // affect strings in another. + if (getCXXABI().getMangleContext().shouldMangleStringLiteral(s) && + !getLangOpts().WritableStrings) { + errorNYI(s->getSourceRange(), + "getGlobalForStringLiteral: mangle string literals"); + } - assert(!cir::MissingFeatures::sanitizers()); + // Unlike LLVM IR, CIR doesn't automatically unique names for globals, so + // we need to do that explicitly. + std::string uniqueName = getUniqueGlobalName(name.str()); + mlir::Location loc = getLoc(s->getSourceRange()); + auto typedC = llvm::cast<mlir::TypedAttr>(c); + gv = generateStringLiteral(loc, typedC, + cir::GlobalLinkageKind::PrivateLinkage, *this, + uniqueName, alignment); + setDSOLocal(static_cast<mlir::Operation *>(gv)); + constantStringMap[c] = gv; + assert(!cir::MissingFeatures::sanitizers()); + } return gv; } @@ -2065,6 +2070,15 @@ CIRGenModule::createCIRFunction(mlir::Location loc, StringRef name, return func; } +cir::FuncOp +CIRGenModule::createCIRBuiltinFunction(mlir::Location loc, StringRef name, + cir::FuncType ty, + const clang::FunctionDecl *fd) { + cir::FuncOp fnOp = createCIRFunction(loc, name, ty, fd); + fnOp.setBuiltin(true); + return fnOp; +} + mlir::SymbolTable::Visibility CIRGenModule::getMLIRVisibility(cir::GlobalOp op) { // MLIR doesn't accept public symbols declarations (only diff --git a/clang/lib/CIR/CodeGen/CIRGenModule.h b/clang/lib/CIR/CodeGen/CIRGenModule.h index 2c4c6dd..c6a6681 100644 --- a/clang/lib/CIR/CodeGen/CIRGenModule.h +++ b/clang/lib/CIR/CodeGen/CIRGenModule.h @@ -274,6 +274,8 @@ public: llvm_unreachable("unknown visibility!"); } + llvm::DenseMap<mlir::Attribute, cir::GlobalOp> constantStringMap; + /// Return a constant array for the given string. mlir::Attribute getConstantArrayFromStringLiteral(const StringLiteral *e); @@ -473,6 +475,13 @@ public: cir::FuncType funcType, const clang::FunctionDecl *funcDecl); + /// Create a CIR function with builtin attribute set. + cir::FuncOp createCIRBuiltinFunction(mlir::Location loc, llvm::StringRef name, + cir::FuncType ty, + const clang::FunctionDecl *fd); + + static constexpr const char *builtinCoroId = "__builtin_coro_id"; + /// Given a builtin id for a function like "__builtin_fabsf", return a /// Function* for "fabsf". cir::FuncOp getBuiltinLibFunction(const FunctionDecl *fd, unsigned builtinID); diff --git a/clang/lib/CIR/CodeGen/CIRGenOpenACCRecipe.cpp b/clang/lib/CIR/CodeGen/CIRGenOpenACCRecipe.cpp index ea6ea2c..bbc45e5 100644 --- a/clang/lib/CIR/CodeGen/CIRGenOpenACCRecipe.cpp +++ b/clang/lib/CIR/CodeGen/CIRGenOpenACCRecipe.cpp @@ -427,22 +427,20 @@ void OpenACCRecipeBuilderBase::makeBoundsInit( cgf.emitAutoVarInit(tempDeclEmission); } -// TODO: OpenACC: When we get this implemented for the reduction/firstprivate, -// this might end up re-merging with createRecipeInitCopy. For now, keep it -// separate until we're sure what everything looks like to keep this as clean -// as possible. -void OpenACCRecipeBuilderBase::createPrivateInitRecipe( +// TODO: OpenACC: when we start doing firstprivate for array/vlas/etc, we +// probably need to do a little work about the 'init' calls to put it in 'copy' +// region instead. +void OpenACCRecipeBuilderBase::createInitRecipe( mlir::Location loc, mlir::Location locEnd, SourceRange exprRange, - mlir::Value mainOp, mlir::acc::PrivateRecipeOp recipe, size_t numBounds, + mlir::Value mainOp, mlir::Region &recipeInitRegion, size_t numBounds, llvm::ArrayRef<QualType> boundTypes, const VarDecl *allocaDecl, QualType origType) { assert(allocaDecl && "Required recipe variable not set?"); CIRGenFunction::DeclMapRevertingRAII declMapRAII{cgf, allocaDecl}; - mlir::Block *block = - createRecipeBlock(recipe.getInitRegion(), mainOp.getType(), loc, - numBounds, /*isInit=*/true); - builder.setInsertionPointToEnd(&recipe.getInitRegion().back()); + mlir::Block *block = createRecipeBlock(recipeInitRegion, mainOp.getType(), + loc, numBounds, /*isInit=*/true); + builder.setInsertionPointToEnd(&recipeInitRegion.back()); CIRGenFunction::LexicalScope ls(cgf, loc, block); const Type *allocaPointeeType = @@ -458,7 +456,7 @@ void OpenACCRecipeBuilderBase::createPrivateInitRecipe( // Sema::TentativeAnalysisScopes in SemaOpenACC::CreateInitRecipe, it'll // emit an error to tell us. However, emitting those errors during // production is a violation of the standard, so we cannot do them. - cgf.cgm.errorNYI(exprRange, "private default-init recipe"); + cgf.cgm.errorNYI(exprRange, "private/reduction default-init recipe"); } if (!numBounds) { @@ -469,7 +467,7 @@ void OpenACCRecipeBuilderBase::createPrivateInitRecipe( cgf.emitAutoVarInit(tempDeclEmission); } else { mlir::Value alloca = makeBoundsAlloca( - block, exprRange, loc, "openacc.private.init", numBounds, boundTypes); + block, exprRange, loc, allocaDecl->getName(), numBounds, boundTypes); // If the initializer is trivial, there is nothing to do here, so save // ourselves some effort. @@ -521,10 +519,10 @@ void OpenACCRecipeBuilderBase::createFirstprivateRecipeCopy( // doesn't restore it aftewards. void OpenACCRecipeBuilderBase::createReductionRecipeCombiner( mlir::Location loc, mlir::Location locEnd, mlir::Value mainOp, - mlir::acc::ReductionRecipeOp recipe) { - mlir::Block *block = builder.createBlock( - &recipe.getCombinerRegion(), recipe.getCombinerRegion().end(), - {mainOp.getType(), mainOp.getType()}, {loc, loc}); + mlir::acc::ReductionRecipeOp recipe, size_t numBounds) { + mlir::Block *block = + createRecipeBlock(recipe.getCombinerRegion(), mainOp.getType(), loc, + numBounds, /*isInit=*/false); builder.setInsertionPointToEnd(&recipe.getCombinerRegion().back()); CIRGenFunction::LexicalScope ls(cgf, loc, block); diff --git a/clang/lib/CIR/CodeGen/CIRGenOpenACCRecipe.h b/clang/lib/CIR/CodeGen/CIRGenOpenACCRecipe.h index a05b0bd..21707ad 100644 --- a/clang/lib/CIR/CodeGen/CIRGenOpenACCRecipe.h +++ b/clang/lib/CIR/CodeGen/CIRGenOpenACCRecipe.h @@ -64,13 +64,13 @@ protected: // doesn't restore it aftewards. void createReductionRecipeCombiner(mlir::Location loc, mlir::Location locEnd, mlir::Value mainOp, - mlir::acc::ReductionRecipeOp recipe); - void createPrivateInitRecipe(mlir::Location loc, mlir::Location locEnd, - SourceRange exprRange, mlir::Value mainOp, - mlir::acc::PrivateRecipeOp recipe, - size_t numBounds, - llvm::ArrayRef<QualType> boundTypes, - const VarDecl *allocaDecl, QualType origType); + mlir::acc::ReductionRecipeOp recipe, + size_t numBounds); + void createInitRecipe(mlir::Location loc, mlir::Location locEnd, + SourceRange exprRange, mlir::Value mainOp, + mlir::Region &recipeInitRegion, size_t numBounds, + llvm::ArrayRef<QualType> boundTypes, + const VarDecl *allocaDecl, QualType origType); void createRecipeDestroySection(mlir::Location loc, mlir::Location locEnd, mlir::Value mainOp, CharUnits alignment, @@ -224,10 +224,10 @@ public: // TODO: OpenACC: This is a bit of a hackery to get this to not change for // the non-private recipes. This will be removed soon, when we get this // 'right' for firstprivate and reduction. - if constexpr (!std::is_same_v<RecipeTy, mlir::acc::PrivateRecipeOp>) { + if constexpr (std::is_same_v<RecipeTy, mlir::acc::FirstprivateRecipeOp>) { if (numBounds) { cgf.cgm.errorNYI(varRef->getSourceRange(), - "firstprivate/reduction-init with bounds"); + "firstprivate-init with bounds"); } boundTypes = {}; numBounds = 0; @@ -260,18 +260,25 @@ public: insertLocation = modBuilder.saveInsertionPoint(); if constexpr (std::is_same_v<RecipeTy, mlir::acc::PrivateRecipeOp>) { - createPrivateInitRecipe(loc, locEnd, varRef->getSourceRange(), mainOp, - recipe, numBounds, boundTypes, varRecipe, - origType); + createInitRecipe(loc, locEnd, varRef->getSourceRange(), mainOp, + recipe.getInitRegion(), numBounds, boundTypes, varRecipe, + origType); + } else if constexpr (std::is_same_v<RecipeTy, + mlir::acc::ReductionRecipeOp>) { + createInitRecipe(loc, locEnd, varRef->getSourceRange(), mainOp, + recipe.getInitRegion(), numBounds, boundTypes, varRecipe, + origType); + createReductionRecipeCombiner(loc, locEnd, mainOp, recipe, numBounds); } else { + static_assert(std::is_same_v<RecipeTy, mlir::acc::FirstprivateRecipeOp>); + // TODO: OpenACC: we probably want this to call createInitRecipe as well, + // but do so in a way that omits the 'initialization', so that we can do + // it separately, since it belongs in the 'copy' region. It also might + // need a way of getting the tempDeclEmission out of it for that purpose. createRecipeInitCopy(loc, locEnd, varRef->getSourceRange(), mainOp, recipe, varRecipe, temporary); } - if constexpr (std::is_same_v<RecipeTy, mlir::acc::ReductionRecipeOp>) { - createReductionRecipeCombiner(loc, locEnd, mainOp, recipe); - } - if (origType.isDestructedType()) createRecipeDestroySection( loc, locEnd, mainOp, cgf.getContext().getDeclAlign(varRecipe), diff --git a/clang/lib/CIR/CodeGen/CIRGenStmt.cpp b/clang/lib/CIR/CodeGen/CIRGenStmt.cpp index 644c383..0b8f8bf 100644 --- a/clang/lib/CIR/CodeGen/CIRGenStmt.cpp +++ b/clang/lib/CIR/CodeGen/CIRGenStmt.cpp @@ -197,6 +197,7 @@ mlir::LogicalResult CIRGenFunction::emitStmt(const Stmt *s, case Stmt::SEHLeaveStmtClass: case Stmt::SYCLKernelCallStmtClass: case Stmt::CoroutineBodyStmtClass: + return emitCoroutineBody(cast<CoroutineBodyStmt>(*s)); case Stmt::CoreturnStmtClass: case Stmt::CXXTryStmtClass: case Stmt::IndirectGotoStmtClass: diff --git a/clang/lib/CIR/CodeGen/CIRGenTypeCache.h b/clang/lib/CIR/CodeGen/CIRGenTypeCache.h index cc3ce09..273ec7f 100644 --- a/clang/lib/CIR/CodeGen/CIRGenTypeCache.h +++ b/clang/lib/CIR/CodeGen/CIRGenTypeCache.h @@ -14,6 +14,7 @@ #define LLVM_CLANG_LIB_CIR_CIRGENTYPECACHE_H #include "clang/AST/CharUnits.h" +#include "clang/Basic/AddressSpaces.h" #include "clang/CIR/Dialect/IR/CIRTypes.h" namespace clang::CIRGen { @@ -73,6 +74,8 @@ struct CIRGenTypeCache { /// The alignment of size_t. unsigned char SizeAlignInBytes; + cir::TargetAddressSpaceAttr cirAllocaAddressSpace; + clang::CharUnits getSizeAlign() const { return clang::CharUnits::fromQuantity(SizeAlignInBytes); } @@ -80,6 +83,10 @@ struct CIRGenTypeCache { clang::CharUnits getPointerAlign() const { return clang::CharUnits::fromQuantity(PointerAlignInBytes); } + + cir::TargetAddressSpaceAttr getCIRAllocaAddressSpace() const { + return cirAllocaAddressSpace; + } }; } // namespace clang::CIRGen diff --git a/clang/lib/CIR/CodeGen/CIRGenTypes.cpp b/clang/lib/CIR/CodeGen/CIRGenTypes.cpp index bb24933..e65896a 100644 --- a/clang/lib/CIR/CodeGen/CIRGenTypes.cpp +++ b/clang/lib/CIR/CodeGen/CIRGenTypes.cpp @@ -417,7 +417,7 @@ mlir::Type CIRGenTypes::convertType(QualType type) { mlir::Type pointeeType = convertType(elemTy); - resultType = builder.getPointerTo(pointeeType); + resultType = builder.getPointerTo(pointeeType, elemTy.getAddressSpace()); break; } diff --git a/clang/lib/CIR/CodeGen/CMakeLists.txt b/clang/lib/CIR/CodeGen/CMakeLists.txt index 3ebf460..36db4bd 100644 --- a/clang/lib/CIR/CodeGen/CMakeLists.txt +++ b/clang/lib/CIR/CodeGen/CMakeLists.txt @@ -14,6 +14,7 @@ add_clang_library(clangCIR CIRGenCall.cpp CIRGenClass.cpp CIRGenCleanup.cpp + CIRGenCoroutine.cpp CIRGenCXX.cpp CIRGenCXXABI.cpp CIRGenBuiltin.cpp diff --git a/clang/lib/CIR/CodeGen/TargetInfo.h b/clang/lib/CIR/CodeGen/TargetInfo.h index a5c548a..dbb0312 100644 --- a/clang/lib/CIR/CodeGen/TargetInfo.h +++ b/clang/lib/CIR/CodeGen/TargetInfo.h @@ -16,6 +16,7 @@ #include "ABIInfo.h" #include "CIRGenTypes.h" +#include "clang/Basic/AddressSpaces.h" #include <memory> #include <utility> @@ -43,6 +44,11 @@ public: /// Returns ABI info helper for the target. const ABIInfo &getABIInfo() const { return *info; } + /// Get the address space for alloca. + virtual cir::TargetAddressSpaceAttr getCIRAllocaAddressSpace() const { + return {}; + } + /// Determine whether a call to an unprototyped functions under /// the given calling convention should use the variadic /// convention or the non-variadic convention. diff --git a/clang/lib/CIR/Dialect/IR/CIRAttrs.cpp b/clang/lib/CIR/Dialect/IR/CIRAttrs.cpp index 95faad6..3484c59 100644 --- a/clang/lib/CIR/Dialect/IR/CIRAttrs.cpp +++ b/clang/lib/CIR/Dialect/IR/CIRAttrs.cpp @@ -43,6 +43,16 @@ parseFloatLiteral(mlir::AsmParser &parser, mlir::FailureOr<llvm::APFloat> &value, cir::FPTypeInterface fpType); +//===----------------------------------------------------------------------===// +// AddressSpaceAttr +//===----------------------------------------------------------------------===// + +mlir::ParseResult parseTargetAddressSpace(mlir::AsmParser &p, + cir::TargetAddressSpaceAttr &attr); + +void printTargetAddressSpace(mlir::AsmPrinter &p, + cir::TargetAddressSpaceAttr attr); + static mlir::ParseResult parseConstPtr(mlir::AsmParser &parser, mlir::IntegerAttr &value); diff --git a/clang/lib/CIR/Dialect/IR/CIRDialect.cpp b/clang/lib/CIR/Dialect/IR/CIRDialect.cpp index 6b5cc80..fba094f 100644 --- a/clang/lib/CIR/Dialect/IR/CIRDialect.cpp +++ b/clang/lib/CIR/Dialect/IR/CIRDialect.cpp @@ -1632,12 +1632,19 @@ ParseResult cir::FuncOp::parse(OpAsmParser &parser, OperationState &state) { llvm::SMLoc loc = parser.getCurrentLocation(); mlir::Builder &builder = parser.getBuilder(); + mlir::StringAttr builtinNameAttr = getBuiltinAttrName(state.name); + mlir::StringAttr coroutineNameAttr = getCoroutineAttrName(state.name); mlir::StringAttr lambdaNameAttr = getLambdaAttrName(state.name); mlir::StringAttr noProtoNameAttr = getNoProtoAttrName(state.name); mlir::StringAttr visNameAttr = getSymVisibilityAttrName(state.name); mlir::StringAttr visibilityNameAttr = getGlobalVisibilityAttrName(state.name); mlir::StringAttr dsoLocalNameAttr = getDsoLocalAttrName(state.name); + if (::mlir::succeeded(parser.parseOptionalKeyword(builtinNameAttr.strref()))) + state.addAttribute(builtinNameAttr, parser.getBuilder().getUnitAttr()); + if (::mlir::succeeded( + parser.parseOptionalKeyword(coroutineNameAttr.strref()))) + state.addAttribute(coroutineNameAttr, parser.getBuilder().getUnitAttr()); if (::mlir::succeeded(parser.parseOptionalKeyword(lambdaNameAttr.strref()))) state.addAttribute(lambdaNameAttr, parser.getBuilder().getUnitAttr()); if (parser.parseOptionalKeyword(noProtoNameAttr).succeeded()) @@ -1747,6 +1754,12 @@ mlir::Region *cir::FuncOp::getCallableRegion() { } void cir::FuncOp::print(OpAsmPrinter &p) { + if (getBuiltin()) + p << " builtin"; + + if (getCoroutine()) + p << " coroutine"; + if (getLambda()) p << " lambda"; diff --git a/clang/lib/CIR/Dialect/IR/CIRTypes.cpp b/clang/lib/CIR/Dialect/IR/CIRTypes.cpp index 35b4513..58973528 100644 --- a/clang/lib/CIR/Dialect/IR/CIRTypes.cpp +++ b/clang/lib/CIR/Dialect/IR/CIRTypes.cpp @@ -13,6 +13,7 @@ #include "clang/CIR/Dialect/IR/CIRTypes.h" #include "mlir/IR/DialectImplementation.h" +#include "clang/CIR/Dialect/IR/CIRAttrs.h" #include "clang/CIR/Dialect/IR/CIRDialect.h" #include "clang/CIR/Dialect/IR/CIRTypesDetails.h" #include "clang/CIR/MissingFeatures.h" @@ -38,6 +39,27 @@ parseFuncTypeParams(mlir::AsmParser &p, llvm::SmallVector<mlir::Type> ¶ms, static void printFuncTypeParams(mlir::AsmPrinter &p, mlir::ArrayRef<mlir::Type> params, bool isVarArg); +//===----------------------------------------------------------------------===// +// CIR Custom Parser/Printer Signatures +//===----------------------------------------------------------------------===// + +static mlir::ParseResult +parseFuncTypeParams(mlir::AsmParser &p, llvm::SmallVector<mlir::Type> ¶ms, + bool &isVarArg); + +static void printFuncTypeParams(mlir::AsmPrinter &p, + mlir::ArrayRef<mlir::Type> params, + bool isVarArg); + +//===----------------------------------------------------------------------===// +// AddressSpace +//===----------------------------------------------------------------------===// + +mlir::ParseResult parseTargetAddressSpace(mlir::AsmParser &p, + cir::TargetAddressSpaceAttr &attr); + +void printTargetAddressSpace(mlir::AsmPrinter &p, + cir::TargetAddressSpaceAttr attr); //===----------------------------------------------------------------------===// // Get autogenerated stuff @@ -298,6 +320,22 @@ bool RecordType::isLayoutIdentical(const RecordType &other) { //===----------------------------------------------------------------------===// llvm::TypeSize +PointerType::getTypeSizeInBits(const ::mlir::DataLayout &dataLayout, + ::mlir::DataLayoutEntryListRef params) const { + // FIXME: improve this in face of address spaces + assert(!cir::MissingFeatures::dataLayoutPtrHandlingBasedOnLangAS()); + return llvm::TypeSize::getFixed(64); +} + +uint64_t +PointerType::getABIAlignment(const ::mlir::DataLayout &dataLayout, + ::mlir::DataLayoutEntryListRef params) const { + // FIXME: improve this in face of address spaces + assert(!cir::MissingFeatures::dataLayoutPtrHandlingBasedOnLangAS()); + return 8; +} + +llvm::TypeSize RecordType::getTypeSizeInBits(const mlir::DataLayout &dataLayout, mlir::DataLayoutEntryListRef params) const { if (isUnion()) @@ -766,30 +804,39 @@ mlir::LogicalResult cir::VectorType::verify( } //===----------------------------------------------------------------------===// -// PointerType Definitions +// TargetAddressSpace definitions //===----------------------------------------------------------------------===// -llvm::TypeSize -PointerType::getTypeSizeInBits(const ::mlir::DataLayout &dataLayout, - ::mlir::DataLayoutEntryListRef params) const { - // FIXME: improve this in face of address spaces - return llvm::TypeSize::getFixed(64); -} +mlir::ParseResult parseTargetAddressSpace(mlir::AsmParser &p, + cir::TargetAddressSpaceAttr &attr) { + if (failed(p.parseKeyword("target_address_space"))) + return mlir::failure(); -uint64_t -PointerType::getABIAlignment(const ::mlir::DataLayout &dataLayout, - ::mlir::DataLayoutEntryListRef params) const { - // FIXME: improve this in face of address spaces - return 8; -} + if (failed(p.parseLParen())) + return mlir::failure(); -mlir::LogicalResult -PointerType::verify(llvm::function_ref<mlir::InFlightDiagnostic()> emitError, - mlir::Type pointee) { - // TODO(CIR): Verification of the address space goes here. + int32_t targetValue; + if (failed(p.parseInteger(targetValue))) + return p.emitError(p.getCurrentLocation(), + "expected integer address space value"); + + if (failed(p.parseRParen())) + return p.emitError(p.getCurrentLocation(), + "expected ')' after address space value"); + + mlir::MLIRContext *context = p.getBuilder().getContext(); + attr = cir::TargetAddressSpaceAttr::get( + context, p.getBuilder().getUI32IntegerAttr(targetValue)); return mlir::success(); } +// The custom printer for the `addrspace` parameter in `!cir.ptr`. +// in the format of `target_address_space(N)`. +void printTargetAddressSpace(mlir::AsmPrinter &p, + cir::TargetAddressSpaceAttr attr) { + p << "target_address_space(" << attr.getValue().getUInt() << ")"; +} + //===----------------------------------------------------------------------===// // CIR Dialect //===----------------------------------------------------------------------===// diff --git a/clang/lib/CIR/Lowering/DirectToLLVM/LowerToLLVM.cpp b/clang/lib/CIR/Lowering/DirectToLLVM/LowerToLLVM.cpp index 4bc7783..bfb1262 100644 --- a/clang/lib/CIR/Lowering/DirectToLLVM/LowerToLLVM.cpp +++ b/clang/lib/CIR/Lowering/DirectToLLVM/LowerToLLVM.cpp @@ -32,6 +32,7 @@ #include "mlir/Transforms/DialectConversion.h" #include "clang/CIR/Dialect/IR/CIRAttrs.h" #include "clang/CIR/Dialect/IR/CIRDialect.h" +#include "clang/CIR/Dialect/IR/CIRTypes.h" #include "clang/CIR/Dialect/Passes.h" #include "clang/CIR/LoweringHelpers.h" #include "clang/CIR/MissingFeatures.h" @@ -2308,11 +2309,9 @@ mlir::LogicalResult CIRToLLVMSelectOpLowering::matchAndRewrite( static void prepareTypeConverter(mlir::LLVMTypeConverter &converter, mlir::DataLayout &dataLayout) { converter.addConversion([&](cir::PointerType type) -> mlir::Type { - // Drop pointee type since LLVM dialect only allows opaque pointers. - assert(!cir::MissingFeatures::addressSpace()); - unsigned targetAS = 0; - - return mlir::LLVM::LLVMPointerType::get(type.getContext(), targetAS); + unsigned addrSpace = + type.getAddrSpace() ? type.getAddrSpace().getValue().getUInt() : 0; + return mlir::LLVM::LLVMPointerType::get(type.getContext(), addrSpace); }); converter.addConversion([&](cir::VPtrType type) -> mlir::Type { assert(!cir::MissingFeatures::addressSpace()); diff --git a/clang/lib/CrossTU/CrossTranslationUnit.cpp b/clang/lib/CrossTU/CrossTranslationUnit.cpp index 847913d..0287845 100644 --- a/clang/lib/CrossTU/CrossTranslationUnit.cpp +++ b/clang/lib/CrossTU/CrossTranslationUnit.cpp @@ -252,9 +252,9 @@ CrossTranslationUnitContext::CrossTranslationUnitContext(CompilerInstance &CI) CrossTranslationUnitContext::~CrossTranslationUnitContext() {} std::optional<std::string> -CrossTranslationUnitContext::getLookupName(const NamedDecl *ND) { +CrossTranslationUnitContext::getLookupName(const Decl *D) { SmallString<128> DeclUSR; - bool Ret = index::generateUSRForDecl(ND, DeclUSR); + bool Ret = index::generateUSRForDecl(D, DeclUSR); if (Ret) return {}; return std::string(DeclUSR); diff --git a/clang/lib/Driver/ToolChains/Clang.cpp b/clang/lib/Driver/ToolChains/Clang.cpp index 412a176..684cc09 100644 --- a/clang/lib/Driver/ToolChains/Clang.cpp +++ b/clang/lib/Driver/ToolChains/Clang.cpp @@ -9224,14 +9224,20 @@ void LinkerWrapper::ConstructJob(Compilation &C, const JobAction &JA, addOffloadCompressArgs(Args, CmdArgs); if (Arg *A = Args.getLastArg(options::OPT_offload_jobs_EQ)) { - int NumThreads; - if (StringRef(A->getValue()).getAsInteger(10, NumThreads) || - NumThreads <= 0) - C.getDriver().Diag(diag::err_drv_invalid_int_value) - << A->getAsString(Args) << A->getValue(); - else - CmdArgs.push_back( - Args.MakeArgString("--wrapper-jobs=" + Twine(NumThreads))); + StringRef Val = A->getValue(); + + if (Val.equals_insensitive("jobserver")) + CmdArgs.push_back(Args.MakeArgString("--wrapper-jobs=jobserver")); + else { + int NumThreads; + if (Val.getAsInteger(10, NumThreads) || NumThreads <= 0) { + C.getDriver().Diag(diag::err_drv_invalid_int_value) + << A->getAsString(Args) << Val; + } else { + CmdArgs.push_back( + Args.MakeArgString("--wrapper-jobs=" + Twine(NumThreads))); + } + } } const char *Exec = diff --git a/clang/lib/Format/Format.cpp b/clang/lib/Format/Format.cpp index 2bf6244..686e541 100644 --- a/clang/lib/Format/Format.cpp +++ b/clang/lib/Format/Format.cpp @@ -3199,7 +3199,7 @@ private: Keywords.kw_NS_OPTIONS, TT_ObjCBlockLBrace, TT_ObjCBlockLParen, TT_ObjCDecl, TT_ObjCForIn, TT_ObjCMethodExpr, TT_ObjCMethodSpecifier, - TT_ObjCProperty)) { + TT_ObjCProperty, TT_ObjCSelector)) { LLVM_DEBUG(llvm::dbgs() << "Detected ObjC at location " << FormatTok->Tok.getLocation().printToString( diff --git a/clang/lib/Format/FormatToken.h b/clang/lib/Format/FormatToken.h index e4ddd61..f015d27 100644 --- a/clang/lib/Format/FormatToken.h +++ b/clang/lib/Format/FormatToken.h @@ -127,9 +127,17 @@ namespace format { TYPE(ObjCBlockLParen) \ TYPE(ObjCDecl) \ TYPE(ObjCForIn) \ + /* The square brackets surrounding a method call, the colon separating the \ + * method or parameter name and the argument inside the square brackets, and \ + * the colon separating the method or parameter name and the type inside the \ + * method declaration. */ \ TYPE(ObjCMethodExpr) \ + /* The '+' or '-' at the start of the line. */ \ TYPE(ObjCMethodSpecifier) \ TYPE(ObjCProperty) \ + /* The parentheses following '@selector' and the colon following the method \ + * or parameter name inside the parentheses. */ \ + TYPE(ObjCSelector) \ TYPE(ObjCStringLiteral) \ TYPE(OverloadedOperator) \ TYPE(OverloadedOperatorLParen) \ @@ -146,6 +154,9 @@ namespace format { TYPE(RequiresExpression) \ TYPE(RequiresExpressionLBrace) \ TYPE(RequiresExpressionLParen) \ + /* The hash key in languages that have hash literals, not including the \ + * field name in the C++ struct literal. Also the method or parameter name \ + * in the Objective-C method declaration or call. */ \ TYPE(SelectorName) \ TYPE(StartOfName) \ TYPE(StatementAttributeLikeMacro) \ diff --git a/clang/lib/Format/TokenAnnotator.cpp b/clang/lib/Format/TokenAnnotator.cpp index 59f81b3..5b784ed 100644 --- a/clang/lib/Format/TokenAnnotator.cpp +++ b/clang/lib/Format/TokenAnnotator.cpp @@ -321,13 +321,13 @@ private: return parseUntouchableParens(); } - bool StartsObjCMethodExpr = false; + bool StartsObjCSelector = false; if (!Style.isVerilog()) { if (FormatToken *MaybeSel = OpeningParen.Previous) { // @selector( starts a selector. if (MaybeSel->is(tok::objc_selector) && MaybeSel->Previous && MaybeSel->Previous->is(tok::at)) { - StartsObjCMethodExpr = true; + StartsObjCSelector = true; } } } @@ -451,10 +451,8 @@ private: } } - if (StartsObjCMethodExpr) { - Contexts.back().ColonIsObjCMethodExpr = true; - OpeningParen.setType(TT_ObjCMethodExpr); - } + if (StartsObjCSelector) + OpeningParen.setType(TT_ObjCSelector); // MightBeFunctionType and ProbablyFunctionType are used for // function pointer and reference types as well as Objective-C @@ -513,8 +511,8 @@ private: } } - if (StartsObjCMethodExpr) { - CurrentToken->setType(TT_ObjCMethodExpr); + if (StartsObjCSelector) { + CurrentToken->setType(TT_ObjCSelector); if (Contexts.back().FirstObjCSelectorName) { Contexts.back().FirstObjCSelectorName->LongestObjCSelectorName = Contexts.back().LongestObjCSelectorName; @@ -1449,7 +1447,7 @@ private: Next->Next->is(tok::colon)))) { // This handles a special macro in ObjC code where selectors including // the colon are passed as macro arguments. - Tok->setType(TT_ObjCMethodExpr); + Tok->setType(TT_ObjCSelector); } break; case tok::pipe: @@ -4608,7 +4606,7 @@ bool TokenAnnotator::spaceRequiredBetween(const AnnotatedLine &Line, return false; } if (Left.is(tok::colon)) - return Left.isNot(TT_ObjCMethodExpr); + return Left.isNoneOf(TT_ObjCSelector, TT_ObjCMethodExpr); if (Left.is(tok::coloncolon)) return false; if (Left.is(tok::less) || Right.isOneOf(tok::greater, tok::less)) { @@ -5464,7 +5462,7 @@ bool TokenAnnotator::spaceRequiredBefore(const AnnotatedLine &Line, // `private:` and `public:`. if (!Right.getNextNonComment()) return false; - if (Right.is(TT_ObjCMethodExpr)) + if (Right.isOneOf(TT_ObjCSelector, TT_ObjCMethodExpr)) return false; if (Left.is(tok::question)) return false; @@ -6288,6 +6286,7 @@ bool TokenAnnotator::canBreakBefore(const AnnotatedLine &Line, return Style.BreakInheritanceList == FormatStyle::BILS_AfterColon; if (Right.is(TT_InheritanceColon)) return Style.BreakInheritanceList != FormatStyle::BILS_AfterColon; + // When the method parameter has no name, allow breaking before the colon. if (Right.is(TT_ObjCMethodExpr) && Right.isNot(tok::r_square) && Left.isNot(TT_SelectorName)) { return true; diff --git a/clang/lib/Headers/avx10_2bf16intrin.h b/clang/lib/Headers/avx10_2bf16intrin.h index 5bcec4b..765cd68 100644 --- a/clang/lib/Headers/avx10_2bf16intrin.h +++ b/clang/lib/Headers/avx10_2bf16intrin.h @@ -519,34 +519,34 @@ _mm_maskz_min_pbh(__mmask8 __U, __m128bh __A, __m128bh __B) { (__mmask8)__U, (__v8bf)_mm_min_pbh(__A, __B), (__v8bf)_mm_setzero_pbh()); } -static __inline__ int __DEFAULT_FN_ATTRS128 _mm_comieq_sbh(__m128bh A, - __m128bh B) { - return __builtin_ia32_vcomisbf16eq((__v8bf)A, (__v8bf)B); +static __inline__ int __DEFAULT_FN_ATTRS128 _mm_comieq_sbh(__m128bh __A, + __m128bh __B) { + return __builtin_ia32_vcomisbf16eq((__v8bf)__A, (__v8bf)__B); } -static __inline__ int __DEFAULT_FN_ATTRS128 _mm_comilt_sbh(__m128bh A, - __m128bh B) { - return __builtin_ia32_vcomisbf16lt((__v8bf)A, (__v8bf)B); +static __inline__ int __DEFAULT_FN_ATTRS128 _mm_comilt_sbh(__m128bh __A, + __m128bh __B) { + return __builtin_ia32_vcomisbf16lt((__v8bf)__A, (__v8bf)__B); } -static __inline__ int __DEFAULT_FN_ATTRS128 _mm_comile_sbh(__m128bh A, - __m128bh B) { - return __builtin_ia32_vcomisbf16le((__v8bf)A, (__v8bf)B); +static __inline__ int __DEFAULT_FN_ATTRS128 _mm_comile_sbh(__m128bh __A, + __m128bh __B) { + return __builtin_ia32_vcomisbf16le((__v8bf)__A, (__v8bf)__B); } -static __inline__ int __DEFAULT_FN_ATTRS128 _mm_comigt_sbh(__m128bh A, - __m128bh B) { - return __builtin_ia32_vcomisbf16gt((__v8bf)A, (__v8bf)B); +static __inline__ int __DEFAULT_FN_ATTRS128 _mm_comigt_sbh(__m128bh __A, + __m128bh __B) { + return __builtin_ia32_vcomisbf16gt((__v8bf)__A, (__v8bf)__B); } -static __inline__ int __DEFAULT_FN_ATTRS128 _mm_comige_sbh(__m128bh A, - __m128bh B) { - return __builtin_ia32_vcomisbf16ge((__v8bf)A, (__v8bf)B); +static __inline__ int __DEFAULT_FN_ATTRS128 _mm_comige_sbh(__m128bh __A, + __m128bh __B) { + return __builtin_ia32_vcomisbf16ge((__v8bf)__A, (__v8bf)__B); } -static __inline__ int __DEFAULT_FN_ATTRS128 _mm_comineq_sbh(__m128bh A, - __m128bh B) { - return __builtin_ia32_vcomisbf16neq((__v8bf)A, (__v8bf)B); +static __inline__ int __DEFAULT_FN_ATTRS128 _mm_comineq_sbh(__m128bh __A, + __m128bh __B) { + return __builtin_ia32_vcomisbf16neq((__v8bf)__A, (__v8bf)__B); } #define _mm256_cmp_pbh_mask(__A, __B, __P) \ diff --git a/clang/lib/Sema/SemaOpenACC.cpp b/clang/lib/Sema/SemaOpenACC.cpp index 9aaf7f4..7ad7049 100644 --- a/clang/lib/Sema/SemaOpenACC.cpp +++ b/clang/lib/Sema/SemaOpenACC.cpp @@ -2894,17 +2894,18 @@ SemaOpenACC::CreateFirstPrivateInitRecipe(const Expr *VarExpr) { OpenACCReductionRecipe SemaOpenACC::CreateReductionInitRecipe( OpenACCReductionOperator ReductionOperator, const Expr *VarExpr) { - // TODO: OpenACC: This shouldn't be necessary, see PrivateInitRecipe - VarExpr = StripOffBounds(VarExpr); - + // We don't strip bounds here, so that we are doing our recipe init at the + // 'lowest' possible level. Codegen is going to have to do its own 'looping'. if (!VarExpr || VarExpr->getType()->isDependentType()) return OpenACCReductionRecipe::Empty(); QualType VarTy = VarExpr->getType().getNonReferenceType().getUnqualifiedType(); - // TODO: OpenACC: for arrays/bounds versions, we're going to have to do a - // different initializer, but for now we can go ahead with this. + // Array sections are special, and we have to treat them that way. + if (const auto *ASE = + dyn_cast<ArraySectionExpr>(VarExpr->IgnoreParenImpCasts())) + VarTy = ArraySectionExpr::getBaseOriginalType(ASE); VarDecl *AllocaDecl = CreateAllocaDecl( getASTContext(), SemaRef.getCurContext(), VarExpr->getBeginLoc(), diff --git a/clang/lib/Sema/SemaTemplateDeductionGuide.cpp b/clang/lib/Sema/SemaTemplateDeductionGuide.cpp index 9a61888..8ba23aa 100644 --- a/clang/lib/Sema/SemaTemplateDeductionGuide.cpp +++ b/clang/lib/Sema/SemaTemplateDeductionGuide.cpp @@ -1189,14 +1189,13 @@ BuildDeductionGuideForTypeAlias(Sema &SemaRef, // parameter list of a synthesized CTAD guide. See also the FIXME in // test/SemaCXX/cxx20-ctad-type-alias.cpp:test25. Sema::CheckTemplateArgumentInfo CTAI; + for (auto TA : Output.arguments()) + if (SemaRef.CheckTemplateArgument( + TP, TA, F, F->getLocation(), F->getLocation(), + /*ArgumentPackIndex=*/-1, CTAI, + Sema::CheckTemplateArgumentKind::CTAK_Specified)) + return nullptr; if (Input.getArgument().getKind() == TemplateArgument::Pack) { - for (auto TA : Output.arguments()) { - if (SemaRef.CheckTemplateArgument( - TP, TA, F, F->getLocation(), F->getLocation(), - /*ArgumentPackIndex=*/-1, CTAI, - Sema::CheckTemplateArgumentKind::CTAK_Specified)) - return nullptr; - } // We will substitute the non-deduced template arguments with these // transformed (unpacked at this point) arguments, where that substitution // requires a pack for the corresponding parameter packs. @@ -1204,12 +1203,6 @@ BuildDeductionGuideForTypeAlias(Sema &SemaRef, TemplateArgument::CreatePackCopy(Context, CTAI.SugaredConverted); } else { assert(Output.arguments().size() == 1); - TemplateArgumentLoc Transformed = Output.arguments()[0]; - if (SemaRef.CheckTemplateArgument( - TP, Transformed, F, F->getLocation(), F->getLocation(), - /*ArgumentPackIndex=*/-1, CTAI, - Sema::CheckTemplateArgumentKind::CTAK_Specified)) - return nullptr; TemplateArgsForBuildingFPrime[Index] = CTAI.SugaredConverted[0]; } } diff --git a/clang/lib/StaticAnalyzer/Core/CMakeLists.txt b/clang/lib/StaticAnalyzer/Core/CMakeLists.txt index d0a9b20..b8095a5 100644 --- a/clang/lib/StaticAnalyzer/Core/CMakeLists.txt +++ b/clang/lib/StaticAnalyzer/Core/CMakeLists.txt @@ -61,6 +61,7 @@ add_clang_library(clangStaticAnalyzerCore clangBasic clangCrossTU clangFrontend + clangIndex clangLex clangRewrite clangToolingCore diff --git a/clang/lib/StaticAnalyzer/Core/EntryPointStats.cpp b/clang/lib/StaticAnalyzer/Core/EntryPointStats.cpp index b7f9044..62ae62f2f 100644 --- a/clang/lib/StaticAnalyzer/Core/EntryPointStats.cpp +++ b/clang/lib/StaticAnalyzer/Core/EntryPointStats.cpp @@ -9,7 +9,9 @@ #include "clang/StaticAnalyzer/Core/PathSensitive/EntryPointStats.h" #include "clang/AST/DeclBase.h" #include "clang/Analysis/AnalysisDeclContext.h" +#include "clang/Index/USRGeneration.h" #include "llvm/ADT/STLExtras.h" +#include "llvm/ADT/SmallVector.h" #include "llvm/ADT/StringExtras.h" #include "llvm/ADT/StringRef.h" #include "llvm/Support/FileSystem.h" @@ -38,6 +40,7 @@ struct Registry { }; std::vector<Snapshot> Snapshots; + std::string EscapedCPPFileName; }; } // namespace @@ -69,7 +72,7 @@ static void checkStatName(const EntryPointStat *M) { } } -void EntryPointStat::lockRegistry() { +void EntryPointStat::lockRegistry(llvm::StringRef CPPFileName) { auto CmpByNames = [](const EntryPointStat *L, const EntryPointStat *R) { return L->name() < R->name(); }; @@ -78,6 +81,8 @@ void EntryPointStat::lockRegistry() { enumerateStatVectors( [](const auto &Stats) { llvm::for_each(Stats, checkStatName); }); StatsRegistry->IsLocked = true; + llvm::raw_string_ostream OS(StatsRegistry->EscapedCPPFileName); + llvm::printEscapedString(CPPFileName, OS); } [[maybe_unused]] static bool isRegistered(llvm::StringLiteral Name) { @@ -144,15 +149,27 @@ static std::vector<llvm::StringLiteral> getStatNames() { return Ret; } +static std::string getUSR(const Decl *D) { + llvm::SmallVector<char> Buf; + if (index::generateUSRForDecl(D, Buf)) { + assert(false && "This should never fail"); + return AnalysisDeclContext::getFunctionName(D); + } + return llvm::toStringRef(Buf).str(); +} + void Registry::Snapshot::dumpAsCSV(llvm::raw_ostream &OS) const { OS << '"'; + llvm::printEscapedString(getUSR(EntryPoint), OS); + OS << "\",\""; + OS << StatsRegistry->EscapedCPPFileName << "\",\""; llvm::printEscapedString( clang::AnalysisDeclContext::getFunctionName(EntryPoint), OS); - OS << "\", "; + OS << "\","; auto PrintAsBool = [&OS](bool B) { OS << (B ? "true" : "false"); }; - llvm::interleaveComma(BoolStatValues, OS, PrintAsBool); - OS << ((BoolStatValues.empty() || UnsignedStatValues.empty()) ? "" : ", "); - llvm::interleaveComma(UnsignedStatValues, OS); + llvm::interleave(BoolStatValues, OS, PrintAsBool, ","); + OS << ((BoolStatValues.empty() || UnsignedStatValues.empty()) ? "" : ","); + llvm::interleave(UnsignedStatValues, OS, [&OS](unsigned U) { OS << U; }, ","); } static std::vector<bool> consumeBoolStats() { @@ -181,8 +198,8 @@ void EntryPointStat::dumpStatsAsCSV(llvm::StringRef FileName) { } void EntryPointStat::dumpStatsAsCSV(llvm::raw_ostream &OS) { - OS << "EntryPoint, "; - llvm::interleaveComma(getStatNames(), OS); + OS << "USR,File,DebugName,"; + llvm::interleave(getStatNames(), OS, [&OS](const auto &a) { OS << a; }, ","); OS << "\n"; std::vector<std::string> Rows; diff --git a/clang/lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp b/clang/lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp index 53466e7..cf01e2f 100644 --- a/clang/lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp +++ b/clang/lib/StaticAnalyzer/Frontend/AnalysisConsumer.cpp @@ -51,6 +51,9 @@ STAT_COUNTER(NumFunctionTopLevel, "The # of functions at top level."); ALWAYS_ENABLED_STATISTIC(NumFunctionsAnalyzed, "The # of functions and blocks analyzed (as top level " "with inlining turned on)."); +ALWAYS_ENABLED_STATISTIC( + NumFunctionsAnalyzedSyntaxOnly, + "The # of functions analyzed by syntax checkers only."); ALWAYS_ENABLED_STATISTIC(NumBlocksInAnalyzedFunctions, "The # of basic blocks in the analyzed functions."); ALWAYS_ENABLED_STATISTIC( @@ -65,6 +68,15 @@ STAT_MAX(MaxCFGSize, "The maximum number of basic blocks in a function."); namespace { +StringRef getMainFileName(const CompilerInvocation &Invocation) { + if (!Invocation.getFrontendOpts().Inputs.empty()) { + const FrontendInputFile &Input = Invocation.getFrontendOpts().Inputs[0]; + return Input.isFile() ? Input.getFile() + : Input.getBuffer().getBufferIdentifier(); + } + return "<no input>"; +} + class AnalysisConsumer : public AnalysisASTConsumer, public DynamicRecursiveASTVisitor { enum { @@ -125,7 +137,8 @@ public: PP(CI.getPreprocessor()), OutDir(outdir), Opts(opts), Plugins(plugins), Injector(std::move(injector)), CTU(CI), MacroExpansions(CI.getLangOpts()) { - EntryPointStat::lockRegistry(); + + EntryPointStat::lockRegistry(getMainFileName(CI.getInvocation())); DigestAnalyzerOptions(); if (Opts.AnalyzerDisplayProgress || Opts.PrintStats || @@ -588,10 +601,10 @@ void AnalysisConsumer::runAnalysisOnTranslationUnit(ASTContext &C) { // If the user wanted to analyze a specific function and the number of basic // blocks analyzed is zero, than the user might not specified the function // name correctly. - // FIXME: The user might have analyzed the requested function in Syntax mode, - // but we are unaware of that. - if (!Opts.AnalyzeSpecificFunction.empty() && NumFunctionsAnalyzed == 0) + if (!Opts.AnalyzeSpecificFunction.empty() && NumFunctionsAnalyzed == 0 && + NumFunctionsAnalyzedSyntaxOnly == 0) { reportAnalyzerFunctionMisuse(Opts, *Ctx); + } } void AnalysisConsumer::reportAnalyzerProgress(StringRef S) { @@ -659,8 +672,11 @@ void AnalysisConsumer::HandleTranslationUnit(ASTContext &C) { AnalysisConsumer::AnalysisMode AnalysisConsumer::getModeForDecl(Decl *D, AnalysisMode Mode) { if (!Opts.AnalyzeSpecificFunction.empty() && - AnalysisDeclContext::getFunctionName(D) != Opts.AnalyzeSpecificFunction) + AnalysisDeclContext::getFunctionName(D) != Opts.AnalyzeSpecificFunction && + cross_tu::CrossTranslationUnitContext::getLookupName(D).value_or("") != + Opts.AnalyzeSpecificFunction) { return AM_None; + } // Unless -analyze-all is specified, treat decls differently depending on // where they came from: @@ -723,6 +739,7 @@ void AnalysisConsumer::HandleCode(Decl *D, AnalysisMode Mode, SyntaxCheckTimer->startTimer(); } checkerMgr->runCheckersOnASTBody(D, *Mgr, BR); + ++NumFunctionsAnalyzedSyntaxOnly; if (SyntaxCheckTimer) { SyntaxCheckTimer->stopTimer(); llvm::TimeRecord CheckerEndTime = SyntaxCheckTimer->getTotalTime(); diff --git a/clang/lib/Tooling/DependencyScanning/CMakeLists.txt b/clang/lib/Tooling/DependencyScanning/CMakeLists.txt index 53a2728..76bdc50 100644 --- a/clang/lib/Tooling/DependencyScanning/CMakeLists.txt +++ b/clang/lib/Tooling/DependencyScanning/CMakeLists.txt @@ -24,6 +24,5 @@ add_clang_library(clangDependencyScanning clangFrontend clangLex clangSerialization - clangTooling ${LLVM_PTHREAD_LIB} ) diff --git a/clang/lib/Tooling/DependencyScanning/DependencyScannerImpl.cpp b/clang/lib/Tooling/DependencyScanning/DependencyScannerImpl.cpp index 010380d..e1f4d0d 100644 --- a/clang/lib/Tooling/DependencyScanning/DependencyScannerImpl.cpp +++ b/clang/lib/Tooling/DependencyScanning/DependencyScannerImpl.cpp @@ -9,8 +9,10 @@ #include "DependencyScannerImpl.h" #include "clang/Basic/DiagnosticFrontend.h" #include "clang/Basic/DiagnosticSerialization.h" +#include "clang/Driver/Driver.h" #include "clang/Frontend/FrontendActions.h" #include "clang/Tooling/DependencyScanning/DependencyScanningWorker.h" +#include "llvm/TargetParser/Host.h" using namespace clang; using namespace tooling; @@ -332,11 +334,9 @@ public: return DepFS->getDirectiveTokens(File.getName()); } }; -} // namespace /// Sanitize diagnostic options for dependency scan. -void clang::tooling::dependencies::sanitizeDiagOpts( - DiagnosticOptions &DiagOpts) { +void sanitizeDiagOpts(DiagnosticOptions &DiagOpts) { // Don't print 'X warnings and Y errors generated'. DiagOpts.ShowCarets = false; // Don't write out diagnostic file. @@ -355,44 +355,146 @@ void clang::tooling::dependencies::sanitizeDiagOpts( .Default(true); }); } +} // namespace -bool DependencyScanningAction::runInvocation( - std::shared_ptr<CompilerInvocation> Invocation, - IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS, - std::shared_ptr<PCHContainerOperations> PCHContainerOps, - DiagnosticConsumer *DiagConsumer) { - // Making sure that we canonicalize the defines before we create the deep - // copy to avoid unnecessary variants in the scanner and in the resulting - // explicit command lines. - if (any(Service.getOptimizeArgs() & ScanningOptimizations::Macros)) - canonicalizeDefines(Invocation->getPreprocessorOpts()); +namespace clang::tooling::dependencies { +std::unique_ptr<DiagnosticOptions> +createDiagOptions(ArrayRef<std::string> CommandLine) { + std::vector<const char *> CLI; + for (const std::string &Arg : CommandLine) + CLI.push_back(Arg.c_str()); + auto DiagOpts = CreateAndPopulateDiagOpts(CLI); + sanitizeDiagOpts(*DiagOpts); + return DiagOpts; +} - // Make a deep copy of the original Clang invocation. - CompilerInvocation OriginalInvocation(*Invocation); +DignosticsEngineWithDiagOpts::DignosticsEngineWithDiagOpts( + ArrayRef<std::string> CommandLine, + IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS, DiagnosticConsumer &DC) { + std::vector<const char *> CCommandLine(CommandLine.size(), nullptr); + llvm::transform(CommandLine, CCommandLine.begin(), + [](const std::string &Str) { return Str.c_str(); }); + DiagOpts = CreateAndPopulateDiagOpts(CCommandLine); + sanitizeDiagOpts(*DiagOpts); + DiagEngine = CompilerInstance::createDiagnostics(*FS, *DiagOpts, &DC, + /*ShouldOwnClient=*/false); +} - if (Scanned) { - // Scanning runs once for the first -cc1 invocation in a chain of driver - // jobs. For any dependent jobs, reuse the scanning result and just - // update the LastCC1Arguments to correspond to the new invocation. - // FIXME: to support multi-arch builds, each arch requires a separate scan - setLastCC1Arguments(std::move(OriginalInvocation)); - return true; +std::pair<std::unique_ptr<driver::Driver>, std::unique_ptr<driver::Compilation>> +buildCompilation(ArrayRef<std::string> ArgStrs, DiagnosticsEngine &Diags, + IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS) { + SmallVector<const char *, 256> Argv; + Argv.reserve(ArgStrs.size()); + for (const std::string &Arg : ArgStrs) + Argv.push_back(Arg.c_str()); + + std::unique_ptr<driver::Driver> Driver = std::make_unique<driver::Driver>( + Argv[0], llvm::sys::getDefaultTargetTriple(), Diags, + "clang LLVM compiler", FS); + Driver->setTitle("clang_based_tool"); + + llvm::BumpPtrAllocator Alloc; + bool CLMode = driver::IsClangCL( + driver::getDriverMode(Argv[0], ArrayRef(Argv).slice(1))); + + 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 std::make_pair(nullptr, nullptr); } - Scanned = true; + std::unique_ptr<driver::Compilation> Compilation( + Driver->BuildCompilation(Argv)); + if (!Compilation) + return std::make_pair(nullptr, nullptr); - // Create a compiler instance to handle the actual work. - auto ModCache = makeInProcessModuleCache(Service.getModuleCacheEntries()); - ScanInstanceStorage.emplace(std::move(Invocation), std::move(PCHContainerOps), - ModCache.get()); - CompilerInstance &ScanInstance = *ScanInstanceStorage; + if (Compilation->containsError()) + return std::make_pair(nullptr, nullptr); + + return std::make_pair(std::move(Driver), std::move(Compilation)); +} + +std::unique_ptr<CompilerInvocation> +createCompilerInvocation(ArrayRef<std::string> CommandLine, + DiagnosticsEngine &Diags) { + llvm::opt::ArgStringList Argv; + for (const std::string &Str : ArrayRef(CommandLine).drop_front()) + Argv.push_back(Str.c_str()); + + auto Invocation = std::make_unique<CompilerInvocation>(); + if (!CompilerInvocation::CreateFromArgs(*Invocation, Argv, Diags)) { + // FIXME: Should we just go on like cc1_main does? + return nullptr; + } + return Invocation; +} + +std::pair<IntrusiveRefCntPtr<llvm::vfs::FileSystem>, std::vector<std::string>> +initVFSForTUBuferScanning(IntrusiveRefCntPtr<llvm::vfs::FileSystem> BaseFS, + ArrayRef<std::string> CommandLine, + StringRef WorkingDirectory, + llvm::MemoryBufferRef TUBuffer) { + // Reset what might have been modified in the previous worker invocation. + BaseFS->setCurrentWorkingDirectory(WorkingDirectory); + + IntrusiveRefCntPtr<llvm::vfs::FileSystem> ModifiedFS; + auto OverlayFS = + llvm::makeIntrusiveRefCnt<llvm::vfs::OverlayFileSystem>(BaseFS); + auto InMemoryFS = llvm::makeIntrusiveRefCnt<llvm::vfs::InMemoryFileSystem>(); + InMemoryFS->setCurrentWorkingDirectory(WorkingDirectory); + auto InputPath = TUBuffer.getBufferIdentifier(); + InMemoryFS->addFile( + InputPath, 0, llvm::MemoryBuffer::getMemBufferCopy(TUBuffer.getBuffer())); + IntrusiveRefCntPtr<llvm::vfs::FileSystem> InMemoryOverlay = InMemoryFS; + + OverlayFS->pushOverlay(InMemoryOverlay); + ModifiedFS = OverlayFS; + std::vector<std::string> ModifiedCommandLine(CommandLine); + ModifiedCommandLine.emplace_back(InputPath); + + return std::make_pair(ModifiedFS, ModifiedCommandLine); +} + +std::pair<IntrusiveRefCntPtr<llvm::vfs::FileSystem>, std::vector<std::string>> +initVFSForByNameScanning(IntrusiveRefCntPtr<llvm::vfs::FileSystem> BaseFS, + ArrayRef<std::string> CommandLine, + StringRef WorkingDirectory, StringRef ModuleName) { + // Reset what might have been modified in the previous worker invocation. + BaseFS->setCurrentWorkingDirectory(WorkingDirectory); + + // If we're scanning based on a module name alone, we don't expect the client + // to provide us with an input file. However, the driver really wants to have + // one. Let's just make it up to make the driver happy. + auto OverlayFS = + llvm::makeIntrusiveRefCnt<llvm::vfs::OverlayFileSystem>(BaseFS); + auto InMemoryFS = llvm::makeIntrusiveRefCnt<llvm::vfs::InMemoryFileSystem>(); + InMemoryFS->setCurrentWorkingDirectory(WorkingDirectory); + SmallString<128> FakeInputPath; + // TODO: We should retry the creation if the path already exists. + llvm::sys::fs::createUniquePath(ModuleName + "-%%%%%%%%.input", FakeInputPath, + /*MakeAbsolute=*/false); + InMemoryFS->addFile(FakeInputPath, 0, llvm::MemoryBuffer::getMemBuffer("")); + IntrusiveRefCntPtr<llvm::vfs::FileSystem> InMemoryOverlay = InMemoryFS; + OverlayFS->pushOverlay(InMemoryOverlay); + + std::vector<std::string> ModifiedCommandLine(CommandLine); + ModifiedCommandLine.emplace_back(FakeInputPath); + + return std::make_pair(OverlayFS, ModifiedCommandLine); +} + +bool initializeScanCompilerInstance( + CompilerInstance &ScanInstance, + IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS, + DiagnosticConsumer *DiagConsumer, DependencyScanningService &Service, + IntrusiveRefCntPtr<DependencyScanningWorkerFilesystem> DepFS) { ScanInstance.setBuildingModule(false); ScanInstance.createVirtualFileSystem(FS, DiagConsumer); // Create the compiler's actual diagnostics engine. sanitizeDiagOpts(ScanInstance.getDiagnosticOpts()); - assert(!DiagConsumerFinished && "attempt to reuse finished consumer"); ScanInstance.createDiagnostics(DiagConsumer, /*ShouldOwnClient=*/false); if (!ScanInstance.hasDiagnostics()) return false; @@ -435,6 +537,26 @@ bool DependencyScanningAction::runInvocation( ScanInstance.createSourceManager(); + // Consider different header search and diagnostic options to create + // different modules. This avoids the unsound aliasing of module PCMs. + // + // TODO: Implement diagnostic bucketing to reduce the impact of strict + // context hashing. + ScanInstance.getHeaderSearchOpts().ModulesStrictContextHash = true; + ScanInstance.getHeaderSearchOpts().ModulesSerializeOnlyPreprocessor = true; + ScanInstance.getHeaderSearchOpts().ModulesSkipDiagnosticOptions = true; + ScanInstance.getHeaderSearchOpts().ModulesSkipHeaderSearchPaths = true; + ScanInstance.getHeaderSearchOpts().ModulesSkipPragmaDiagnosticMappings = true; + ScanInstance.getHeaderSearchOpts().ModulesForceValidateUserHeaders = false; + + // Avoid some checks and module map parsing when loading PCM files. + ScanInstance.getPreprocessorOpts().ModulesCheckRelocated = false; + + return true; +} + +llvm::SmallVector<StringRef> +getInitialStableDirs(const CompilerInstance &ScanInstance) { // Create a collection of stable directories derived from the ScanInstance // for determining whether module dependencies would fully resolve from // those directories. @@ -442,7 +564,12 @@ bool DependencyScanningAction::runInvocation( const StringRef Sysroot = ScanInstance.getHeaderSearchOpts().Sysroot; if (!Sysroot.empty() && (llvm::sys::path::root_directory(Sysroot) != Sysroot)) StableDirs = {Sysroot, ScanInstance.getHeaderSearchOpts().ResourceDir}; + return StableDirs; +} +std::optional<PrebuiltModulesAttrsMap> +computePrebuiltModulesASTMap(CompilerInstance &ScanInstance, + llvm::SmallVector<StringRef> &StableDirs) { // Store a mapping of prebuilt module files and their properties like header // search options. This will prevent the implicit build to create duplicate // modules and will force reuse of the existing prebuilt module files @@ -454,12 +581,14 @@ bool DependencyScanningAction::runInvocation( ScanInstance.getPreprocessorOpts().ImplicitPCHInclude, ScanInstance, ScanInstance.getHeaderSearchOpts().PrebuiltModuleFiles, PrebuiltModulesASTMap, ScanInstance.getDiagnostics(), StableDirs)) - return false; + return {}; - // Create the dependency collector that will collect the produced - // dependencies. - // - // This also moves the existing dependency output options from the + return PrebuiltModulesASTMap; +} + +std::unique_ptr<DependencyOutputOptions> +takeDependencyOutputOptionsFrom(CompilerInstance &ScanInstance) { + // This function moves the existing dependency output options from the // invocation to the collector. The options in the invocation are reset, // which ensures that the compiler won't create new dependency collectors, // and thus won't write out the extra '.d' files to disk. @@ -472,35 +601,85 @@ bool DependencyScanningAction::runInvocation( ScanInstance.getFrontendOpts().Inputs)}; Opts->IncludeSystemHeaders = true; + return Opts; +} + +std::shared_ptr<ModuleDepCollector> initializeScanInstanceDependencyCollector( + CompilerInstance &ScanInstance, + std::unique_ptr<DependencyOutputOptions> DepOutputOpts, + StringRef WorkingDirectory, DependencyConsumer &Consumer, + DependencyScanningService &Service, CompilerInvocation &Inv, + DependencyActionController &Controller, + PrebuiltModulesAttrsMap PrebuiltModulesASTMap, + llvm::SmallVector<StringRef> &StableDirs) { + std::shared_ptr<ModuleDepCollector> MDC; switch (Service.getFormat()) { case ScanningOutputFormat::Make: ScanInstance.addDependencyCollector( std::make_shared<DependencyConsumerForwarder>( - std::move(Opts), WorkingDirectory, Consumer)); + std::move(DepOutputOpts), WorkingDirectory, Consumer)); break; case ScanningOutputFormat::P1689: case ScanningOutputFormat::Full: MDC = std::make_shared<ModuleDepCollector>( - Service, std::move(Opts), ScanInstance, Consumer, Controller, - OriginalInvocation, std::move(PrebuiltModulesASTMap), StableDirs); + Service, std::move(DepOutputOpts), ScanInstance, Consumer, Controller, + Inv, std::move(PrebuiltModulesASTMap), StableDirs); ScanInstance.addDependencyCollector(MDC); break; } - // Consider different header search and diagnostic options to create - // different modules. This avoids the unsound aliasing of module PCMs. - // - // TODO: Implement diagnostic bucketing to reduce the impact of strict - // context hashing. - ScanInstance.getHeaderSearchOpts().ModulesStrictContextHash = true; - ScanInstance.getHeaderSearchOpts().ModulesSerializeOnlyPreprocessor = true; - ScanInstance.getHeaderSearchOpts().ModulesSkipDiagnosticOptions = true; - ScanInstance.getHeaderSearchOpts().ModulesSkipHeaderSearchPaths = true; - ScanInstance.getHeaderSearchOpts().ModulesSkipPragmaDiagnosticMappings = true; - ScanInstance.getHeaderSearchOpts().ModulesForceValidateUserHeaders = false; + return MDC; +} +} // namespace clang::tooling::dependencies - // Avoid some checks and module map parsing when loading PCM files. - ScanInstance.getPreprocessorOpts().ModulesCheckRelocated = false; +bool DependencyScanningAction::runInvocation( + std::unique_ptr<CompilerInvocation> Invocation, + IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS, + std::shared_ptr<PCHContainerOperations> PCHContainerOps, + DiagnosticConsumer *DiagConsumer) { + // Making sure that we canonicalize the defines before we create the deep + // copy to avoid unnecessary variants in the scanner and in the resulting + // explicit command lines. + if (any(Service.getOptimizeArgs() & ScanningOptimizations::Macros)) + canonicalizeDefines(Invocation->getPreprocessorOpts()); + + // Make a deep copy of the original Clang invocation. + CompilerInvocation OriginalInvocation(*Invocation); + + if (Scanned) { + // Scanning runs once for the first -cc1 invocation in a chain of driver + // jobs. For any dependent jobs, reuse the scanning result and just + // update the LastCC1Arguments to correspond to the new invocation. + // FIXME: to support multi-arch builds, each arch requires a separate scan + setLastCC1Arguments(std::move(OriginalInvocation)); + return true; + } + + Scanned = true; + + // Create a compiler instance to handle the actual work. + auto ModCache = makeInProcessModuleCache(Service.getModuleCacheEntries()); + ScanInstanceStorage.emplace(std::move(Invocation), std::move(PCHContainerOps), + ModCache.get()); + CompilerInstance &ScanInstance = *ScanInstanceStorage; + + assert(!DiagConsumerFinished && "attempt to reuse finished consumer"); + if (!initializeScanCompilerInstance(ScanInstance, FS, DiagConsumer, Service, + DepFS)) + return false; + + llvm::SmallVector<StringRef> StableDirs = getInitialStableDirs(ScanInstance); + auto MaybePrebuiltModulesASTMap = + computePrebuiltModulesASTMap(ScanInstance, StableDirs); + if (!MaybePrebuiltModulesASTMap) + return false; + + auto DepOutputOpts = takeDependencyOutputOptionsFrom(ScanInstance); + + MDC = initializeScanInstanceDependencyCollector( + ScanInstance, std::move(DepOutputOpts), WorkingDirectory, Consumer, + Service, OriginalInvocation, Controller, *MaybePrebuiltModulesASTMap, + StableDirs); std::unique_ptr<FrontendAction> Action; diff --git a/clang/lib/Tooling/DependencyScanning/DependencyScannerImpl.h b/clang/lib/Tooling/DependencyScanning/DependencyScannerImpl.h index 32fbcff..71c6731 100644 --- a/clang/lib/Tooling/DependencyScanning/DependencyScannerImpl.h +++ b/clang/lib/Tooling/DependencyScanning/DependencyScannerImpl.h @@ -9,8 +9,10 @@ #ifndef LLVM_CLANG_TOOLING_DEPENDENCYSCANNING_DEPENDENCYSCANNER_H #define LLVM_CLANG_TOOLING_DEPENDENCYSCANNING_DEPENDENCYSCANNER_H +#include "clang/Driver/Compilation.h" #include "clang/Frontend/CompilerInstance.h" #include "clang/Frontend/CompilerInvocation.h" +#include "clang/Frontend/TextDiagnosticPrinter.h" #include "clang/Serialization/ObjectFilePCHContainerReader.h" #include "clang/Tooling/DependencyScanning/DependencyScanningFilesystem.h" #include "clang/Tooling/DependencyScanning/ModuleDepCollector.h" @@ -30,12 +32,12 @@ public: DependencyScanningAction( DependencyScanningService &Service, StringRef WorkingDirectory, DependencyConsumer &Consumer, DependencyActionController &Controller, - llvm::IntrusiveRefCntPtr<DependencyScanningWorkerFilesystem> DepFS, + IntrusiveRefCntPtr<DependencyScanningWorkerFilesystem> DepFS, std::optional<StringRef> ModuleName = std::nullopt) : Service(Service), WorkingDirectory(WorkingDirectory), Consumer(Consumer), Controller(Controller), DepFS(std::move(DepFS)), ModuleName(ModuleName) {} - bool runInvocation(std::shared_ptr<CompilerInvocation> Invocation, + bool runInvocation(std::unique_ptr<CompilerInvocation> Invocation, IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS, std::shared_ptr<PCHContainerOperations> PCHContainerOps, DiagnosticConsumer *DiagConsumer); @@ -63,7 +65,7 @@ private: StringRef WorkingDirectory; DependencyConsumer &Consumer; DependencyActionController &Controller; - llvm::IntrusiveRefCntPtr<DependencyScanningWorkerFilesystem> DepFS; + IntrusiveRefCntPtr<DependencyScanningWorkerFilesystem> DepFS; std::optional<StringRef> ModuleName; std::optional<CompilerInstance> ScanInstanceStorage; std::shared_ptr<ModuleDepCollector> MDC; @@ -72,9 +74,81 @@ private: bool DiagConsumerFinished = false; }; -// Helper functions -void sanitizeDiagOpts(DiagnosticOptions &DiagOpts); +// Helper functions and data types. +std::unique_ptr<DiagnosticOptions> +createDiagOptions(ArrayRef<std::string> CommandLine); +struct DignosticsEngineWithDiagOpts { + // We need to bound the lifetime of the DiagOpts used to create the + // DiganosticsEngine with the DiagnosticsEngine itself. + std::unique_ptr<DiagnosticOptions> DiagOpts; + IntrusiveRefCntPtr<DiagnosticsEngine> DiagEngine; + + DignosticsEngineWithDiagOpts(ArrayRef<std::string> CommandLine, + IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS, + DiagnosticConsumer &DC); +}; + +struct TextDiagnosticsPrinterWithOutput { + // We need to bound the lifetime of the data that supports the DiagPrinter + // with it together so they have the same lifetime. + std::string DiagnosticOutput; + llvm::raw_string_ostream DiagnosticsOS; + std::unique_ptr<DiagnosticOptions> DiagOpts; + TextDiagnosticPrinter DiagPrinter; + + TextDiagnosticsPrinterWithOutput(ArrayRef<std::string> CommandLine) + : DiagnosticsOS(DiagnosticOutput), + DiagOpts(createDiagOptions(CommandLine)), + DiagPrinter(DiagnosticsOS, *DiagOpts) {} +}; + +std::pair<std::unique_ptr<driver::Driver>, std::unique_ptr<driver::Compilation>> +buildCompilation(ArrayRef<std::string> ArgStrs, DiagnosticsEngine &Diags, + IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS); + +std::unique_ptr<CompilerInvocation> +createCompilerInvocation(ArrayRef<std::string> CommandLine, + DiagnosticsEngine &Diags); + +std::pair<IntrusiveRefCntPtr<llvm::vfs::FileSystem>, std::vector<std::string>> +initVFSForTUBuferScanning(IntrusiveRefCntPtr<llvm::vfs::FileSystem> BaseFS, + ArrayRef<std::string> CommandLine, + StringRef WorkingDirectory, + llvm::MemoryBufferRef TUBuffer); + +std::pair<IntrusiveRefCntPtr<llvm::vfs::FileSystem>, std::vector<std::string>> +initVFSForByNameScanning(IntrusiveRefCntPtr<llvm::vfs::FileSystem> BaseFS, + ArrayRef<std::string> CommandLine, + StringRef WorkingDirectory, StringRef ModuleName); + +bool initializeScanCompilerInstance( + CompilerInstance &ScanInstance, + IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS, + DiagnosticConsumer *DiagConsumer, DependencyScanningService &Service, + IntrusiveRefCntPtr<DependencyScanningWorkerFilesystem> DepFS); + +SmallVector<StringRef> +getInitialStableDirs(const CompilerInstance &ScanInstance); + +std::optional<PrebuiltModulesAttrsMap> +computePrebuiltModulesASTMap(CompilerInstance &ScanInstance, + SmallVector<StringRef> &StableDirs); + +std::unique_ptr<DependencyOutputOptions> +takeDependencyOutputOptionsFrom(CompilerInstance &ScanInstance); + +/// Create the dependency collector that will collect the produced +/// dependencies. May return the created ModuleDepCollector depending +/// on the scanning format. +std::shared_ptr<ModuleDepCollector> initializeScanInstanceDependencyCollector( + CompilerInstance &ScanInstance, + std::unique_ptr<DependencyOutputOptions> DepOutputOpts, + StringRef WorkingDirectory, DependencyConsumer &Consumer, + DependencyScanningService &Service, CompilerInvocation &Inv, + DependencyActionController &Controller, + PrebuiltModulesAttrsMap PrebuiltModulesASTMap, + llvm::SmallVector<StringRef> &StableDirs); } // namespace dependencies } // namespace tooling } // namespace clang diff --git a/clang/lib/Tooling/DependencyScanning/DependencyScanningWorker.cpp b/clang/lib/Tooling/DependencyScanning/DependencyScanningWorker.cpp index 796e587..9515421 100644 --- a/clang/lib/Tooling/DependencyScanning/DependencyScanningWorker.cpp +++ b/clang/lib/Tooling/DependencyScanning/DependencyScanningWorker.cpp @@ -8,29 +8,9 @@ #include "clang/Tooling/DependencyScanning/DependencyScanningWorker.h" #include "DependencyScannerImpl.h" -#include "clang/Basic/DiagnosticDriver.h" #include "clang/Basic/DiagnosticFrontend.h" -#include "clang/Basic/DiagnosticSerialization.h" -#include "clang/Driver/Compilation.h" #include "clang/Driver/Driver.h" -#include "clang/Driver/Job.h" #include "clang/Driver/Tool.h" -#include "clang/Frontend/CompilerInstance.h" -#include "clang/Frontend/CompilerInvocation.h" -#include "clang/Frontend/FrontendActions.h" -#include "clang/Frontend/TextDiagnosticPrinter.h" -#include "clang/Frontend/Utils.h" -#include "clang/Lex/PreprocessorOptions.h" -#include "clang/Serialization/ObjectFilePCHContainerReader.h" -#include "clang/Tooling/DependencyScanning/DependencyScanningService.h" -#include "clang/Tooling/DependencyScanning/InProcessModuleCache.h" -#include "clang/Tooling/DependencyScanning/ModuleDepCollector.h" -#include "llvm/ADT/IntrusiveRefCntPtr.h" -#include "llvm/Support/Allocator.h" -#include "llvm/Support/Error.h" -#include "llvm/Support/MemoryBuffer.h" -#include "llvm/TargetParser/Host.h" -#include <optional> using namespace clang; using namespace tooling; @@ -63,32 +43,19 @@ DependencyScanningWorker::DependencyScanningWorker( } } -static std::unique_ptr<DiagnosticOptions> -createDiagOptions(const std::vector<std::string> &CommandLine) { - std::vector<const char *> CLI; - for (const std::string &Arg : CommandLine) - CLI.push_back(Arg.c_str()); - auto DiagOpts = CreateAndPopulateDiagOpts(CLI); - sanitizeDiagOpts(*DiagOpts); - return DiagOpts; -} - llvm::Error DependencyScanningWorker::computeDependencies( StringRef WorkingDirectory, const std::vector<std::string> &CommandLine, DependencyConsumer &Consumer, DependencyActionController &Controller, std::optional<llvm::MemoryBufferRef> TUBuffer) { // Capture the emitted diagnostics and report them to the client // in the case of a failure. - std::string DiagnosticOutput; - llvm::raw_string_ostream DiagnosticsOS(DiagnosticOutput); - auto DiagOpts = createDiagOptions(CommandLine); - TextDiagnosticPrinter DiagPrinter(DiagnosticsOS, *DiagOpts); + TextDiagnosticsPrinterWithOutput DiagPrinterWithOS(CommandLine); if (computeDependencies(WorkingDirectory, CommandLine, Consumer, Controller, - DiagPrinter, TUBuffer)) + DiagPrinterWithOS.DiagPrinter, TUBuffer)) return llvm::Error::success(); - return llvm::make_error<llvm::StringError>(DiagnosticsOS.str(), - llvm::inconvertibleErrorCode()); + return llvm::make_error<llvm::StringError>( + DiagPrinterWithOS.DiagnosticsOS.str(), llvm::inconvertibleErrorCode()); } llvm::Error DependencyScanningWorker::computeDependencies( @@ -97,51 +64,24 @@ llvm::Error DependencyScanningWorker::computeDependencies( StringRef ModuleName) { // Capture the emitted diagnostics and report them to the client // in the case of a failure. - std::string DiagnosticOutput; - llvm::raw_string_ostream DiagnosticsOS(DiagnosticOutput); - auto DiagOpts = createDiagOptions(CommandLine); - TextDiagnosticPrinter DiagPrinter(DiagnosticsOS, *DiagOpts); + TextDiagnosticsPrinterWithOutput DiagPrinterWithOS(CommandLine); if (computeDependencies(WorkingDirectory, CommandLine, Consumer, Controller, - DiagPrinter, ModuleName)) + DiagPrinterWithOS.DiagPrinter, ModuleName)) return llvm::Error::success(); - return llvm::make_error<llvm::StringError>(DiagnosticsOS.str(), - llvm::inconvertibleErrorCode()); + return llvm::make_error<llvm::StringError>( + DiagPrinterWithOS.DiagnosticsOS.str(), llvm::inconvertibleErrorCode()); } static bool forEachDriverJob( 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()); - - std::unique_ptr<driver::Driver> Driver = std::make_unique<driver::Driver>( - Argv[0], llvm::sys::getDefaultTargetTriple(), Diags, - "clang LLVM compiler", FS); - Driver->setTitle("clang_based_tool"); - - llvm::BumpPtrAllocator Alloc; - bool CLMode = driver::IsClangCL( - driver::getDriverMode(Argv[0], ArrayRef(Argv).slice(1))); - - 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; - } - - const std::unique_ptr<driver::Compilation> Compilation( - Driver->BuildCompilation(llvm::ArrayRef(Argv))); + // Compilation holds a non-owning a reference to the Driver, hence we need to + // keep the Driver alive when we use Compilation. + auto [Driver, Compilation] = buildCompilation(ArgStrs, Diags, FS); if (!Compilation) return false; - - if (Compilation->containsError()) - return false; - for (const driver::Command &Job : Compilation->getJobs()) { if (!Callback(Job)) return false; @@ -150,30 +90,21 @@ static bool forEachDriverJob( } static bool createAndRunToolInvocation( - std::vector<std::string> CommandLine, DependencyScanningAction &Action, + const std::vector<std::string> &CommandLine, + DependencyScanningAction &Action, 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]; - - 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? + auto Invocation = createCompilerInvocation(CommandLine, Diags); + if (!Invocation) return false; - } if (!Action.runInvocation(std::move(Invocation), std::move(FS), PCHContainerOps, Diags.getClient())) return false; std::vector<std::string> Args = Action.takeLastCC1Arguments(); - Consumer.handleBuildCommand({std::move(Executable), std::move(Args)}); + Consumer.handleBuildCommand({CommandLine[0], std::move(Args)}); return true; } @@ -182,24 +113,19 @@ bool DependencyScanningWorker::scanDependencies( DependencyConsumer &Consumer, DependencyActionController &Controller, DiagnosticConsumer &DC, llvm::IntrusiveRefCntPtr<llvm::vfs::FileSystem> FS, std::optional<StringRef> ModuleName) { - 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); - auto Diags = CompilerInstance::createDiagnostics(*FS, *DiagOpts, &DC, - /*ShouldOwnClient=*/false); - + DignosticsEngineWithDiagOpts DiagEngineWithCmdAndOpts(CommandLine, FS, DC); DependencyScanningAction Action(Service, WorkingDirectory, Consumer, Controller, DepFS, ModuleName); bool Success = false; if (CommandLine[1] == "-cc1") { - Success = createAndRunToolInvocation(CommandLine, Action, FS, - PCHContainerOps, *Diags, Consumer); + Success = createAndRunToolInvocation( + CommandLine, Action, FS, PCHContainerOps, + *DiagEngineWithCmdAndOpts.DiagEngine, Consumer); } else { Success = forEachDriverJob( - CommandLine, *Diags, FS, [&](const driver::Command &Cmd) { + CommandLine, *DiagEngineWithCmdAndOpts.DiagEngine, FS, + [&](const driver::Command &Cmd) { if (StringRef(Cmd.getCreator().getName()) != "clang") { // Non-clang command. Just pass through to the dependency // consumer. @@ -218,13 +144,15 @@ 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, FS, - PCHContainerOps, *Diags, Consumer); + return createAndRunToolInvocation( + std::move(Argv), Action, FS, PCHContainerOps, + *DiagEngineWithCmdAndOpts.DiagEngine, Consumer); }); } if (Success && !Action.hasScanned()) - Diags->Report(diag::err_fe_expected_compiler_job) + DiagEngineWithCmdAndOpts.DiagEngine->Report( + diag::err_fe_expected_compiler_job) << llvm::join(CommandLine, " "); // Ensure finish() is called even if we never reached ExecuteAction(). @@ -238,66 +166,25 @@ bool DependencyScanningWorker::computeDependencies( StringRef WorkingDirectory, const std::vector<std::string> &CommandLine, DependencyConsumer &Consumer, DependencyActionController &Controller, DiagnosticConsumer &DC, std::optional<llvm::MemoryBufferRef> TUBuffer) { - // Reset what might have been modified in the previous worker invocation. - BaseFS->setCurrentWorkingDirectory(WorkingDirectory); - - std::optional<std::vector<std::string>> ModifiedCommandLine; - llvm::IntrusiveRefCntPtr<llvm::vfs::FileSystem> ModifiedFS; - - // If we're scanning based on a module name alone, we don't expect the client - // to provide us with an input file. However, the driver really wants to have - // one. Let's just make it up to make the driver happy. if (TUBuffer) { - auto OverlayFS = - llvm::makeIntrusiveRefCnt<llvm::vfs::OverlayFileSystem>(BaseFS); - auto InMemoryFS = - llvm::makeIntrusiveRefCnt<llvm::vfs::InMemoryFileSystem>(); - InMemoryFS->setCurrentWorkingDirectory(WorkingDirectory); - auto InputPath = TUBuffer->getBufferIdentifier(); - InMemoryFS->addFile( - InputPath, 0, - llvm::MemoryBuffer::getMemBufferCopy(TUBuffer->getBuffer())); - llvm::IntrusiveRefCntPtr<llvm::vfs::FileSystem> InMemoryOverlay = - InMemoryFS; - - OverlayFS->pushOverlay(InMemoryOverlay); - ModifiedFS = OverlayFS; - ModifiedCommandLine = CommandLine; - ModifiedCommandLine->emplace_back(InputPath); + auto [FinalFS, FinalCommandLine] = initVFSForTUBuferScanning( + BaseFS, CommandLine, WorkingDirectory, *TUBuffer); + return scanDependencies(WorkingDirectory, FinalCommandLine, Consumer, + Controller, DC, FinalFS, + /*ModuleName=*/std::nullopt); + } else { + BaseFS->setCurrentWorkingDirectory(WorkingDirectory); + return scanDependencies(WorkingDirectory, CommandLine, Consumer, Controller, + DC, BaseFS, /*ModuleName=*/std::nullopt); } - - const std::vector<std::string> &FinalCommandLine = - ModifiedCommandLine ? *ModifiedCommandLine : CommandLine; - auto &FinalFS = ModifiedFS ? ModifiedFS : BaseFS; - - return scanDependencies(WorkingDirectory, FinalCommandLine, Consumer, - Controller, DC, FinalFS, /*ModuleName=*/std::nullopt); } bool DependencyScanningWorker::computeDependencies( StringRef WorkingDirectory, const std::vector<std::string> &CommandLine, DependencyConsumer &Consumer, DependencyActionController &Controller, DiagnosticConsumer &DC, StringRef ModuleName) { - // Reset what might have been modified in the previous worker invocation. - BaseFS->setCurrentWorkingDirectory(WorkingDirectory); - - // If we're scanning based on a module name alone, we don't expect the client - // to provide us with an input file. However, the driver really wants to have - // one. Let's just make it up to make the driver happy. - auto OverlayFS = - llvm::makeIntrusiveRefCnt<llvm::vfs::OverlayFileSystem>(BaseFS); - auto InMemoryFS = llvm::makeIntrusiveRefCnt<llvm::vfs::InMemoryFileSystem>(); - InMemoryFS->setCurrentWorkingDirectory(WorkingDirectory); - SmallString<128> FakeInputPath; - // TODO: We should retry the creation if the path already exists. - llvm::sys::fs::createUniquePath(ModuleName + "-%%%%%%%%.input", FakeInputPath, - /*MakeAbsolute=*/false); - InMemoryFS->addFile(FakeInputPath, 0, llvm::MemoryBuffer::getMemBuffer("")); - llvm::IntrusiveRefCntPtr<llvm::vfs::FileSystem> InMemoryOverlay = InMemoryFS; - - OverlayFS->pushOverlay(InMemoryOverlay); - auto ModifiedCommandLine = CommandLine; - ModifiedCommandLine.emplace_back(FakeInputPath); + auto [OverlayFS, ModifiedCommandLine] = initVFSForByNameScanning( + BaseFS, CommandLine, WorkingDirectory, ModuleName); return scanDependencies(WorkingDirectory, ModifiedCommandLine, Consumer, Controller, DC, OverlayFS, ModuleName); diff --git a/clang/test/Analysis/analyze-function-guide.cpp b/clang/test/Analysis/analyze-function-guide.cpp index 96f10010..e260fc4 100644 --- a/clang/test/Analysis/analyze-function-guide.cpp +++ b/clang/test/Analysis/analyze-function-guide.cpp @@ -46,14 +46,17 @@ int fizzbuzz(int x, bool y) { // CHECK-ADVOCATE-DISPLAY-PROGRESS-NEXT: Pass the -analyzer-display-progress for tracking which functions are analyzed. // CHECK-ADVOCATE-DISPLAY-PROGRESS-NOT: For analyzing -// Same as the previous but syntax mode only. -// FIXME: This should have empty standard output. +// The user only enables syntax-only analysis, like `debug.DumpDominators`. +// `-analyze-function` should only match the given function. // -// RUN: %clang_analyze_cc1 -analyzer-checker=core -analyzer-config ipa=none \ +// RUN: %clang_analyze_cc1 -analyzer-checker=core,debug.DumpDominators -analyzer-config ipa=none \ // RUN: -analyze-function='fizzbuzz(int, _Bool)' -x c++ \ // RUN: -triple x86_64-pc-linux-gnu 2>&1 %s \ -// RUN: | FileCheck %s -check-prefix=CHECK-EMPTY3 --allow-empty -// -// FIXME: This should have empty standard output. -// CHECK-EMPTY3: Every top-level function was skipped. -// CHECK-EMPTY3-NEXT: Pass the -analyzer-display-progress for tracking which functions are analyzed. +// RUN: | FileCheck %s -check-prefix=CHECK-SYNTAX-ONLY --allow-empty +// +// With syntax-only analysis, the function is found and analyzed, so no error message. +// CHECK-SYNTAX-ONLY: Immediate dominance tree (Node#,IDom#): +// CHECK-SYNTAX-ONLY-NEXT: (0,1) +// CHECK-SYNTAX-ONLY-NEXT: (1,2) +// CHECK-SYNTAX-ONLY-NEXT: (2,2) +// CHECK-SYNTAX-ONLY-NOT: Every top-level function was skipped. diff --git a/clang/test/Analysis/analyzeOneFunction.cpp b/clang/test/Analysis/analyzeOneFunction.cpp new file mode 100644 index 0000000..3a362df --- /dev/null +++ b/clang/test/Analysis/analyzeOneFunction.cpp @@ -0,0 +1,18 @@ +// RUN: %clang_analyze_cc1 -analyzer-checker=core,debug.ExprInspection -verify %s \ +// RUN: -analyze-function="Window::overloaded(int)" + +// RUN: %clang_analyze_cc1 -analyzer-checker=core,debug.ExprInspection -verify %s \ +// RUN: -analyze-function="c:@S@Window@F@overloaded#I#" + +// RUN: %clang_extdef_map %s | FileCheck %s +// CHECK: 27:c:@S@Window@F@overloaded#I# +// CHECK-NEXT: 27:c:@S@Window@F@overloaded#C# +// CHECK-NEXT: 27:c:@S@Window@F@overloaded#d# + +void clang_analyzer_warnIfReached(); + +struct Window { + void overloaded(double) { clang_analyzer_warnIfReached(); } // not analyzed, thus not reachable + void overloaded(char) { clang_analyzer_warnIfReached(); } // not analyzed, thus not reachable + void overloaded(int) { clang_analyzer_warnIfReached(); } // expected-warning {{REACHABLE}} +}; diff --git a/clang/test/Analysis/analyzer-stats/entry-point-stats.cpp b/clang/test/Analysis/analyzer-stats/entry-point-stats.cpp index 1ff31d1..9cbe045 100644 --- a/clang/test/Analysis/analyzer-stats/entry-point-stats.cpp +++ b/clang/test/Analysis/analyzer-stats/entry-point-stats.cpp @@ -5,7 +5,9 @@ // RUN: %csv2json "%t.csv" | FileCheck --check-prefix=CHECK %s // // CHECK: { -// CHECK-NEXT: "fib(unsigned int)": { +// CHECK-NEXT: "c:@F@fib#i#": { +// CHECK-NEXT: "File": "{{.*}}entry-point-stats.cpp", +// CHECK-NEXT: "DebugName": "fib(unsigned int)", // CHECK-NEXT: "NumBlocks": "{{[0-9]+}}", // CHECK-NEXT: "NumBlocksUnreachable": "{{[0-9]+}}", // CHECK-NEXT: "NumCTUSteps": "{{[0-9]+}}", @@ -40,7 +42,9 @@ // CHECK-NEXT: "MaxValidBugClassSize": "{{[0-9]+}}", // CHECK-NEXT: "PathRunningTime": "{{[0-9]+}}" // CHECK-NEXT: }, -// CHECK-NEXT: "main(int, char **)": { +// CHECK-NEXT: "c:@F@main#I#**C#": { +// CHECK-NEXT: "File": "{{.*}}entry-point-stats.cpp", +// CHECK-NEXT: "DebugName": "main(int, char **)", // CHECK-NEXT: "NumBlocks": "{{[0-9]+}}", // CHECK-NEXT: "NumBlocksUnreachable": "{{[0-9]+}}", // CHECK-NEXT: "NumCTUSteps": "{{[0-9]+}}", diff --git a/clang/test/Analysis/csv2json.py b/clang/test/Analysis/csv2json.py index 3c20d68..6e1aca9 100644 --- a/clang/test/Analysis/csv2json.py +++ b/clang/test/Analysis/csv2json.py @@ -44,7 +44,7 @@ def csv_to_json_dict(csv_filepath): """ try: with open(csv_filepath, "r", encoding="utf-8") as csvfile: - reader = csv.reader(csvfile) + reader = csv.reader(csvfile, skipinitialspace=True) # Read the header row (column names) try: @@ -58,12 +58,13 @@ def csv_to_json_dict(csv_filepath): json.dumps({}, indent=2) return - other_column_names = [name.strip() for name in header[1:]] + header_length = len(header) + other_column_names = header[1:] data_dict = {} for row in reader: - if len(row) != len(header): + if len(row) != header_length: raise csv.Error("Inconsistent CSV file") exit(1) diff --git a/clang/test/CIR/CodeGen/address-space.c b/clang/test/CIR/CodeGen/address-space.c new file mode 100644 index 0000000..a334b8a --- /dev/null +++ b/clang/test/CIR/CodeGen/address-space.c @@ -0,0 +1,30 @@ +// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -fclangir -emit-cir %s -o %t.cir +// RUN: FileCheck --input-file=%t.cir %s -check-prefix=CIR +// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -fclangir -emit-llvm %s -o %t.ll +// RUN: FileCheck --input-file=%t.ll %s -check-prefix=LLVM +// RUN: %clang_cc1 -triple x86_64-unknown-linux-gnu -emit-llvm %s -o %t.ll +// RUN: FileCheck --input-file=%t.ll %s -check-prefix=OGCG + +// Test address space 1 +// CIR: cir.func dso_local @foo(%arg0: !cir.ptr<!s32i, target_address_space(1)> +// LLVM: define dso_local void @foo(ptr addrspace(1) %0) +// OGCG: define dso_local void @foo(ptr addrspace(1) noundef %arg) +void foo(int __attribute__((address_space(1))) *arg) { + return; +} + +// Test explicit address space 0 (should be same as default) +// CIR: cir.func dso_local @bar(%arg0: !cir.ptr<!s32i, target_address_space(0)> +// LLVM: define dso_local void @bar(ptr %0) +// OGCG: define dso_local void @bar(ptr noundef %arg) +void bar(int __attribute__((address_space(0))) *arg) { + return; +} + +// Test default address space (no attribute) +// CIR: cir.func dso_local @baz(%arg0: !cir.ptr<!s32i> +// LLVM: define dso_local void @baz(ptr %0) +// OGCG: define dso_local void @baz(ptr noundef %arg) +void baz(int *arg) { + return; +} diff --git a/clang/test/CIR/CodeGen/coro-task.cpp b/clang/test/CIR/CodeGen/coro-task.cpp new file mode 100644 index 0000000..1fc7d77 --- /dev/null +++ b/clang/test/CIR/CodeGen/coro-task.cpp @@ -0,0 +1,123 @@ +// RUN: %clang_cc1 -std=c++20 -triple x86_64-unknown-linux-gnu -fclangir -emit-cir %s -o %t.cir +// RUN: FileCheck --input-file=%t.cir %s -check-prefix=CIR + +namespace std { + +template<typename T> struct remove_reference { typedef T type; }; +template<typename T> struct remove_reference<T &> { typedef T type; }; +template<typename T> struct remove_reference<T &&> { typedef T type; }; + +template<typename T> +typename remove_reference<T>::type &&move(T &&t) noexcept; + +template <class Ret, typename... T> +struct coroutine_traits { using promise_type = typename Ret::promise_type; }; + +template <class Promise = void> +struct coroutine_handle { + static coroutine_handle from_address(void *) noexcept; +}; +template <> +struct coroutine_handle<void> { + template <class PromiseType> + coroutine_handle(coroutine_handle<PromiseType>) noexcept; + static coroutine_handle from_address(void *); +}; + +struct suspend_always { + bool await_ready() noexcept { return false; } + void await_suspend(coroutine_handle<>) noexcept {} + void await_resume() noexcept {} +}; + +struct suspend_never { + bool await_ready() noexcept { return true; } + void await_suspend(coroutine_handle<>) noexcept {} + void await_resume() noexcept {} +}; + +} // namespace std + +namespace folly { +namespace coro { + +using std::suspend_always; +using std::suspend_never; +using std::coroutine_handle; + +using SemiFuture = int; + +template<class T> +struct Task { + struct promise_type { + Task<T> get_return_object() noexcept; + suspend_always initial_suspend() noexcept; + suspend_always final_suspend() noexcept; + void return_value(T); + void unhandled_exception(); + auto yield_value(Task<T>) noexcept { return final_suspend(); } + }; + bool await_ready() noexcept { return false; } + void await_suspend(coroutine_handle<>) noexcept {} + T await_resume(); +}; + +template<> +struct Task<void> { + struct promise_type { + Task<void> get_return_object() noexcept; + suspend_always initial_suspend() noexcept; + suspend_always final_suspend() noexcept; + void return_void() noexcept; + void unhandled_exception() noexcept; + auto yield_value(Task<void>) noexcept { return final_suspend(); } + }; + bool await_ready() noexcept { return false; } + void await_suspend(coroutine_handle<>) noexcept {} + void await_resume() noexcept {} + SemiFuture semi(); +}; + +// FIXME: add CIRGen support here. +// struct blocking_wait_fn { +// template <typename T> +// T operator()(Task<T>&& awaitable) const { +// return T(); +// } +// }; + +// inline constexpr blocking_wait_fn blocking_wait{}; +// static constexpr blocking_wait_fn const& blockingWait = blocking_wait; + +struct co_invoke_fn { + template <typename F, typename... A> + Task<void> operator()(F&& f, A&&... a) const { + return Task<void>(); + } +}; + +co_invoke_fn co_invoke; + +}} // namespace folly::coro + +// CIR-DAG: ![[VoidTask:.*]] = !cir.record<struct "folly::coro::Task<void>" padded {!u8i}> + +// CIR: module {{.*}} { +// CIR-NEXT: cir.global external @_ZN5folly4coro9co_invokeE = #cir.zero : !rec_folly3A3Acoro3A3Aco_invoke_fn + +// CIR: cir.func builtin private @__builtin_coro_id(!u32i, !cir.ptr<!void>, !cir.ptr<!void>, !cir.ptr<!void>) -> !u32i + +using VoidTask = folly::coro::Task<void>; + +VoidTask silly_task() { + co_await std::suspend_always(); +} + +// CIR: cir.func coroutine dso_local @_Z10silly_taskv() -> ![[VoidTask]] +// CHECK: %[[#VoidTaskAddr:]] = cir.alloca ![[VoidTask]], {{.*}}, ["__retval"] + +// Get coroutine id with __builtin_coro_id. + +// CIR: %[[NullPtr:.*]] = cir.const #cir.ptr<null> : !cir.ptr<!void> +// CIR: %[[Align:.*]] = cir.const #cir.int<16> : !u32i +// CIR: %[[CoroId:.*]] = cir.call @__builtin_coro_id(%[[Align]], %[[NullPtr]], %[[NullPtr]], %[[NullPtr]]) diff --git a/clang/test/CIR/CodeGen/predefined-expr.c b/clang/test/CIR/CodeGen/predefined-expr.c new file mode 100644 index 0000000..674c9bd0 --- /dev/null +++ b/clang/test/CIR/CodeGen/predefined-expr.c @@ -0,0 +1,71 @@ +// RUN: %clang_cc1 %s -triple x86_64-unknown-linux-gnu -fclangir -emit-cir -o %t.cir +// RUN: FileCheck %s --input-file=%t.cir --check-prefix=CIR +// RUN: %clang_cc1 %s -triple x86_64-unknown-linux-gnu -fclangir -emit-llvm -o %t-cir.ll +// RUN: FileCheck %s --input-file=%t-cir.ll --check-prefix=LLVM +// RUN: %clang_cc1 %s -triple x86_64-unknown-linux-gnu -emit-llvm -o %t.ll +// RUN: FileCheck %s --input-file=%t.ll --check-prefix=OGCG + +// CIR: cir.global "private" constant cir_private dso_local @__func__.plainFunction = #cir.const_array<"plainFunction\00" : !cir.array<!s8i x 14>> +// CIR: cir.global "private" constant cir_private dso_local @__PRETTY_FUNCTION__.plainFunction = #cir.const_array<"void plainFunction(void)\00" : !cir.array<!s8i x 25>> +// CIR: cir.global "private" constant cir_private dso_local @__func__.externFunction = #cir.const_array<"externFunction\00" : !cir.array<!s8i x 15>> +// CIR: cir.global "private" constant cir_private dso_local @__PRETTY_FUNCTION__.externFunction = #cir.const_array<"void externFunction(void)\00" : !cir.array<!s8i x 26>> +// CIR: cir.global "private" constant cir_private dso_local @__func__.privateExternFunction = #cir.const_array<"privateExternFunction\00" : !cir.array<!s8i x 22>> +// CIR: cir.global "private" constant cir_private dso_local @__PRETTY_FUNCTION__.privateExternFunction = #cir.const_array<"void privateExternFunction(void)\00" : !cir.array<!s8i x 33>> +// CIR: cir.global "private" constant cir_private dso_local @__func__.staticFunction = #cir.const_array<"staticFunction\00" : !cir.array<!s8i x 15>> +// CIR: cir.global "private" constant cir_private dso_local @__PRETTY_FUNCTION__.staticFunction = #cir.const_array<"void staticFunction(void)\00" : !cir.array<!s8i x 26>> + +// TODO(cir): These should be unnamed_addr +// LLVM: @__func__.plainFunction = private constant [14 x i8] c"plainFunction\00" +// LLVM: @__PRETTY_FUNCTION__.plainFunction = private constant [25 x i8] c"void plainFunction(void)\00" +// LLVM: @__func__.externFunction = private constant [15 x i8] c"externFunction\00" +// LLVM: @__PRETTY_FUNCTION__.externFunction = private constant [26 x i8] c"void externFunction(void)\00" +// LLVM: @__func__.privateExternFunction = private constant [22 x i8] c"privateExternFunction\00" +// LLVM: @__PRETTY_FUNCTION__.privateExternFunction = private constant [33 x i8] c"void privateExternFunction(void)\00" +// LLVM: @__func__.staticFunction = private constant [15 x i8] c"staticFunction\00" +// LLVM: @__PRETTY_FUNCTION__.staticFunction = private constant [26 x i8] c"void staticFunction(void)\00" + +// OGCG: @__func__.plainFunction = private unnamed_addr constant [14 x i8] c"plainFunction\00" +// OGCG: @__PRETTY_FUNCTION__.plainFunction = private unnamed_addr constant [25 x i8] c"void plainFunction(void)\00" +// OGCG: @__func__.externFunction = private unnamed_addr constant [15 x i8] c"externFunction\00" +// OGCG: @__PRETTY_FUNCTION__.externFunction = private unnamed_addr constant [26 x i8] c"void externFunction(void)\00" +// OGCG: @__func__.privateExternFunction = private unnamed_addr constant [22 x i8] c"privateExternFunction\00" +// OGCG: @__PRETTY_FUNCTION__.privateExternFunction = private unnamed_addr constant [33 x i8] c"void privateExternFunction(void)\00" +// OGCG: @__func__.staticFunction = private unnamed_addr constant [15 x i8] c"staticFunction\00" +// OGCG: @__PRETTY_FUNCTION__.staticFunction = private unnamed_addr constant [26 x i8] c"void staticFunction(void)\00" + +int printf(const char *, ...); + +void plainFunction(void) { + printf("__func__ %s\n", __func__); + printf("__FUNCTION__ %s\n", __FUNCTION__); + printf("__PRETTY_FUNCTION__ %s\n\n", __PRETTY_FUNCTION__); +} + +extern void externFunction(void) { + printf("__func__ %s\n", __func__); + printf("__FUNCTION__ %s\n", __FUNCTION__); + printf("__PRETTY_FUNCTION__ %s\n\n", __PRETTY_FUNCTION__); +} + +__private_extern__ void privateExternFunction(void) { + printf("__func__ %s\n", __func__); + printf("__FUNCTION__ %s\n", __FUNCTION__); + printf("__PRETTY_FUNCTION__ %s\n\n", __PRETTY_FUNCTION__); +} + +// TODO(cir): Add support for __captured_stmt + +static void staticFunction(void) { + printf("__func__ %s\n", __func__); + printf("__FUNCTION__ %s\n", __FUNCTION__); + printf("__PRETTY_FUNCTION__ %s\n\n", __PRETTY_FUNCTION__); +} + +int main(void) { + plainFunction(); + externFunction(); + privateExternFunction(); + staticFunction(); + + return 0; +} diff --git a/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-default-ops.cpp b/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-default-ops.cpp index 3d295d5..36d8c5ed 100644 --- a/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-default-ops.cpp +++ b/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-default-ops.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s struct DefaultOperators { int i; @@ -944,22 +944,436 @@ void acc_combined() { for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(+:someVarArr[1:1]) @@ -980,8 +1394,6 @@ void acc_combined() { for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(||:someVarArr[1:1]) for(int i=0;i < 5; ++i); - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z12acc_combined } diff --git a/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-float.cpp b/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-float.cpp index be33afe..d3d500d 100644 --- a/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-float.cpp +++ b/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-float.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s template<typename T> void acc_combined() { T someVar; @@ -403,22 +403,319 @@ void acc_combined() { for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[LEAST]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[LARGEST]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ALL_ONES]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(+:someVarArr[1:1]) @@ -439,8 +736,6 @@ void acc_combined() { for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(||:someVarArr[1:1]) for(int i=0;i < 5; ++i); - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z12acc_combined } diff --git a/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-inline-ops.cpp b/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-inline-ops.cpp index f13d96d..df7dc5d 100644 --- a/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-inline-ops.cpp +++ b/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-inline-ops.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s struct HasOperatorsInline { int i; @@ -1172,22 +1172,697 @@ void acc_combined() { for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(+:someVarArr[1:1]) @@ -1208,8 +1883,6 @@ void acc_combined() { for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(||:someVarArr[1:1]) for(int i=0;i < 5; ++i); - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z12acc_combined } diff --git a/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-int.cpp b/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-int.cpp index 952fee9b..8ca4ffa 100644 --- a/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-int.cpp +++ b/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-int.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s template<typename T> void acc_combined() { @@ -406,22 +406,319 @@ void acc_combined() { for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[LEAST]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[LARGEST]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ALL_ONES]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(+:someVarArr[1:1]) @@ -442,8 +739,6 @@ void acc_combined() { for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(||:someVarArr[1:1]) for(int i=0;i < 5; ++i); - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z12acc_combined } diff --git a/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-outline-ops.cpp b/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-outline-ops.cpp index 15646ed..99d5bd2 100644 --- a/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-outline-ops.cpp +++ b/clang/test/CIR/CodeGenOpenACC/combined-reduction-clause-outline-ops.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s struct HasOperatorsOutline { int i; unsigned u; @@ -1172,22 +1172,697 @@ void acc_combined() { for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc parallel loop reduction(+:someVarArr[1:1]) @@ -1209,8 +1884,6 @@ void acc_combined() { #pragma acc parallel loop reduction(||:someVarArr[1:1]) for(int i=0;i < 5; ++i); - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z12acc_combined } diff --git a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-default-ops.c b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-default-ops.c index e357f44..8f45c77 100644 --- a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-default-ops.c +++ b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-default-ops.c @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -std=c23 -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -std=c23 -triple x86_64-linux-pc %s -o - | FileCheck %s struct DefaultOperators { int i; @@ -888,22 +888,436 @@ void acc_compute() { ; #pragma acc parallel reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(+:someVarArr[1:1]) @@ -925,8 +1339,6 @@ void acc_compute() { #pragma acc parallel reduction(||:someVarArr[1:1]) ; - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@acc_compute } diff --git a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-default-ops.cpp b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-default-ops.cpp index e0098bc..c61d047 100644 --- a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-default-ops.cpp +++ b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-default-ops.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s struct DefaultOperators { int i; @@ -944,22 +944,436 @@ void acc_compute() { ; #pragma acc parallel reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(+:someVarArr[1:1]) @@ -980,8 +1394,6 @@ void acc_compute() { ; #pragma acc parallel reduction(||:someVarArr[1:1]) ; - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z11acc_compute } diff --git a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-float.c b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-float.c index 5336fad..3e4aa6f 100644 --- a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-float.c +++ b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-float.c @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s void acc_compute() { float someVar; @@ -403,22 +403,319 @@ void acc_compute() { ; #pragma acc parallel reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[LEAST]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[LARGEST]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ALL_ONES]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(+:someVarArr[1:1]) @@ -439,7 +736,5 @@ void acc_compute() { ; #pragma acc parallel reduction(||:someVarArr[1:1]) ; - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@acc_compute } diff --git a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-float.cpp b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-float.cpp index a513882..fce4c93 100644 --- a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-float.cpp +++ b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-float.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s template<typename T> void acc_compute() { @@ -404,22 +404,319 @@ void acc_compute() { ; #pragma acc parallel reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[LEAST]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[LARGEST]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ALL_ONES]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(+:someVarArr[1:1]) @@ -440,8 +737,6 @@ void acc_compute() { ; #pragma acc parallel reduction(||:someVarArr[1:1]) ; - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z11acc_compute } diff --git a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-inline-ops.cpp b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-inline-ops.cpp index 1968c0a..635de6a 100644 --- a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-inline-ops.cpp +++ b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-inline-ops.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s struct HasOperatorsInline { int i; @@ -1172,22 +1172,697 @@ void acc_compute() { ; #pragma acc parallel reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(+:someVarArr[1:1]) @@ -1208,8 +1883,6 @@ void acc_compute() { ; #pragma acc parallel reduction(||:someVarArr[1:1]) ; - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z11acc_compute } diff --git a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-int.c b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-int.c index f63e340..da5f4c0 100644 --- a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-int.c +++ b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-int.c @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s void acc_compute() { int someVar; @@ -404,22 +404,319 @@ void acc_compute() { ; #pragma acc parallel reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[LEAST]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[LARGEST]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ALL_ONES]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(+:someVarArr[1:1]) @@ -440,7 +737,5 @@ void acc_compute() { ; #pragma acc parallel reduction(||:someVarArr[1:1]) ; - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@acc_compute } diff --git a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-int.cpp b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-int.cpp index 48e5ac9..933a7a4 100644 --- a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-int.cpp +++ b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-int.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s template<typename T> void acc_compute() { @@ -406,22 +406,319 @@ void acc_compute() { ; #pragma acc parallel reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[LEAST]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[LARGEST]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ALL_ONES]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(+:someVarArr[1:1]) @@ -442,8 +739,6 @@ void acc_compute() { ; #pragma acc parallel reduction(||:someVarArr[1:1]) ; - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z11acc_compute } diff --git a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-outline-ops.cpp b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-outline-ops.cpp index 6d204bc..b078eba 100644 --- a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-outline-ops.cpp +++ b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-outline-ops.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s struct HasOperatorsOutline { int i; unsigned u; @@ -1172,22 +1172,697 @@ void acc_compute() { ; #pragma acc parallel reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } ; #pragma acc parallel reduction(+:someVarArr[1:1]) @@ -1209,8 +1884,6 @@ void acc_compute() { #pragma acc parallel reduction(||:someVarArr[1:1]) ; - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z11acc_compute } diff --git a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-unsigned-int.c b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-unsigned-int.c index 35a7e7a..81139a7 100644 --- a/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-unsigned-int.c +++ b/clang/test/CIR/CodeGenOpenACC/compute-reduction-clause-unsigned-int.c @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s void acc_compute() { unsigned int someVar; @@ -404,22 +404,319 @@ void acc_compute() { ; #pragma acc parallel reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_j : !cir.ptr<!cir.array<!u32i x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!u32i x 5>, !cir.ptr<!cir.array<!u32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!u32i x 5>> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!u32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!u32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_j : !cir.ptr<!cir.array<!u32i x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!u32i x 5>, !cir.ptr<!cir.array<!u32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!u32i x 5>> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!u32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!u32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_j : !cir.ptr<!cir.array<!u32i x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!u32i x 5>, !cir.ptr<!cir.array<!u32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!u32i x 5>> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!u32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!u32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store{{.*}} %[[LEAST]], %[[STRIDE]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!u32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_j : !cir.ptr<!cir.array<!u32i x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!u32i x 5>, !cir.ptr<!cir.array<!u32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!u32i x 5>> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!u32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!u32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store{{.*}} %[[LARGEST]], %[[STRIDE]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!u32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_j : !cir.ptr<!cir.array<!u32i x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!u32i x 5>, !cir.ptr<!cir.array<!u32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!u32i x 5>> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!u32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!u32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store{{.*}} %[[ALL_ONES]], %[[STRIDE]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!u32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_j : !cir.ptr<!cir.array<!u32i x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!u32i x 5>, !cir.ptr<!cir.array<!u32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!u32i x 5>> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!u32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!u32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_j : !cir.ptr<!cir.array<!u32i x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!u32i x 5>, !cir.ptr<!cir.array<!u32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!u32i x 5>> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!u32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!u32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_j : !cir.ptr<!cir.array<!u32i x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!u32i x 5>, !cir.ptr<!cir.array<!u32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!u32i x 5>> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!u32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!u32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_j : !cir.ptr<!cir.array<!u32i x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!u32i x 5>, !cir.ptr<!cir.array<!u32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!u32i x 5>> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!u32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!u32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!u32i x 5>> +// CHECK-NEXT: } ; #pragma acc parallel reduction(+:someVarArr[1:1]) @@ -440,7 +737,5 @@ void acc_compute() { ; #pragma acc parallel reduction(||:someVarArr[1:1]) ; - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@acc_compute } diff --git a/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-default-ops.cpp b/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-default-ops.cpp index 73b8fe2..bc4768e 100644 --- a/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-default-ops.cpp +++ b/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-default-ops.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s struct DefaultOperators { int i; @@ -944,22 +944,436 @@ void acc_loop() { for(int i=0;i < 5; ++i); #pragma acc loop reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_16DefaultOperators : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_DefaultOperators x 5>, !cir.ptr<!cir.array<!rec_DefaultOperators x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> -> !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_DefaultOperators>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_DefaultOperators> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][3] {name = "d"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][4] {name = "b"} : !cir.ptr<!rec_DefaultOperators> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_DefaultOperators x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(+:someVarArr[1:1]) @@ -980,8 +1394,6 @@ void acc_loop() { for(int i=0;i < 5; ++i); #pragma acc loop reduction(||:someVarArr[1:1]) for(int i=0;i < 5; ++i); - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z8acc_loop } diff --git a/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-float.cpp b/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-float.cpp index 77c6138..6b29ab5 100644 --- a/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-float.cpp +++ b/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-float.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s template<typename T> void acc_loop() { @@ -404,22 +404,319 @@ void acc_loop() { for(int i=0;i < 5; ++i); #pragma acc loop reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[LEAST]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[LARGEST]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ALL_ONES]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_f : !cir.ptr<!cir.array<!cir.float x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!cir.float x 5>, !cir.ptr<!cir.array<!cir.float x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!cir.float x 5>> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.float>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!cir.float x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!cir.float x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(+:someVarArr[1:1]) @@ -440,8 +737,6 @@ void acc_loop() { for(int i=0;i < 5; ++i); #pragma acc loop reduction(||:someVarArr[1:1]) for(int i=0;i < 5; ++i); - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z8acc_loop } diff --git a/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-inline-ops.cpp b/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-inline-ops.cpp index 6ca0654..df07041 100644 --- a/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-inline-ops.cpp +++ b/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-inline-ops.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s struct HasOperatorsInline { int i; @@ -1172,22 +1172,697 @@ void acc_loop() { for(int i=0;i < 5; ++i); #pragma acc loop reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_18HasOperatorsInline : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsInline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsInline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsInline x 5>> -> !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsInline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsInline> +// CHECK-NEXT: cir.call @_ZN18HasOperatorsInlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsInline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(+:someVarArr[1:1]) @@ -1208,8 +1883,6 @@ void acc_loop() { for(int i=0;i < 5; ++i); #pragma acc loop reduction(||:someVarArr[1:1]) for(int i=0;i < 5; ++i); - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z8acc_loop } diff --git a/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-int.cpp b/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-int.cpp index dd3c54f..19f96f2 100644 --- a/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-int.cpp +++ b/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-int.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s template<typename T> void acc_loop() { @@ -406,22 +406,319 @@ void acc_loop() { for(int i=0;i < 5; ++i); #pragma acc loop reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[LEAST]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[LARGEST]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ALL_ONES]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_i : !cir.ptr<!cir.array<!s32i x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!s32i x 5>, !cir.ptr<!cir.array<!s32i x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!s32i x 5>> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!s32i>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[STRIDE]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!s32i x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!s32i x 5>> +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(+:someVarArr[1:1]) @@ -442,8 +739,6 @@ void acc_loop() { for(int i=0;i < 5; ++i); #pragma acc loop reduction(||:someVarArr[1:1]) for(int i=0;i < 5; ++i); - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z8acc_loop } diff --git a/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-outline-ops.cpp b/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-outline-ops.cpp index d36f9c6..ccc5db6 100644 --- a/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-outline-ops.cpp +++ b/clang/test/CIR/CodeGenOpenACC/loop-reduction-clause-outline-ops.cpp @@ -1,4 +1,4 @@ -// RUN: not %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s struct HasOperatorsOutline { int i; unsigned u; @@ -1172,24 +1172,698 @@ void acc_loop() { for(int i=0;i < 5; ++i); #pragma acc loop reduction(+:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_add__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(*:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_mul__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(max:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_max__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <max> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<-2147483648> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #cir.fp<-1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LEAST:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[LEAST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(min:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_min__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <min> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<2147483647> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<3.4{{.*}}E+38> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #cir.fp<1.7{{.*}}E+308> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[LARGEST:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[LARGEST]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_iand__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <iand> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<-1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.int<4294967295> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #cir.fp<0xFF{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ALL_ONES:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ALL_ONES]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(|:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_ior__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <ior> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(^:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_xor__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <xor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(&&:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_land__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <land> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.fp<1{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #true +// CHECK-NEXT: cir.store {{.*}} %[[ONE]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); #pragma acc loop reduction(||:someVarArr[2]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt1__ZTSA5_19HasOperatorsOutline : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>{{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: %[[ALLOCA:.*]] = cir.alloca !cir.array<!rec_HasOperatorsOutline x 5>, !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>>, ["openacc.reduction.init"] +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB_CAST]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ALLOCA]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: %[[GET_U:.*]] = cir.get_member %[[STRIDE]][1] {name = "u"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!u32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u32i +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_U]] : !u32i, !cir.ptr<!u32i> +// CHECK-NEXT: %[[GET_F:.*]] = cir.get_member %[[STRIDE]][2] {name = "f"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.float> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.float +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_F]] : !cir.float, !cir.ptr<!cir.float> +// CHECK-NEXT: %[[GET_D:.*]] = cir.get_member %[[STRIDE]][4] {name = "d"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.double> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.fp<0{{.*}}> : !cir.double +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_D]] : !cir.double, !cir.ptr<!cir.double> +// CHECK-NEXT: %[[GET_B:.*]] = cir.get_member %[[STRIDE]][5] {name = "b"} : !cir.ptr<!rec_HasOperatorsOutline> -> !cir.ptr<!cir.bool> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #false +// CHECK-NEXT: cir.store {{.*}} %[[ZERO]], %[[GET_B]] : !cir.bool, !cir.ptr<!cir.bool> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[LHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[RHSARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}})) +// CHECK-NEXT: acc.yield %[[LHSARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[ORIG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[ARG:.*]]: !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty{{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB]] : index to !u64i +// CHECK-NEXT: %[[UB:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB]] : index to !u64i +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR_LOAD]], %[[LB_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARG]] : !cir.ptr<!cir.array<!rec_HasOperatorsOutline x 5>> -> !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!rec_HasOperatorsOutline>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!rec_HasOperatorsOutline> +// CHECK-NEXT: cir.call @_ZN19HasOperatorsOutlineD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_HasOperatorsOutline>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } for(int i=0;i < 5; ++i); - #pragma acc loop reduction(+:someVarArr[1:1]) for(int i=0;i < 5; ++i); #pragma acc loop reduction(*:someVarArr[1:1]) @@ -1209,8 +1883,6 @@ void acc_loop() { #pragma acc loop reduction(||:someVarArr[1:1]) for(int i=0;i < 5; ++i); - // TODO OpenACC: When pointers/arrays are handled correctly, we should see all - // of the above repeated for arrays/pointers. // CHECK-NEXT: cir.func {{.*}}@_Z8acc_loop } diff --git a/clang/test/CIR/CodeGenOpenACC/reduction-clause-recipes.cpp b/clang/test/CIR/CodeGenOpenACC/reduction-clause-recipes.cpp new file mode 100644 index 0000000..4c012aa --- /dev/null +++ b/clang/test/CIR/CodeGenOpenACC/reduction-clause-recipes.cpp @@ -0,0 +1,677 @@ +// RUN: %clang_cc1 -fopenacc -triple x86_64-linux-gnu -Wno-openacc-self-if-potential-conflict -emit-cir -fclangir -triple x86_64-linux-pc %s -o - | FileCheck %s + +// Note: unlike the 'private' recipe checks, this is just for spot-checking, +// so this test isn't as comprehensive. The same code paths are used for +// 'private', so we just count on those to catch the errors. +struct NoOps { + int i; + ~NoOps(); +}; +void do_things(unsigned A, unsigned B) { + NoOps ThreeArr[5][5][5]; + +#pragma acc parallel reduction(+:ThreeArr[B][B][B]) +// CHECK:acc.reduction.recipe @reduction_add__Bcnt3__ZTSA5_A5_A5_5NoOps : !cir.ptr<!cir.array<!cir.array<!cir.array<!rec_NoOps x 5> x 5> x 5>> reduction_operator <add> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.array<!cir.array<!cir.array<!rec_NoOps x 5> x 5> x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND2:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND3:.*]]: !acc.data_bounds_ty {{.*}}): +// CHECK-NEXT: %[[TL_ALLOCA:.*]] = cir.alloca !cir.array<!cir.array<!cir.array<!rec_NoOps x 5> x 5> x 5>, !cir.ptr<!cir.array<!cir.array<!cir.array<!rec_NoOps x 5> x 5> x 5>>, ["openacc.reduction.init"] {alignment = 4 : i64} +// +// Init Section: +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB3:.*]] = acc.get_lowerbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB3]] : index to !u64i +// CHECK-NEXT: %[[UB3:.*]] = acc.get_upperbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB3]] : index to !u64i +// CHECK-NEXT: %[[ITR3:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB3_CAST]], %[[ITR3]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR3_LOAD:.*]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR3_LOAD]], %[[UB3_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR3_LOAD:.*]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[TLA_DECAY:.*]] = cir.cast array_to_ptrdecay %[[TL_ALLOCA]] : !cir.ptr<!cir.array<!cir.array<!cir.array<!rec_NoOps x 5> x 5> x 5>> -> !cir.ptr<!cir.array<!cir.array<!rec_NoOps x 5> x 5>> +// CHECK-NEXT: %[[BOUND3_STRIDE:.*]] = cir.ptr_stride(%[[TLA_DECAY]] : !cir.ptr<!cir.array<!cir.array<!rec_NoOps x 5> x 5>>, %[[ITR3_LOAD]] : !u64i), !cir.ptr<!cir.array<!cir.array<!rec_NoOps x 5> x 5>> +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB2:.*]] = acc.get_lowerbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB2]] : index to !u64i +// CHECK-NEXT: %[[UB2:.*]] = acc.get_upperbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB2]] : index to !u64i +// CHECK-NEXT: %[[ITR2:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB2_CAST]], %[[ITR2]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR2_LOAD:.*]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR2_LOAD]], %[[UB2_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR2_LOAD:.*]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[BOUND3_STRIDE_DECAY:.*]] = cir.cast array_to_ptrdecay %[[BOUND3_STRIDE]] : !cir.ptr<!cir.array<!cir.array<!rec_NoOps x 5> x 5>> -> !cir.ptr<!cir.array<!rec_NoOps x 5>> +// CHECK-NEXT: %[[BOUND2_STRIDE:.*]] = cir.ptr_stride(%[[BOUND3_STRIDE_DECAY]] : !cir.ptr<!cir.array<!rec_NoOps x 5>>, %[[ITR2_LOAD]] : !u64i), !cir.ptr<!cir.array<!rec_NoOps x 5>> +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB1:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB1]] : index to !u64i +// CHECK-NEXT: %[[UB1:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB1]] : index to !u64i +// CHECK-NEXT: %[[ITR1:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB1_CAST]], %[[ITR1]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR1_LOAD:.*]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR1_LOAD]], %[[UB1_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR1_LOAD:.*]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[BOUND2_STRIDE_DECAY:.*]] = cir.cast array_to_ptrdecay %[[BOUND2_STRIDE]] : !cir.ptr<!cir.array<!rec_NoOps x 5>> -> !cir.ptr<!rec_NoOps> +// CHECK-NEXT: %[[BOUND1_STRIDE:.*]] = cir.ptr_stride(%[[BOUND2_STRIDE_DECAY]] : !cir.ptr<!rec_NoOps>, %[[ITR1_LOAD]] : !u64i), !cir.ptr<!rec_NoOps> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[BOUND1_STRIDE]][0] {name = "i"} : !cir.ptr<!rec_NoOps> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR1_LOAD]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR1_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR1]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR2_LOAD]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR2_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR2]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR3_LOAD]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR3_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR3]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[REF:.*]]: !cir.ptr<!cir.array<!cir.array<!cir.array<!rec_NoOps x 5> x 5> x 5>> {{.*}}, %[[PRIVATE:.*]]: !cir.ptr<!cir.array<!cir.array<!cir.array<!rec_NoOps x 5> x 5> x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND2:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND3:.*]]: !acc.data_bounds_ty {{.*}}): +// CHECK-NEXT: acc.yield +// CHECK-NEXT:} destroy { +// CHECK-NEXT: ^bb0(%[[REF:.*]]: !cir.ptr<!cir.array<!cir.array<!cir.array<!rec_NoOps x 5> x 5> x 5>> {{.*}}, %[[PRIVATE:.*]]: !cir.ptr<!cir.array<!cir.array<!cir.array<!rec_NoOps x 5> x 5> x 5>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND2:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND3:.*]]: !acc.data_bounds_ty {{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB3:.*]] = acc.get_lowerbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB3]] : index to !u64i +// CHECK-NEXT: %[[UB3:.*]] = acc.get_upperbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB3]] : index to !u64i +// CHECK-NEXT: %[[ITR3:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB3_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR3]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR3_LOAD:.*]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR3_LOAD]], %[[LB3_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR3_LOAD:.*]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[TLA_DECAY:.*]] = cir.cast array_to_ptrdecay %[[PRIVATE]] : !cir.ptr<!cir.array<!cir.array<!cir.array<!rec_NoOps x 5> x 5> x 5>> -> !cir.ptr<!cir.array<!cir.array<!rec_NoOps x 5> x 5>> +// CHECK-NEXT: %[[BOUND3_STRIDE:.*]] = cir.ptr_stride(%[[TLA_DECAY]] : !cir.ptr<!cir.array<!cir.array<!rec_NoOps x 5> x 5>>, %[[ITR3_LOAD]] : !u64i), !cir.ptr<!cir.array<!cir.array<!rec_NoOps x 5> x 5>> +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB2:.*]] = acc.get_lowerbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB2]] : index to !u64i +// CHECK-NEXT: %[[UB2:.*]] = acc.get_upperbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB2]] : index to !u64i +// CHECK-NEXT: %[[ITR2:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB2_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR2]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR2_LOAD:.*]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR2_LOAD]], %[[LB2_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR2_LOAD:.*]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[BOUND3_STRIDE_DECAY:.*]] = cir.cast array_to_ptrdecay %[[BOUND3_STRIDE]] : !cir.ptr<!cir.array<!cir.array<!rec_NoOps x 5> x 5>> -> !cir.ptr<!cir.array<!rec_NoOps x 5>> +// CHECK-NEXT: %[[BOUND2_STRIDE:.*]] = cir.ptr_stride(%[[BOUND3_STRIDE_DECAY]] : !cir.ptr<!cir.array<!rec_NoOps x 5>>, %[[ITR2_LOAD]] : !u64i), !cir.ptr<!cir.array<!rec_NoOps x 5>> +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB1:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB1]] : index to !u64i +// CHECK-NEXT: %[[UB1:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB1]] : index to !u64i +// CHECK-NEXT: %[[ITR1:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[LAST_SUB_ONE:.*]] = cir.binop(sub, %[[UB1_CAST]], %[[ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[LAST_SUB_ONE]], %[[ITR1]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR1_LOAD:.*]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR1_LOAD]], %[[LB1_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR1_LOAD:.*]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[BOUND2_STRIDE_DECAY:.*]] = cir.cast array_to_ptrdecay %[[BOUND2_STRIDE]] : !cir.ptr<!cir.array<!rec_NoOps x 5>> -> !cir.ptr<!rec_NoOps> +// CHECK-NEXT: %[[BOUND1_STRIDE:.*]] = cir.ptr_stride(%[[BOUND2_STRIDE_DECAY]] : !cir.ptr<!rec_NoOps>, %[[ITR1_LOAD]] : !u64i), !cir.ptr<!rec_NoOps> +// CHECK-NEXT: cir.call @_ZN5NoOpsD1Ev(%[[BOUND1_STRIDE]]) nothrow : (!cir.ptr<!rec_NoOps>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR1_LOAD]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR1_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR1]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR2_LOAD]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR2_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR2]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR3_LOAD]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR3_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR3]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT:acc.yield +// CHECK-NEXT:} + ; + + NoOps ***ThreePtr; +#pragma acc parallel reduction(*:ThreePtr[B][B][A:B]) +// CHECK: acc.reduction.recipe @reduction_mul__Bcnt3__ZTSPPP5NoOps : !cir.ptr<!cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>> reduction_operator <mul> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND2:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND3:.*]]: !acc.data_bounds_ty {{.*}}): +// CHECK-NEXT: %[[TOP_LEVEL_ALLOCA:.*]] = cir.alloca !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>, !cir.ptr<!cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>>, ["openacc.reduction.init"] {alignment = 8 : i64} +// +// CHECK-NEXT: %[[INT_PTR_PTR_PTR_UPPER_BOUND:.*]] = acc.get_upperbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UPPER_BOUND_CAST:.*]] = builtin.unrealized_conversion_cast %[[INT_PTR_PTR_PTR_UPPER_BOUND]] : index to !u64i +// CHECK-NEXT: %[[SIZEOF_PTR:.*]] = cir.const #cir.int<8> : !u64i +// CHECK-NEXT: %[[CALC_ALLOCA_SIZE:.*]] = cir.binop(mul, %[[UPPER_BOUND_CAST]], %[[SIZEOF_PTR]]) : !u64i +// CHECK-NEXT: %[[INT_PTR_PTR_VLA_ALLOCA:.*]] = cir.alloca !cir.ptr<!cir.ptr<!rec_NoOps>>, !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>, %[[CALC_ALLOCA_SIZE]] : !u64i, ["openacc.init.bounds"] {alignment = 8 : i64} +// +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["itr"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u64i +// CHECK-NEXT: cir.store %[[ZERO]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[UPPER_LIMIT:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[CMP:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UPPER_LIMIT]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[CMP]]) +// +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[SRC_IDX:.*]] = cir.binop(mul, %[[UPPER_BOUND_CAST]], %[[ITR_LOAD]]) : !u64i +// CHECK-NEXT: %[[SRC_STRIDE:.*]] = cir.ptr_stride(%[[INT_PTR_PTR_VLA_ALLOCA]] : !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>, %[[SRC_IDX]] : !u64i), !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>> +// CHECK-NEXT: %[[DEST_STRIDE:.*]] = cir.ptr_stride(%[[TOP_LEVEL_ALLOCA]] : !cir.ptr<!cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>> +// CHECK-NEXT: cir.store %[[SRC_STRIDE]], %[[DEST_STRIDE]] : !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>, !cir.ptr<!cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>> +// CHECK-NEXT: cir.yield +// +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// +// CHECK-NEXT: %[[INT_PTR_PTR_UPPER_BOUND:.*]] = acc.get_upperbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UPPER_BOUND_CAST_2:.*]] = builtin.unrealized_conversion_cast %[[INT_PTR_PTR_UPPER_BOUND]] : index to !u64i +// CHECK-NEXT: %[[NUM_ELTS:.*]] = cir.binop(mul, %[[UPPER_BOUND_CAST_2]], %[[UPPER_BOUND_CAST]]) : !u64i +// CHECK-NEXT: %[[SIZEOF_PTR_PTR:.*]] = cir.const #cir.int<8> : !u64i +// CHECK-NEXT: %[[CALC_ALLOCA_SIZE:.*]] = cir.binop(mul, %[[NUM_ELTS]], %[[SIZEOF_PTR_PTR]]) : !u64i +// CHECK-NEXT: %[[INT_PTR_PTR_ALLOCA:.*]] = cir.alloca !cir.ptr<!rec_NoOps>, !cir.ptr<!cir.ptr<!rec_NoOps>>, %[[CALC_ALLOCA_SIZE]] : !u64i, ["openacc.init.bounds"] {alignment = 8 : i64} +// +// +// Copy array pointer to the original alloca. +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["itr"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u64i +// CHECK-NEXT: cir.store %[[ZERO]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[CMP:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UPPER_BOUND_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[CMP]]) +// +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[SRC_IDX:.*]] = cir.binop(mul, %[[UPPER_BOUND_CAST_2]], %[[ITR_LOAD]]) : !u64i +// CHECK-NEXT: %[[SRC_STRIDE:.*]] = cir.ptr_stride(%[[INT_PTR_PTR_ALLOCA]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, %[[SRC_IDX]] : !u64i), !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: %[[DEST_STRIDE:.*]] = cir.ptr_stride(%[[INT_PTR_PTR_VLA_ALLOCA]] : !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>> +// CHECK-NEXT: cir.store %[[SRC_STRIDE]], %[[DEST_STRIDE]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>> +// CHECK-NEXT: cir.yield +// +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// +// CHECK-NEXT: %[[INT_PTR_UPPER_BOUND:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UPPER_BOUND_CAST_3:.*]] = builtin.unrealized_conversion_cast %[[INT_PTR_UPPER_BOUND]] : index to !u64i +// CHECK-NEXT: %[[NUM_ELTS_2:.*]] = cir.binop(mul, %[[UPPER_BOUND_CAST_3]], %[[NUM_ELTS]]) : !u64i +// CHECK-NEXT: %[[SIZEOF_INT:.*]] = cir.const #cir.int<4> : !u64i +// CHECK-NEXT: %[[CALC_ALLOCA_SIZE:.*]] = cir.binop(mul, %[[NUM_ELTS_2]], %[[SIZEOF_INT]]) : !u64i +// CHECK-NEXT: %[[INT_PTR_ALLOCA:.*]] = cir.alloca !rec_NoOps, !cir.ptr<!rec_NoOps>, %[[CALC_ALLOCA_SIZE]] : !u64i, ["openacc.init.bounds"] {alignment = 4 : i64} +// +// Copy array pointer to the original alloca. +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["itr"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u64i +// CHECK-NEXT: cir.store %[[ZERO]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[CMP:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[NUM_ELTS]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[CMP]]) +// +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[SRC_IDX:.*]] = cir.binop(mul, %[[UPPER_BOUND_CAST_3]], %[[ITR_LOAD]]) : !u64i +// CHECK-NEXT: %[[SRC_STRIDE:.*]] = cir.ptr_stride(%[[INT_PTR_ALLOCA]] : !cir.ptr<!rec_NoOps>, %[[SRC_IDX]] : !u64i), !cir.ptr<!rec_NoOps> +// CHECK-NEXT: %[[DEST_STRIDE:.*]] = cir.ptr_stride(%[[INT_PTR_PTR_ALLOCA]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: cir.store %[[SRC_STRIDE]], %[[DEST_STRIDE]] : !cir.ptr<!rec_NoOps>, !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: cir.yield +// +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// Initialization Section +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB3:.*]] = acc.get_lowerbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB3]] : index to !u64i +// CHECK-NEXT: %[[UB3:.*]] = acc.get_upperbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB3]] : index to !u64i +// CHECK-NEXT: %[[ITR3:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB3_CAST]], %[[ITR3]] : !u64i, !cir.ptr<!u64i> + +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR3_LOAD:.*]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR3_LOAD]], %[[UB3_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR3_LOAD:.*]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[TLA_LOAD:.*]] = cir.load %[[TOP_LEVEL_ALLOCA]] : !cir.ptr<!cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>>, !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>> +// CHECK-NEXT: %[[BOUND3_STRIDE:.*]] = cir.ptr_stride(%[[TLA_LOAD]] : !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>, %[[ITR3_LOAD]] : !u64i), !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>> +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB2:.*]] = acc.get_lowerbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB2]] : index to !u64i +// CHECK-NEXT: %[[UB2:.*]] = acc.get_upperbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB2]] : index to !u64i +// CHECK-NEXT: %[[ITR2:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB2_CAST]], %[[ITR2]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR2_LOAD:.*]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR2_LOAD]], %[[UB2_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR2_LOAD:.*]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[BOUND3_STRIDE_LOAD:.*]] = cir.load %[[BOUND3_STRIDE]] : !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>, !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: %[[BOUND2_STRIDE:.*]] = cir.ptr_stride(%[[BOUND3_STRIDE_LOAD]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, %[[ITR2_LOAD]] : !u64i), !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB1:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB1]] : index to !u64i +// CHECK-NEXT: %[[UB1:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB1]] : index to !u64i +// CHECK-NEXT: %[[ITR1:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB1_CAST]], %[[ITR1]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR1_LOAD:.*]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR1_LOAD]], %[[UB1_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR1_LOAD:.*]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[BOUND2_STRIDE_LOAD:.*]] = cir.load %[[BOUND2_STRIDE]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, !cir.ptr<!rec_NoOps> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[BOUND2_STRIDE_LOAD]] : !cir.ptr<!rec_NoOps>, %[[ITR1_LOAD]] : !u64i), !cir.ptr<!rec_NoOps> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_NoOps> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ONE:.*]] = cir.const #cir.int<1> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ONE]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR1_LOAD]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR1_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR1]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR2_LOAD]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR2_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR2]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR3_LOAD]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR3_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR3]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[REF:.*]]: !cir.ptr<!cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>> {{.*}}, %[[PRIVATE:.*]]: !cir.ptr<!cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND2:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND3:.*]]: !acc.data_bounds_ty {{.*}}): +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[REF:.*]]: !cir.ptr<!cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>> {{.*}}, %[[PRIVATE:.*]]: !cir.ptr<!cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND2:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND3:.*]]: !acc.data_bounds_ty {{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB3:.*]] = acc.get_lowerbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB3]] : index to !u64i +// CHECK-NEXT: %[[UB3:.*]] = acc.get_upperbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB3]] : index to !u64i +// CHECK-NEXT: %[[ITR3:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[CONST_ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[ONE_BELOW_UB3:.*]] = cir.binop(sub, %[[UB3_CAST]], %[[CONST_ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[ONE_BELOW_UB3]], %[[ITR3]] : !u64i, !cir.ptr<!u64i> + +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR3_LOAD:.*]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR3_LOAD]], %[[LB3_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR3_LOAD:.*]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[TLA_LOAD:.*]] = cir.load %[[PRIVATE]] : !cir.ptr<!cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>>, !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>> +// CHECK-NEXT: %[[BOUND3_STRIDE:.*]] = cir.ptr_stride(%[[TLA_LOAD]] : !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>, %[[ITR3_LOAD]] : !u64i), !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>> +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB2:.*]] = acc.get_lowerbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB2]] : index to !u64i +// CHECK-NEXT: %[[UB2:.*]] = acc.get_upperbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB2]] : index to !u64i +// CHECK-NEXT: %[[ITR2:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[CONST_ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[ONE_BELOW_UB2:.*]] = cir.binop(sub, %[[UB2_CAST]], %[[CONST_ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[ONE_BELOW_UB2]], %[[ITR2]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR2_LOAD:.*]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR2_LOAD]], %[[LB2_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR2_LOAD:.*]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[BOUND3_STRIDE_LOAD:.*]] = cir.load %[[BOUND3_STRIDE]] : !cir.ptr<!cir.ptr<!cir.ptr<!rec_NoOps>>>, !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: %[[BOUND2_STRIDE:.*]] = cir.ptr_stride(%[[BOUND3_STRIDE_LOAD]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, %[[ITR2_LOAD]] : !u64i), !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB1:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB1]] : index to !u64i +// CHECK-NEXT: %[[UB1:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB1]] : index to !u64i +// CHECK-NEXT: %[[ITR1:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[CONST_ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[ONE_BELOW_UB1:.*]] = cir.binop(sub, %[[UB1_CAST]], %[[CONST_ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[ONE_BELOW_UB1]], %[[ITR1]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR1_LOAD:.*]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR1_LOAD]], %[[LB1_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR1_LOAD:.*]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[BOUND2_STRIDE_LOAD:.*]] = cir.load %[[BOUND2_STRIDE]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, !cir.ptr<!rec_NoOps> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[BOUND2_STRIDE_LOAD]] : !cir.ptr<!rec_NoOps>, %[[ITR1_LOAD]] : !u64i), !cir.ptr<!rec_NoOps> +// CHECK-NEXT: cir.call @_ZN5NoOpsD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_NoOps>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR1_LOAD]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR1_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR1]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR2_LOAD]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR2_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR2]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR3_LOAD]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR3_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR3]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } +; + using PtrTArrayTy = NoOps*[5]; + PtrTArrayTy *PtrArrayPtr; + +#pragma acc parallel reduction(||:PtrArrayPtr[B][B][B]) +// CHECK-NEXT: acc.reduction.recipe @reduction_lor__Bcnt3__ZTSPA5_P5NoOps : !cir.ptr<!cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>> reduction_operator <lor> init { +// CHECK-NEXT: ^bb0(%[[ARG:.*]]: !cir.ptr<!cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND2:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND3:.*]]: !acc.data_bounds_ty {{.*}}): +// CHECK-NEXT: %[[TL_ALLOCA:.*]] = cir.alloca !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>, !cir.ptr<!cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>>, ["openacc.reduction.init"] {alignment = 8 : i64} +// +// CHECK-NEXT: %[[UB3:.*]] = acc.get_upperbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB3]] : index to !u64i +// CHECK-NEXT: %[[ARR_SIZE:.*]] = cir.const #cir.int<40> : !u64i +// CHECK-NEXT: %[[ALLOCA_SIZE:.*]] = cir.binop(mul, %[[UB3_CAST]], %[[ARR_SIZE]]) : !u64i +// CHECK-NEXT: %[[ARR_ALLOCA:.*]] = cir.alloca !cir.array<!cir.ptr<!rec_NoOps> x 5>, !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>, %[[ALLOCA_SIZE]] : !u64i, ["openacc.init.bounds"] {alignment = 8 : i64} +// +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["itr"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u64i +// CHECK-NEXT: cir.store %[[ZERO]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[UPP_BOUND:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[CMP:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[UPP_BOUND]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[CMP]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[SRC_IDX:.*]] = cir.binop(mul, %[[UB3_CAST]], %[[ITR_LOAD]]) : !u64i +// CHECK-NEXT: %[[SRC:.*]] = cir.ptr_stride(%[[ARR_ALLOCA]] : !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>, %[[SRC_IDX]] : !u64i), !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>> +// CHECK-NEXT: %[[DEST:.*]] = cir.ptr_stride(%[[TL_ALLOCA]] : !cir.ptr<!cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>> +// CHECK-NEXT: cir.store %[[SRC]], %[[DEST]] : !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>, !cir.ptr<!cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// +// CHECK-NEXT: %[[UB2:.*]] = acc.get_upperbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB2]] : index to !u64i +// CHECK-NEXT: %[[NUM_ELTS:.*]] = cir.binop(mul, %[[UB2_CAST]], %[[UB3_CAST]]) : !u64i +// +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u64i +// CHECK-NEXT: %[[DECAY:.*]] = cir.cast array_to_ptrdecay %[[ARR_ALLOCA]] : !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>> -> !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[DECAY]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, %[[ZERO]] : !u64i), !cir.ptr<!cir.ptr<!rec_NoOps>> +// +// CHECK-NEXT: %[[UB1:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB1]] : index to !u64i +// CHECK-NEXT: %[[NUM_ELTS2:.*]] = cir.binop(mul, %[[UB1_CAST]], %[[NUM_ELTS]]) : !u64i +// CHECK-NEXT: %[[ELT_SIZE:.*]] = cir.const #cir.int<4> : !u64i +// CHECK-NEXT: %[[ALLOCA_SIZE:.*]] = cir.binop(mul, %[[NUM_ELTS2]], %[[ELT_SIZE]]) : !u64i +// CHECK-NEXT: %[[ARR_ALLOCA2:.*]] = cir.alloca !rec_NoOps, !cir.ptr<!rec_NoOps>, %[[ALLOCA_SIZE]] : !u64i, ["openacc.init.bounds"] {alignment = 4 : i64} +// +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[ITR:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["itr"] {alignment = 8 : i64} +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !u64i +// CHECK-NEXT: cir.store %[[ZERO]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[CMP:.*]] = cir.cmp(lt, %[[ITR_LOAD]], %[[NUM_ELTS]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[CMP]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[SRC_IDX:.*]] = cir.binop(mul, %[[UB1_CAST]], %[[ITR_LOAD]]) : !u64i +// CHECK-NEXT: %[[SRC:.*]] = cir.ptr_stride(%[[ARR_ALLOCA2]] : !cir.ptr<!rec_NoOps>, %[[SRC_IDX]] : !u64i), !cir.ptr<!rec_NoOps> +// CHECK-NEXT: %[[DEST:.*]] = cir.ptr_stride(%[[STRIDE]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, %[[ITR_LOAD]] : !u64i), !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: cir.store %[[SRC]], %[[DEST]] : !cir.ptr<!rec_NoOps>, !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR_LOAD:.*]] = cir.load %[[ITR]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// Init Section: +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB3:.*]] = acc.get_lowerbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB3]] : index to !u64i +// CHECK-NEXT: %[[UB3:.*]] = acc.get_upperbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB3]] : index to !u64i +// CHECK-NEXT: %[[ITR3:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB3_CAST]], %[[ITR3]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR3_LOAD:.*]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[CMP:.*]] = cir.cmp(lt, %[[ITR3_LOAD]], %[[UB3_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[CMP]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR3_LOAD:.*]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[TLA_LOAD:.*]] = cir.load %[[TL_ALLOCA]] : !cir.ptr<!cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>>, !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>> +// CHECK-NEXT: %[[BOUND3_STRIDE:.*]] = cir.ptr_stride(%[[TLA_LOAD]] : !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>, %[[ITR3_LOAD]] : !u64i), !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>> +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB2:.*]] = acc.get_lowerbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB2]] : index to !u64i +// CHECK-NEXT: %[[UB2:.*]] = acc.get_upperbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB2]] : index to !u64i +// CHECK-NEXT: %[[ITR2:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB2_CAST]], %[[ITR2]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR2_LOAD:.*]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR2_LOAD]], %[[UB2_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR2_LOAD:.*]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[BOUND3_STRIDE_DECAY:.*]] = cir.cast array_to_ptrdecay %[[BOUND3_STRIDE]] : !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>> -> !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: %[[BOUND2_STRIDE:.*]] = cir.ptr_stride(%[[BOUND3_STRIDE_DECAY]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, %[[ITR2_LOAD]] : !u64i), !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB1:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB1]] : index to !u64i +// CHECK-NEXT: %[[UB1:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB1]] : index to !u64i +// CHECK-NEXT: %[[ITR1:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: cir.store %[[LB1_CAST]], %[[ITR1]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR1_LOAD:.*]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(lt, %[[ITR1_LOAD]], %[[UB1_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR1_LOAD:.*]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[BOUND2_STRIDE_LOAD:.*]] = cir.load %[[BOUND2_STRIDE]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, !cir.ptr<!rec_NoOps> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[BOUND2_STRIDE_LOAD]] : !cir.ptr<!rec_NoOps>, %[[ITR1_LOAD]] : !u64i), !cir.ptr<!rec_NoOps> +// CHECK-NEXT: %[[GET_I:.*]] = cir.get_member %[[STRIDE]][0] {name = "i"} : !cir.ptr<!rec_NoOps> -> !cir.ptr<!s32i> +// CHECK-NEXT: %[[ZERO:.*]] = cir.const #cir.int<0> : !s32i +// CHECK-NEXT: cir.store{{.*}} %[[ZERO]], %[[GET_I]] : !s32i, !cir.ptr<!s32i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR1_LOAD]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR1_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR1]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR2_LOAD]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR2_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR2]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR3_LOAD]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[INC:.*]] = cir.unary(inc, %[[ITR3_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[INC]], %[[ITR3]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } combiner { +// CHECK-NEXT: ^bb0(%[[REF:.*]]: !cir.ptr<!cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>> {{.*}}, %[[PRIVATE:.*]]: !cir.ptr<!cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND2:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND3:.*]]: !acc.data_bounds_ty {{.*}}): +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } destroy { +// CHECK-NEXT: ^bb0(%[[REF:.*]]: !cir.ptr<!cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>> {{.*}}, %[[PRIVATE:.*]]: !cir.ptr<!cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>> {{.*}}, %[[BOUND1:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND2:.*]]: !acc.data_bounds_ty {{.*}}, %[[BOUND3:.*]]: !acc.data_bounds_ty {{.*}}): +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB3:.*]] = acc.get_lowerbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB3]] : index to !u64i +// CHECK-NEXT: %[[UB3:.*]] = acc.get_upperbound %[[BOUND3]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB3_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB3]] : index to !u64i +// CHECK-NEXT: %[[ITR3:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[CONST_ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[ONE_BELOW_UB3:.*]] = cir.binop(sub, %[[UB3_CAST]], %[[CONST_ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[ONE_BELOW_UB3]], %[[ITR3]] : !u64i, !cir.ptr<!u64i> +// +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR3_LOAD:.*]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR3_LOAD]], %[[LB3_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR3_LOAD:.*]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[TLA_LOAD:.*]] = cir.load %[[PRIVATE]] : !cir.ptr<!cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>>, !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>> +// CHECK-NEXT: %[[BOUND3_STRIDE:.*]] = cir.ptr_stride(%[[TLA_LOAD]] : !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>>, %[[ITR3_LOAD]] : !u64i), !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>> +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB2:.*]] = acc.get_lowerbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB2]] : index to !u64i +// CHECK-NEXT: %[[UB2:.*]] = acc.get_upperbound %[[BOUND2]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB2_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB2]] : index to !u64i +// CHECK-NEXT: %[[ITR2:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[CONST_ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[ONE_BELOW_UB2:.*]] = cir.binop(sub, %[[UB2_CAST]], %[[CONST_ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[ONE_BELOW_UB2]], %[[ITR2]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR2_LOAD:.*]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR2_LOAD]], %[[LB2_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR2_LOAD:.*]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[BOUND3_STRIDE_DECAY:.*]] = cir.cast array_to_ptrdecay %[[BOUND3_STRIDE]] : !cir.ptr<!cir.array<!cir.ptr<!rec_NoOps> x 5>> -> !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: %[[BOUND2_STRIDE:.*]] = cir.ptr_stride(%[[BOUND3_STRIDE_DECAY]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, %[[ITR2_LOAD]] : !u64i), !cir.ptr<!cir.ptr<!rec_NoOps>> +// CHECK-NEXT: cir.scope { +// CHECK-NEXT: %[[LB1:.*]] = acc.get_lowerbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[LB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[LB1]] : index to !u64i +// CHECK-NEXT: %[[UB1:.*]] = acc.get_upperbound %[[BOUND1]] : (!acc.data_bounds_ty) -> index +// CHECK-NEXT: %[[UB1_CAST:.*]] = builtin.unrealized_conversion_cast %[[UB1]] : index to !u64i +// CHECK-NEXT: %[[ITR1:.*]] = cir.alloca !u64i, !cir.ptr<!u64i>, ["iter"] {alignment = 8 : i64} +// CHECK-NEXT: %[[CONST_ONE:.*]] = cir.const #cir.int<1> : !u64i +// CHECK-NEXT: %[[ONE_BELOW_UB1:.*]] = cir.binop(sub, %[[UB1_CAST]], %[[CONST_ONE]]) : !u64i +// CHECK-NEXT: cir.store %[[ONE_BELOW_UB1]], %[[ITR1]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.for : cond { +// CHECK-NEXT: %[[ITR1_LOAD:.*]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[COND:.*]] = cir.cmp(ge, %[[ITR1_LOAD]], %[[LB1_CAST]]) : !u64i, !cir.bool +// CHECK-NEXT: cir.condition(%[[COND]]) +// CHECK-NEXT: } body { +// CHECK-NEXT: %[[ITR1_LOAD:.*]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[BOUND2_STRIDE_LOAD:.*]] = cir.load %[[BOUND2_STRIDE]] : !cir.ptr<!cir.ptr<!rec_NoOps>>, !cir.ptr<!rec_NoOps> +// CHECK-NEXT: %[[STRIDE:.*]] = cir.ptr_stride(%[[BOUND2_STRIDE_LOAD]] : !cir.ptr<!rec_NoOps>, %[[ITR1_LOAD]] : !u64i), !cir.ptr<!rec_NoOps> +// CHECK-NEXT: cir.call @_ZN5NoOpsD1Ev(%[[STRIDE]]) nothrow : (!cir.ptr<!rec_NoOps>) -> () +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR1_LOAD]] = cir.load %[[ITR1]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR1_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR1]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR2_LOAD]] = cir.load %[[ITR2]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR2_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR2]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } step { +// CHECK-NEXT: %[[ITR3_LOAD]] = cir.load %[[ITR3]] : !cir.ptr<!u64i>, !u64i +// CHECK-NEXT: %[[DEC:.*]] = cir.unary(dec, %[[ITR3_LOAD]]) : !u64i, !u64i +// CHECK-NEXT: cir.store %[[DEC]], %[[ITR3]] : !u64i, !cir.ptr<!u64i> +// CHECK-NEXT: cir.yield +// CHECK-NEXT: } +// CHECK-NEXT: } +// CHECK-NEXT: acc.yield +// CHECK-NEXT: } + ; +} diff --git a/clang/test/CIR/IR/func.cir b/clang/test/CIR/IR/func.cir index 9532859..d7e8184 100644 --- a/clang/test/CIR/IR/func.cir +++ b/clang/test/CIR/IR/func.cir @@ -99,4 +99,15 @@ cir.func @ullfunc() -> !u64i { // CHECK: %[[VAL:.*]] = cir.const #cir.int<42> : !u64i // CHECK: cir.return %[[VAL:.*]] : !u64i // CHECK: } + +cir.func coroutine @coro() { + cir.return +} +// CHECK: cir.func{{.*}} coroutine @coro() + +cir.func builtin @builtin() { + cir.return +} +// CHECK: cir.func{{.*}} builtin @builtin() + } diff --git a/clang/test/CIR/IR/invalid-addrspace.cir b/clang/test/CIR/IR/invalid-addrspace.cir new file mode 100644 index 0000000..8f188b8 --- /dev/null +++ b/clang/test/CIR/IR/invalid-addrspace.cir @@ -0,0 +1,27 @@ +// RUN: cir-opt %s -verify-diagnostics -split-input-file + +// ----- + +!u64i = !cir.int<u, 64> +// expected-error @below {{expected 'target_address_space'}} +cir.func @address_space1(%p : !cir.ptr<!u64i, foobar>) { + cir.return +} + +// ----- + +!u64i = !cir.int<u, 64> +// expected-error@below {{expected '('}} +cir.func @address_space2(%p : !cir.ptr<!u64i, target_address_space>) { + cir.return +} + +// ----- + +!u64i = !cir.int<u, 64> +// expected-error@+2 {{expected integer value}} +// expected-error@below {{expected integer address space value}} +cir.func @address_space3(%p : !cir.ptr<!u64i, target_address_space()>) { + cir.return +} + diff --git a/clang/test/CodeGen/X86/avx-builtins.c b/clang/test/CodeGen/X86/avx-builtins.c index 3018bb97..5f08b6b 100644 --- a/clang/test/CodeGen/X86/avx-builtins.c +++ b/clang/test/CodeGen/X86/avx-builtins.c @@ -1039,6 +1039,7 @@ int test_mm256_extract_epi8(__m256i A) { // CHECK: zext i8 %{{.*}} to i32 return _mm256_extract_epi8(A, 31); } +TEST_CONSTEXPR(_mm256_extract_epi8(((__m256i)(__v32qs){0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31}), 45) == 13); int test_mm256_extract_epi16(__m256i A) { // CHECK-LABEL: test_mm256_extract_epi16 @@ -1046,12 +1047,14 @@ int test_mm256_extract_epi16(__m256i A) { // CHECK: zext i16 %{{.*}} to i32 return _mm256_extract_epi16(A, 15); } +TEST_CONSTEXPR(_mm256_extract_epi16(((__m256i)(__v16hi){0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30}), 50) == 4); int test_mm256_extract_epi32(__m256i A) { // CHECK-LABEL: test_mm256_extract_epi32 // CHECK: extractelement <8 x i32> %{{.*}}, {{i32|i64}} 7 return _mm256_extract_epi32(A, 7); } +TEST_CONSTEXPR(_mm256_extract_epi32(((__m256i)(__v8si){0, 5, 10, 15, 20, 25, 30, 35}), 18) == 10); #if __x86_64__ long long test_mm256_extract_epi64(__m256i A) { @@ -1059,6 +1062,7 @@ long long test_mm256_extract_epi64(__m256i A) { // X64: extractelement <4 x i64> %{{.*}}, {{i32|i64}} 3 return _mm256_extract_epi64(A, 3); } +TEST_CONSTEXPR(_mm256_extract_epi64(((__m256i)(__v4di){5, 15, 25, 35}), 14) == 25); #endif __m128d test_mm256_extractf128_pd(__m256d A) { @@ -1120,18 +1124,21 @@ __m256i test_mm256_insert_epi8(__m256i x, char b) { // CHECK: insertelement <32 x i8> %{{.*}}, i8 %{{.*}}, {{i32|i64}} 14 return _mm256_insert_epi8(x, b, 14); } +TEST_CONSTEXPR(match_v32qi(_mm256_insert_epi8(((__m256i)(__v32qs){0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31}), 77, 47), 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 77, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31)); __m256i test_mm256_insert_epi16(__m256i x, int b) { // CHECK-LABEL: test_mm256_insert_epi16 // CHECK: insertelement <16 x i16> %{{.*}}, i16 %{{.*}}, {{i32|i64}} 4 return _mm256_insert_epi16(x, b, 4); } +TEST_CONSTEXPR(match_v16hi(_mm256_insert_epi16(((__m256i)(__v16hi){0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30}), 909, 62), 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 909, 30)); __m256i test_mm256_insert_epi32(__m256i x, int b) { // CHECK-LABEL: test_mm256_insert_epi32 // CHECK: insertelement <8 x i32> %{{.*}}, i32 %{{.*}}, {{i32|i64}} 5 return _mm256_insert_epi32(x, b, 5); } +TEST_CONSTEXPR(match_v8si(_mm256_insert_epi32(((__m256i)(__v8si){ 0, 5, 10, 15, 20, 25, 30, 35}), 4321, 18), 0, 5, 4321, 15, 20, 25, 30, 35)); #if __x86_64__ __m256i test_mm256_insert_epi64(__m256i x, long long b) { @@ -1139,6 +1146,7 @@ __m256i test_mm256_insert_epi64(__m256i x, long long b) { // X64: insertelement <4 x i64> %{{.*}}, i64 %{{.*}}, {{i32|i64}} 2 return _mm256_insert_epi64(x, b, 2); } +TEST_CONSTEXPR(match_v4di(_mm256_insert_epi64(((__m256i)(__v4di){5, 15, 25, 35}), -123456789LL, 10), 5, 15, -123456789LL, 35)); #endif __m256d test_mm256_insertf128_pd(__m256d A, __m128d B) { diff --git a/clang/test/CodeGen/X86/avx2-builtins.c b/clang/test/CodeGen/X86/avx2-builtins.c index eff2797..4299b18 100644 --- a/clang/test/CodeGen/X86/avx2-builtins.c +++ b/clang/test/CodeGen/X86/avx2-builtins.c @@ -1109,19 +1109,19 @@ __m256i test_mm256_shuffle_epi32(__m256i a) { // CHECK: shufflevector <8 x i32> %{{.*}}, <8 x i32> poison, <8 x i32> <i32 3, i32 3, i32 0, i32 0, i32 7, i32 7, i32 4, i32 4> return _mm256_shuffle_epi32(a, 15); } - +TEST_CONSTEXPR(match_v8si(_mm256_shuffle_epi32((((__m256i)(__v8si){0,1,2,3,4,5,6,7})), 15), 3,3,0,0, 7,7,4,4)); __m256i test_mm256_shufflehi_epi16(__m256i a) { // CHECK-LABEL: test_mm256_shufflehi_epi16 // CHECK: shufflevector <16 x i16> %{{.*}}, <16 x i16> poison, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 7, i32 6, i32 6, i32 5, i32 8, i32 9, i32 10, i32 11, i32 15, i32 14, i32 14, i32 13> return _mm256_shufflehi_epi16(a, 107); } - +TEST_CONSTEXPR(match_v16hi(_mm256_shufflehi_epi16((((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15})), 107), 0,1,2,3, 7,6,6,5, 8,9,10,11, 15,14,14,13)); __m256i test_mm256_shufflelo_epi16(__m256i a) { // CHECK-LABEL: test_mm256_shufflelo_epi16 // CHECK: shufflevector <16 x i16> %{{.*}}, <16 x i16> poison, <16 x i32> <i32 3, i32 0, i32 1, i32 1, i32 4, i32 5, i32 6, i32 7, i32 11, i32 8, i32 9, i32 9, i32 12, i32 13, i32 14, i32 15> return _mm256_shufflelo_epi16(a, 83); } - +TEST_CONSTEXPR(match_v16hi(_mm256_shufflelo_epi16(((__m256i)(__v16hi){ 0,1,2,3, 4,5,6,7, 8,9,10,11, 12,13,14,15}), 83), 3,0,1,1, 4,5,6,7, 11,8,9,9, 12,13,14,15) ); __m256i test_mm256_sign_epi8(__m256i a, __m256i b) { // CHECK-LABEL: test_mm256_sign_epi8 // CHECK: call <32 x i8> @llvm.x86.avx2.psign.b(<32 x i8> %{{.*}}, <32 x i8> %{{.*}}) diff --git a/clang/test/CodeGen/X86/avx512bw-builtins.c b/clang/test/CodeGen/X86/avx512bw-builtins.c index 3f42ac0..bd19363 100644 --- a/clang/test/CodeGen/X86/avx512bw-builtins.c +++ b/clang/test/CodeGen/X86/avx512bw-builtins.c @@ -1876,13 +1876,15 @@ __m512i test_mm512_shufflehi_epi16(__m512i __A) { // CHECK: shufflevector <32 x i16> %{{.*}}, <32 x i16> poison, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 5, i32 5, i32 4, i32 4, i32 8, i32 9, i32 10, i32 11, i32 13, i32 13, i32 12, i32 12, i32 16, i32 17, i32 18, i32 19, i32 21, i32 21, i32 20, i32 20, i32 24, i32 25, i32 26, i32 27, i32 29, i32 29, i32 28, i32 28> return _mm512_shufflehi_epi16(__A, 5); } - +TEST_CONSTEXPR(match_v32hi(_mm512_shufflehi_epi16((((__m512i)(__v32hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31})), 5), 0,1,2,3, 5,5,4,4, 8,9,10,11, 13,13,12,12, 16,17,18,19, 21,21,20,20, 24,25,26,27, 29,29,28,28)); __m512i test_mm512_mask_shufflehi_epi16(__m512i __W, __mmask32 __U, __m512i __A) { // CHECK-LABEL: test_mm512_mask_shufflehi_epi16 // CHECK: shufflevector <32 x i16> %{{.*}}, <32 x i16> poison, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 5, i32 5, i32 4, i32 4, i32 8, i32 9, i32 10, i32 11, i32 13, i32 13, i32 12, i32 12, i32 16, i32 17, i32 18, i32 19, i32 21, i32 21, i32 20, i32 20, i32 24, i32 25, i32 26, i32 27, i32 29, i32 29, i32 28, i32 28> // CHECK: select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> %{{.*}} return _mm512_mask_shufflehi_epi16(__W, __U, __A, 5); } +TEST_CONSTEXPR(match_v32hi(_mm512_mask_shufflehi_epi16((((__m512i)(__v32hi){100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131})), 0xFFFF0000u, (((__m512i)(__v32hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31})), 5), 100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115, 16,17,18,19,21,21,20,20, 24,25,26,27,29,29,28,28)); +TEST_CONSTEXPR(match_v32hi(_mm512_mask_shufflehi_epi16(((__m512i)(__v32hi){100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131}), 0x0000FFFFu, ((__m512i)(__v32hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31}), 5), 0,1,2,3,5,5,4,4, 8,9,10,11,13,13,12,12, 116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131)); __m512i test_mm512_maskz_shufflehi_epi16(__mmask32 __U, __m512i __A) { // CHECK-LABEL: test_mm512_maskz_shufflehi_epi16 @@ -1890,12 +1892,15 @@ __m512i test_mm512_maskz_shufflehi_epi16(__mmask32 __U, __m512i __A) { // CHECK: select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> %{{.*}} return _mm512_maskz_shufflehi_epi16(__U, __A, 5); } +TEST_CONSTEXPR(match_v32hi(_mm512_maskz_shufflehi_epi16(0xAAAAAAAAu, (((__m512i)(__v32hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31})), 5), 0,1,0,3,0,5,0,4, 0,9,0,11,0,13,0,12, 0,17,0,19,0,21,0,20, 0,25,0,27,0,29,0,28)); +TEST_CONSTEXPR(match_v32hi(_mm512_maskz_shufflehi_epi16(0x0000FFFFu, ((__m512i)(__v32hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31}), 5), 0,1,2,3,5,5,4,4, 8,9,10,11,13,13,12,12, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0)); __m512i test_mm512_shufflelo_epi16(__m512i __A) { // CHECK-LABEL: test_mm512_shufflelo_epi16 // CHECK: shufflevector <32 x i16> %{{.*}}, <32 x i16> poison, <32 x i32> <i32 1, i32 1, i32 0, i32 0, i32 4, i32 5, i32 6, i32 7, i32 9, i32 9, i32 8, i32 8, i32 12, i32 13, i32 14, i32 15, i32 17, i32 17, i32 16, i32 16, i32 20, i32 21, i32 22, i32 23, i32 25, i32 25, i32 24, i32 24, i32 28, i32 29, i32 30, i32 31> return _mm512_shufflelo_epi16(__A, 5); } +TEST_CONSTEXPR( match_v32hi(_mm512_shufflelo_epi16(((__m512i)(__v32hi){ 0,1,2,3, 4,5,6,7, 8,9,10,11, 12,13,14,15, 16,17,18,19, 20,21,22,23, 24,25,26,27, 28,29,30,31}), 5), 1,1,0,0, 4,5,6,7, 9,9,8,8, 12,13,14,15, 17,17,16,16, 20,21,22,23, 25,25,24,24, 28,29,30,31)); __m512i test_mm512_mask_shufflelo_epi16(__m512i __W, __mmask32 __U, __m512i __A) { // CHECK-LABEL: test_mm512_mask_shufflelo_epi16 @@ -1903,6 +1908,8 @@ __m512i test_mm512_mask_shufflelo_epi16(__m512i __W, __mmask32 __U, __m512i __A) // CHECK: select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> %{{.*}} return _mm512_mask_shufflelo_epi16(__W, __U, __A, 5); } +TEST_CONSTEXPR(match_v32hi(_mm512_mask_shufflelo_epi16((((__m512i)(__v32hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31})), 0xFFFFFFFF, (((__m512i)(__v32hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31})), 5), 1,1,0,0, 4,5,6,7, 9,9,8,8, 12,13,14,15, 17,17,16,16, 20,21,22,23, 25,25,24,24, 28,29,30,31)); +TEST_CONSTEXPR(match_v32hi(_mm512_mask_shufflelo_epi16(((__m512i)(__v32hi){100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131}), 0x0000FFFFu, ((__m512i)(__v32hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31}), 5), 1,1,0,0,4,5,6,7, 9,9,8,8,12,13,14,15, 116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131)); __m512i test_mm512_maskz_shufflelo_epi16(__mmask32 __U, __m512i __A) { // CHECK-LABEL: test_mm512_maskz_shufflelo_epi16 @@ -1910,6 +1917,8 @@ __m512i test_mm512_maskz_shufflelo_epi16(__mmask32 __U, __m512i __A) { // CHECK: select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> %{{.*}} return _mm512_maskz_shufflelo_epi16(__U, __A, 5); } +TEST_CONSTEXPR(match_v32hi(_mm512_maskz_shufflelo_epi16(0xFFFFFFFF, (((__m512i)(__v32hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31})), 5), 1,1,0,0, 4,5,6,7, 9,9,8,8, 12,13,14,15, 17,17,16,16, 20,21,22,23, 25,25,24,24, 28,29,30,31)); +TEST_CONSTEXPR(match_v32hi(_mm512_maskz_shufflelo_epi16(0x0000FFFFu, ((__m512i)(__v32hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31}), 5), 1,1,0,0,4,5,6,7, 9,9,8,8,12,13,14,15, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0)); __m512i test_mm512_sllv_epi16(__m512i __A, __m512i __B) { // CHECK-LABEL: test_mm512_sllv_epi16 diff --git a/clang/test/CodeGen/X86/avx512f-builtins.c b/clang/test/CodeGen/X86/avx512f-builtins.c index 84eaad8..47cb485a 100644 --- a/clang/test/CodeGen/X86/avx512f-builtins.c +++ b/clang/test/CodeGen/X86/avx512f-builtins.c @@ -9073,20 +9073,25 @@ __m512i test_mm512_shuffle_epi32(__m512i __A) { // CHECK: shufflevector <16 x i32> %{{.*}}, <16 x i32> poison, <16 x i32> <i32 1, i32 0, i32 0, i32 0, i32 5, i32 4, i32 4, i32 4, i32 9, i32 8, i32 8, i32 8, i32 13, i32 12, i32 12, i32 12> return _mm512_shuffle_epi32(__A, 1); } - +TEST_CONSTEXPR(match_v16si(_mm512_shuffle_epi32((((__m512i)(__v16si){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15})), 1), 1,0,0,0, 5,4,4,4, 9,8,8,8, 13,12,12,12)); __m512i test_mm512_mask_shuffle_epi32(__m512i __W, __mmask16 __U, __m512i __A) { // CHECK-LABEL: test_mm512_mask_shuffle_epi32 // CHECK: shufflevector <16 x i32> %{{.*}}, <16 x i32> poison, <16 x i32> <i32 1, i32 0, i32 0, i32 0, i32 5, i32 4, i32 4, i32 4, i32 9, i32 8, i32 8, i32 8, i32 13, i32 12, i32 12, i32 12> // CHECK: select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> %{{.*}} return _mm512_mask_shuffle_epi32(__W, __U, __A, 1); } - +TEST_CONSTEXPR(match_v16si(_mm512_mask_shuffle_epi32(((__m512i)(__v16si){100,101,102,103,104,105,106,107,200,201,202,203,204,205,206,207}), 0xFFFFu, ((__m512i)(__v16si){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}), 1), 1,0,0,0, 5,4,4,4, 9,8,8,8, 13,12,12,12)); +TEST_CONSTEXPR(match_v16si(_mm512_mask_shuffle_epi32(((__m512i)(__v16si){100,101,102,103,104,105,106,107,200,201,202,203,204,205,206,207}), 0x0000u, ((__m512i)(__v16si){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}), 1), 100,101,102,103,104,105,106,107,200,201,202,203,204,205,206,207)); +TEST_CONSTEXPR(match_v16si(_mm512_mask_shuffle_epi32(((__m512i)(__v16si){100,101,102,103,104,105,106,107,200,201,202,203,204,205,206,207}), 0x00FFu, ((__m512i)(__v16si){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}), 1), 1,0,0,0, 5,4,4,4, 200,201,202,203,204,205,206,207)); __m512i test_mm512_maskz_shuffle_epi32(__mmask16 __U, __m512i __A) { // CHECK-LABEL: test_mm512_maskz_shuffle_epi32 // CHECK: shufflevector <16 x i32> %{{.*}}, <16 x i32> poison, <16 x i32> <i32 1, i32 0, i32 0, i32 0, i32 5, i32 4, i32 4, i32 4, i32 9, i32 8, i32 8, i32 8, i32 13, i32 12, i32 12, i32 12> // CHECK: select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> %{{.*}} return _mm512_maskz_shuffle_epi32(__U, __A, 1); } +TEST_CONSTEXPR(match_v16si(_mm512_maskz_shuffle_epi32(0xFFFFu, ((__m512i)(__v16si){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}), 1), 1,0,0,0, 5,4,4,4, 9,8,8,8, 13,12,12,12)); +TEST_CONSTEXPR(match_v16si(_mm512_maskz_shuffle_epi32(0x5555u, ((__m512i)(__v16si){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}), 1), 1,0,0,0, 5,0,4,0, 9,0,8,0, 13,0,12,0)); +TEST_CONSTEXPR(match_v16si(_mm512_maskz_shuffle_epi32(0x8001u, ((__m512i)(__v16si){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}), 1), 1,0,0,0, 0,0,0,0, 0,0,0,0, 0,0,0,12)); __m512d test_mm512_mask_expand_pd(__m512d __W, __mmask8 __U, __m512d __A) { // CHECK-LABEL: test_mm512_mask_expand_pd diff --git a/clang/test/CodeGen/X86/avx512vl-builtins.c b/clang/test/CodeGen/X86/avx512vl-builtins.c index 5282c7a..8800623 100644 --- a/clang/test/CodeGen/X86/avx512vl-builtins.c +++ b/clang/test/CodeGen/X86/avx512vl-builtins.c @@ -10025,6 +10025,11 @@ __m128i test_mm_mask_shuffle_epi32(__m128i __W, __mmask8 __U, __m128i __A) { return _mm_mask_shuffle_epi32(__W, __U, __A, 1); } +TEST_CONSTEXPR(match_v4si(_mm_mask_shuffle_epi32(((__m128i)(__v4si){100,101,102,103}), 0x0Fu, ((__m128i)(__v4si){0,1,2,3}), 1), 1,0,0,0)); +TEST_CONSTEXPR(match_v4si(_mm_mask_shuffle_epi32(((__m128i)(__v4si){100,101,102,103}), 0x0Au, ((__m128i)(__v4si){0,1,2,3}), 1), 100,0,102,0)); +TEST_CONSTEXPR(match_v4si(_mm_mask_shuffle_epi32(((__m128i)(__v4si){100,101,102,103}), 0x05u, ((__m128i)(__v4si){0,1,2,3}), 1), 1,101,0,103)); +TEST_CONSTEXPR(match_v4si(_mm_mask_shuffle_epi32(((__m128i)(__v4si){100,101,102,103}), 0x00u, ((__m128i)(__v4si){0,1,2,3}), 1), 100,101,102,103)); + __m128i test_mm_maskz_shuffle_epi32(__mmask8 __U, __m128i __A) { // CHECK-LABEL: test_mm_maskz_shuffle_epi32 // CHECK: shufflevector <4 x i32> %{{.*}}, <4 x i32> poison, <4 x i32> <i32 2, i32 0, i32 0, i32 0> @@ -10032,6 +10037,10 @@ __m128i test_mm_maskz_shuffle_epi32(__mmask8 __U, __m128i __A) { return _mm_maskz_shuffle_epi32(__U, __A, 2); } +TEST_CONSTEXPR(match_v4si(_mm_maskz_shuffle_epi32(0x01u, ((__m128i)(__v4si){0,1,2,3}), 2), 2,0,0,0)); +TEST_CONSTEXPR(match_v4si(_mm_maskz_shuffle_epi32(0x0Au, ((__m128i)(__v4si){0,1,2,3}), 2), 0,0,0,0)); +TEST_CONSTEXPR(match_v4si(_mm_maskz_shuffle_epi32(0x0Fu, ((__m128i)(__v4si){0,1,2,3}), 2), 2,0,0,0)); + __m256i test_mm256_mask_shuffle_epi32(__m256i __W, __mmask8 __U, __m256i __A) { // CHECK-LABEL: test_mm256_mask_shuffle_epi32 // CHECK: shufflevector <8 x i32> %{{.*}}, <8 x i32> poison, <8 x i32> <i32 2, i32 0, i32 0, i32 0, i32 6, i32 4, i32 4, i32 4> @@ -10039,6 +10048,10 @@ __m256i test_mm256_mask_shuffle_epi32(__m256i __W, __mmask8 __U, __m256i __A) { return _mm256_mask_shuffle_epi32(__W, __U, __A, 2); } +TEST_CONSTEXPR(match_v8si(_mm256_mask_shuffle_epi32(((__m256i)(__v8si){100,101,102,103,104,105,106,107}), 0xF0u, ((__m256i)(__v8si){0,1,2,3,4,5,6,7}), 2), 100,101,102,103, 6,4,4,4)); +TEST_CONSTEXPR(match_v8si(_mm256_mask_shuffle_epi32(((__m256i)(__v8si){100,101,102,103,104,105,106,107}), 0x33u, ((__m256i)(__v8si){0,1,2,3,4,5,6,7}), 2), 2,0,102,103, 6,4,106,107)); +TEST_CONSTEXPR(match_v8si(_mm256_mask_shuffle_epi32(((__m256i)(__v8si){100,101,102,103,104,105,106,107}), 0x00u, ((__m256i)(__v8si){0,1,2,3,4,5,6,7}), 2), 100,101,102,103,104,105,106,107)); + __m256i test_mm256_maskz_shuffle_epi32(__mmask8 __U, __m256i __A) { // CHECK-LABEL: test_mm256_maskz_shuffle_epi32 // CHECK: shufflevector <8 x i32> %{{.*}}, <8 x i32> poison, <8 x i32> <i32 2, i32 0, i32 0, i32 0, i32 6, i32 4, i32 4, i32 4> @@ -10046,6 +10059,10 @@ __m256i test_mm256_maskz_shuffle_epi32(__mmask8 __U, __m256i __A) { return _mm256_maskz_shuffle_epi32(__U, __A, 2); } +TEST_CONSTEXPR(match_v8si(_mm256_maskz_shuffle_epi32(0x33u, ((__m256i)(__v8si){0,1,2,3,4,5,6,7}), 2), 2,0,0,0, 6,4,0,0)); +TEST_CONSTEXPR(match_v8si(_mm256_maskz_shuffle_epi32(0xAAu, ((__m256i)(__v8si){0,1,2,3,4,5,6,7}), 2), 0,0,0,0, 0,4,0,4)); +TEST_CONSTEXPR(match_v8si(_mm256_maskz_shuffle_epi32(0xFFu, ((__m256i)(__v8si){0,1,2,3,4,5,6,7}), 2), 2,0,0,0, 6,4,4,4)); + __m128d test_mm_mask_mov_pd(__m128d __W, __mmask8 __U, __m128d __A) { // CHECK-LABEL: test_mm_mask_mov_pd // CHECK: select <2 x i1> %{{.*}}, <2 x double> %{{.*}}, <2 x double> %{{.*}} diff --git a/clang/test/CodeGen/X86/avx512vlbw-builtins.c b/clang/test/CodeGen/X86/avx512vlbw-builtins.c index 6c9c80e..1fe1ec0 100644 --- a/clang/test/CodeGen/X86/avx512vlbw-builtins.c +++ b/clang/test/CodeGen/X86/avx512vlbw-builtins.c @@ -3393,6 +3393,13 @@ __m128i test_mm_mask_shufflehi_epi16(__m128i __W, __mmask8 __U, __m128i __A) { return _mm_mask_shufflehi_epi16(__W, __U, __A, 5); } +TEST_CONSTEXPR(match_v8hi(_mm_mask_shufflehi_epi16(((__m128i)(__v8hi){100,101,102,103,104,105,106,107}),0xF0u,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),100,101,102,103,5,5,4,4)); +TEST_CONSTEXPR(match_v8hi(_mm_mask_shufflehi_epi16(((__m128i)(__v8hi){100,101,102,103,104,105,106,107}),0x00u,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),100,101,102,103,104,105,106,107)); +TEST_CONSTEXPR(match_v8hi(_mm_mask_shufflehi_epi16(((__m128i)(__v8hi){100,101,102,103,104,105,106,107}),0xFFu,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),0,1,2,3,5,5,4,4)); +TEST_CONSTEXPR(match_v8hi(_mm_mask_shufflehi_epi16(((__m128i)(__v8hi){100,101,102,103,104,105,106,107}),0x0Fu,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),0,1,2,3,104,105,106,107)); +TEST_CONSTEXPR(match_v8hi(_mm_mask_shufflehi_epi16(((__m128i)(__v8hi){100,101,102,103,104,105,106,107}),0x55u,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),0,101,2,103,5,105,4,107)); +TEST_CONSTEXPR(match_v8hi(_mm_mask_shufflehi_epi16(((__m128i)(__v8hi){100,101,102,103,104,105,106,107}),0xAAu,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),100,1,102,3,104,5,106,4)); + __m128i test_mm_maskz_shufflehi_epi16(__mmask8 __U, __m128i __A) { // CHECK-LABEL: test_mm_maskz_shufflehi_epi16 // CHECK: shufflevector <8 x i16> %{{.*}}, <8 x i16> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 5, i32 5, i32 4, i32 4> @@ -3400,6 +3407,13 @@ __m128i test_mm_maskz_shufflehi_epi16(__mmask8 __U, __m128i __A) { return _mm_maskz_shufflehi_epi16(__U, __A, 5); } +TEST_CONSTEXPR(match_v8hi(_mm_maskz_shufflehi_epi16(0xF0u,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),0,0,0,0,5,5,4,4)); +TEST_CONSTEXPR(match_v8hi(_mm_maskz_shufflehi_epi16(0x00u,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),0,0,0,0,0,0,0,0)); +TEST_CONSTEXPR(match_v8hi(_mm_maskz_shufflehi_epi16(0xFFu,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),0,1,2,3,5,5,4,4)); +TEST_CONSTEXPR(match_v8hi(_mm_maskz_shufflehi_epi16(0x0Fu,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),0,1,2,3,0,0,0,0)); +TEST_CONSTEXPR(match_v8hi(_mm_maskz_shufflehi_epi16(0x55u,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),0,0,2,0,5,0,4,0)); +TEST_CONSTEXPR(match_v8hi(_mm_maskz_shufflehi_epi16(0xAAu,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),0,1,0,3,0,5,0,4)); + __m128i test_mm_mask_shufflelo_epi16(__m128i __W, __mmask8 __U, __m128i __A) { // CHECK-LABEL: test_mm_mask_shufflelo_epi16 // CHECK: shufflevector <8 x i16> %{{.*}}, <8 x i16> poison, <8 x i32> <i32 1, i32 1, i32 0, i32 0, i32 4, i32 5, i32 6, i32 7> @@ -3407,6 +3421,13 @@ __m128i test_mm_mask_shufflelo_epi16(__m128i __W, __mmask8 __U, __m128i __A) { return _mm_mask_shufflelo_epi16(__W, __U, __A, 5); } +TEST_CONSTEXPR(match_v8hi(_mm_mask_shufflelo_epi16(((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),0xFF,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),1,1,0,0,4,5,6,7)); +TEST_CONSTEXPR(match_v8hi(_mm_mask_shufflelo_epi16(((__m128i)(__v8hi){100,101,102,103,104,105,106,107}),0x00u,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),100,101,102,103,104,105,106,107)); +TEST_CONSTEXPR(match_v8hi(_mm_mask_shufflelo_epi16(((__m128i)(__v8hi){100,101,102,103,104,105,106,107}),0x0Fu,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),1,1,0,0,104,105,106,107)); +TEST_CONSTEXPR(match_v8hi(_mm_mask_shufflelo_epi16(((__m128i)(__v8hi){100,101,102,103,104,105,106,107}),0xF0u,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),100,101,102,103,4,5,6,7)); +TEST_CONSTEXPR(match_v8hi(_mm_mask_shufflelo_epi16(((__m128i)(__v8hi){100,101,102,103,104,105,106,107}),0xAAu,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),100,1,102,0,104,5,106,7)); +TEST_CONSTEXPR(match_v8hi(_mm_mask_shufflelo_epi16(((__m128i)(__v8hi){100,101,102,103,104,105,106,107}),0x55u,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),1,101,0,103,4,105,6,107)); + __m128i test_mm_maskz_shufflelo_epi16(__mmask8 __U, __m128i __A) { // CHECK-LABEL: test_mm_maskz_shufflelo_epi16 // CHECK: shufflevector <8 x i16> %{{.*}}, <8 x i16> poison, <8 x i32> <i32 1, i32 1, i32 0, i32 0, i32 4, i32 5, i32 6, i32 7> @@ -3414,6 +3435,12 @@ __m128i test_mm_maskz_shufflelo_epi16(__mmask8 __U, __m128i __A) { return _mm_maskz_shufflelo_epi16(__U, __A, 5); } +TEST_CONSTEXPR(match_v8hi(_mm_maskz_shufflelo_epi16(0xFF,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),1,1,0,0,4,5,6,7)); +TEST_CONSTEXPR(match_v8hi(_mm_maskz_shufflelo_epi16(0x0Fu,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),1,1,0,0,0,0,0,0)); +TEST_CONSTEXPR(match_v8hi(_mm_maskz_shufflelo_epi16(0xF0u,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),0,0,0,0,4,5,6,7)); +TEST_CONSTEXPR(match_v8hi(_mm_maskz_shufflelo_epi16(0xAAu,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),0,1,0,0,0,5,0,7)); +TEST_CONSTEXPR(match_v8hi(_mm_maskz_shufflelo_epi16(0x55u,((__m128i)(__v8hi){0,1,2,3,4,5,6,7}),5),1,0,0,0,4,0,6,0)); + __m256i test_mm256_mask_shufflehi_epi16(__m256i __W, __mmask16 __U, __m256i __A) { // CHECK-LABEL: test_mm256_mask_shufflehi_epi16 // CHECK: shufflevector <16 x i16> %{{.*}}, <16 x i16> poison, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 5, i32 5, i32 4, i32 4, i32 8, i32 9, i32 10, i32 11, i32 13, i32 13, i32 12, i32 12> @@ -3421,6 +3448,12 @@ __m256i test_mm256_mask_shufflehi_epi16(__m256i __W, __mmask16 __U, __m256i __A) return _mm256_mask_shufflehi_epi16(__W, __U, __A, 5); } +TEST_CONSTEXPR(match_v16hi(_mm256_mask_shufflehi_epi16(((__m256i)(__v16hi){100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115}),0xFF00u,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),100,101,102,103,104,105,106,107,8,9,10,11,13,13,12,12)); +TEST_CONSTEXPR(match_v16hi(_mm256_mask_shufflehi_epi16(((__m256i)(__v16hi){100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115}),0x0000u,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115)); +TEST_CONSTEXPR(match_v16hi(_mm256_mask_shufflehi_epi16(((__m256i)(__v16hi){100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115}),0xFFFFu,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),0,1,2,3,5,5,4,4,8,9,10,11,13,13,12,12)); +TEST_CONSTEXPR(match_v16hi(_mm256_mask_shufflehi_epi16(((__m256i)(__v16hi){100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115}),0x00FFu,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),0,1,2,3,5,5,4,4,108,109,110,111,112,113,114,115)); +TEST_CONSTEXPR(match_v16hi(_mm256_mask_shufflehi_epi16(((__m256i)(__v16hi){100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115}),0x5555u,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),0,101,2,103,5,105,4,107,8,109,10,111,13,113,12,115)); + __m256i test_mm256_maskz_shufflehi_epi16(__mmask16 __U, __m256i __A) { // CHECK-LABEL: test_mm256_maskz_shufflehi_epi16 // CHECK: shufflevector <16 x i16> %{{.*}}, <16 x i16> poison, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 5, i32 5, i32 4, i32 4, i32 8, i32 9, i32 10, i32 11, i32 13, i32 13, i32 12, i32 12> @@ -3428,6 +3461,13 @@ __m256i test_mm256_maskz_shufflehi_epi16(__mmask16 __U, __m256i __A) { return _mm256_maskz_shufflehi_epi16(__U, __A, 5); } +TEST_CONSTEXPR(match_v16hi(_mm256_maskz_shufflehi_epi16(0x0000u,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0)); +TEST_CONSTEXPR(match_v16hi(_mm256_maskz_shufflehi_epi16(0xFFFFu,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),0,1,2,3,5,5,4,4,8,9,10,11,13,13,12,12)); +TEST_CONSTEXPR(match_v16hi(_mm256_maskz_shufflehi_epi16(0x00FFu,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),0,1,2,3,5,5,4,4,0,0,0,0,0,0,0,0)); +TEST_CONSTEXPR(match_v16hi(_mm256_maskz_shufflehi_epi16(0xFF00u,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),0,0,0,0,0,0,0,0,8,9,10,11,13,13,12,12)); +TEST_CONSTEXPR(match_v16hi(_mm256_maskz_shufflehi_epi16(0x5555u,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),0,0,2,0,5,0,4,0,8,0,10,0,13,0,12,0)); +TEST_CONSTEXPR(match_v16hi(_mm256_maskz_shufflehi_epi16(0xAAAAu,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),0,1,0,3,0,5,0,4,0,9,0,11,0,13,0,12)); + __m256i test_mm256_mask_shufflelo_epi16(__m256i __W, __mmask16 __U, __m256i __A) { // CHECK-LABEL: test_mm256_mask_shufflelo_epi16 // CHECK: shufflevector <16 x i16> %{{.*}}, <16 x i16> poison, <16 x i32> <i32 1, i32 1, i32 0, i32 0, i32 4, i32 5, i32 6, i32 7, i32 9, i32 9, i32 8, i32 8, i32 12, i32 13, i32 14, i32 15> @@ -3435,6 +3475,11 @@ __m256i test_mm256_mask_shufflelo_epi16(__m256i __W, __mmask16 __U, __m256i __A) return _mm256_mask_shufflelo_epi16(__W, __U, __A, 5); } +TEST_CONSTEXPR(match_v16hi(_mm256_mask_shufflelo_epi16(((__m256i)(__v16hi){100,101,102,103,104,105,106,107,200,201,202,203,204,205,206,207}),0xFFFF,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),1,1,0,0,4,5,6,7,9,9,8,8,12,13,14,15)); +TEST_CONSTEXPR(match_v16hi(_mm256_mask_shufflelo_epi16(((__m256i)(__v16hi){100,101,102,103,104,105,106,107,200,201,202,203,204,205,206,207}),0x000Fu,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),1,1,0,0,104,105,106,107,200,201,202,203,204,205,206,207)); +TEST_CONSTEXPR(match_v16hi(_mm256_mask_shufflelo_epi16(((__m256i)(__v16hi){100,101,102,103,104,105,106,107,200,201,202,203,204,205,206,207}),0x00FFu,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),1,1,0,0,4,5,6,7,200,201,202,203,204,205,206,207)); +TEST_CONSTEXPR(match_v16hi(_mm256_mask_shufflelo_epi16(((__m256i)(__v16hi){100,101,102,103,104,105,106,107,200,201,202,203,204,205,206,207}),0xF00Fu,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),1,1,0,0,104,105,106,107,200,201,202,203,12,13,14,15)); + __m256i test_mm256_maskz_shufflelo_epi16(__mmask16 __U, __m256i __A) { // CHECK-LABEL: test_mm256_maskz_shufflelo_epi16 // CHECK: shufflevector <16 x i16> %{{.*}}, <16 x i16> poison, <16 x i32> <i32 1, i32 1, i32 0, i32 0, i32 4, i32 5, i32 6, i32 7, i32 9, i32 9, i32 8, i32 8, i32 12, i32 13, i32 14, i32 15> @@ -3442,6 +3487,11 @@ __m256i test_mm256_maskz_shufflelo_epi16(__mmask16 __U, __m256i __A) { return _mm256_maskz_shufflelo_epi16(__U, __A, 5); } +TEST_CONSTEXPR(match_v16hi(_mm256_maskz_shufflelo_epi16(0xFFFF,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),1,1,0,0,4,5,6,7,9,9,8,8,12,13,14,15)); +TEST_CONSTEXPR(match_v16hi(_mm256_maskz_shufflelo_epi16(0x000Fu,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),1,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0)); +TEST_CONSTEXPR(match_v16hi(_mm256_maskz_shufflelo_epi16(0x00FFu,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),1,1,0,0,4,5,6,7,0,0,0,0,0,0,0,0)); +TEST_CONSTEXPR(match_v16hi(_mm256_maskz_shufflelo_epi16(0xF0F0u,((__m256i)(__v16hi){0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}),5),0,0,0,0,4,5,6,7,0,0,0,0,12,13,14,15)); + void test_mm_mask_cvtepi16_storeu_epi8 (void * __P, __mmask8 __M, __m128i __A) { // CHECK-LABEL: test_mm_mask_cvtepi16_storeu_epi8 diff --git a/clang/test/CodeGen/X86/mmx-builtins.c b/clang/test/CodeGen/X86/mmx-builtins.c index 26c5f73..a4494b69 100644 --- a/clang/test/CodeGen/X86/mmx-builtins.c +++ b/clang/test/CodeGen/X86/mmx-builtins.c @@ -292,6 +292,7 @@ int test_mm_extract_pi16(__m64 a) { // CHECK: extractelement <4 x i16> {{%.*}}, i64 2 return _mm_extract_pi16(a, 2); } +TEST_CONSTEXPR(_mm_extract_pi16(((__m64)(__v4hi){10, 20, 30, 40}), 7) == 40); __m64 test_m_from_int(int a) { // CHECK-LABEL: test_m_from_int @@ -347,6 +348,7 @@ __m64 test_mm_insert_pi16(__m64 a, int d) { // CHECK: insertelement <4 x i16> return _mm_insert_pi16(a, d, 2); } +TEST_CONSTEXPR(match_v4hi(_mm_insert_pi16(((__m64)(__v4hi){0, 1, 2, 3}), 77, 10), 0, 1, 77, 3)); __m64 test_mm_madd_pi16(__m64 a, __m64 b) { // CHECK-LABEL: test_mm_madd_pi16 @@ -584,7 +586,7 @@ __m64 test_mm_shuffle_pi16(__m64 a) { // CHECK: shufflevector <4 x i16> {{%.*}}, <4 x i16> {{%.*}}, <4 x i32> <i32 3, i32 0, i32 0, i32 0> return _mm_shuffle_pi16(a, 3); } - +TEST_CONSTEXPR(match_v4hi(_mm_shuffle_pi16(((__m64)(__v4hi){0,1,2,3}), 3), 3,0,0,0)); __m64 test_mm_sign_pi8(__m64 a, __m64 b) { // CHECK-LABEL: test_mm_sign_pi8 // CHECK: call <16 x i8> @llvm.x86.ssse3.psign.b.128( diff --git a/clang/test/CodeGen/X86/sse2-builtins.c b/clang/test/CodeGen/X86/sse2-builtins.c index 84b90c0..8428fd6 100644 --- a/clang/test/CodeGen/X86/sse2-builtins.c +++ b/clang/test/CodeGen/X86/sse2-builtins.c @@ -723,12 +723,14 @@ int test_mm_extract_epi16(__m128i A) { // CHECK: zext i16 %{{.*}} to i32 return _mm_extract_epi16(A, 1); } +TEST_CONSTEXPR(_mm_extract_epi16(((__m128i)(__v8hi){0, 10, 20, 30, 40, 50, 60, 70}), 25) == 10); __m128i test_mm_insert_epi16(__m128i A, int B) { // CHECK-LABEL: test_mm_insert_epi16 // CHECK: insertelement <8 x i16> %{{.*}}, {{i32|i64}} 0 return _mm_insert_epi16(A, B, 0); } +TEST_CONSTEXPR(match_v8hi(_mm_insert_epi16(((__m128i)(__v8hi){0, 10, 20, 30, 40, 50, 60, 70}), 555, 17), 0, 555, 20, 30, 40, 50, 60, 70)); void test_mm_lfence(void) { // CHECK-LABEL: test_mm_lfence @@ -1299,7 +1301,7 @@ __m128i test_mm_shuffle_epi32(__m128i A) { // CHECK: shufflevector <4 x i32> %{{.*}}, <4 x i32> poison, <4 x i32> zeroinitializer return _mm_shuffle_epi32(A, 0); } - +TEST_CONSTEXPR(match_v4si(_mm_shuffle_epi32(((__m128i)(__v4si){0,1,2,3}), 0), 0,0,0,0)); __m128d test_mm_shuffle_pd(__m128d A, __m128d B) { // CHECK-LABEL: test_mm_shuffle_pd // CHECK: shufflevector <2 x double> %{{.*}}, <2 x double> %{{.*}}, <2 x i32> <i32 1, i32 2> @@ -1311,13 +1313,13 @@ __m128i test_mm_shufflehi_epi16(__m128i A) { // CHECK: shufflevector <8 x i16> %{{.*}}, <8 x i16> poison, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 4, i32 4, i32 4> return _mm_shufflehi_epi16(A, 0); } - +TEST_CONSTEXPR(match_v8hi(_mm_shufflehi_epi16(((__m128i)(__v8hi){0,1,2,3,4,5,6,7}), 0), 0,1,2,3, 4,4,4,4)); __m128i test_mm_shufflelo_epi16(__m128i A) { // CHECK-LABEL: test_mm_shufflelo_epi16 // CHECK: shufflevector <8 x i16> %{{.*}}, <8 x i16> poison, <8 x i32> <i32 0, i32 0, i32 0, i32 0, i32 4, i32 5, i32 6, i32 7> return _mm_shufflelo_epi16(A, 0); } - +TEST_CONSTEXPR(match_v8hi(_mm_shufflelo_epi16(((__m128i)(__v8hi){0,1,2,3,4,5,6,7}), 0), 0,0,0,0, 4,5,6,7)); __m128i test_mm_sll_epi16(__m128i A, __m128i B) { // CHECK-LABEL: test_mm_sll_epi16 // CHECK: call <8 x i16> @llvm.x86.sse2.psll.w(<8 x i16> %{{.*}}, <8 x i16> %{{.*}}) diff --git a/clang/test/CodeGen/X86/sse41-builtins.c b/clang/test/CodeGen/X86/sse41-builtins.c index 3c37246..eee479a 100644 --- a/clang/test/CodeGen/X86/sse41-builtins.c +++ b/clang/test/CodeGen/X86/sse41-builtins.c @@ -231,24 +231,28 @@ int test_mm_extract_epi8(__m128i x) { // CHECK: zext i8 %{{.*}} to i32 return _mm_extract_epi8(x, 1); } +TEST_CONSTEXPR(_mm_extract_epi8(((__m128i)(__v16qi){0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15}), 20) == 4); int test_mm_extract_epi32(__m128i x) { // CHECK-LABEL: test_mm_extract_epi32 // CHECK: extractelement <4 x i32> %{{.*}}, {{i32|i64}} 1 return _mm_extract_epi32(x, 1); } +TEST_CONSTEXPR(_mm_extract_epi32(((__m128i)(__v4si){1, 3, 5, 7}), 10) == 5); long long test_mm_extract_epi64(__m128i x) { // CHECK-LABEL: test_mm_extract_epi64 // CHECK: extractelement <2 x i64> %{{.*}}, {{i32|i64}} 1 return _mm_extract_epi64(x, 1); } +TEST_CONSTEXPR(_mm_extract_epi64(((__m128i)(__v2di){11, 22}), 5) == 22); int test_mm_extract_ps(__m128 x) { // CHECK-LABEL: test_mm_extract_ps // CHECK: extractelement <4 x float> %{{.*}}, {{i32|i64}} 1 return _mm_extract_ps(x, 1); } +TEST_CONSTEXPR(_mm_extract_ps(((__m128){1.25f, 2.5f, 3.75f, 5.0f}), 6) == __builtin_bit_cast(int, 3.75f)); __m128d test_mm_floor_pd(__m128d x) { // CHECK-LABEL: test_mm_floor_pd @@ -279,12 +283,14 @@ __m128i test_mm_insert_epi8(__m128i x, char b) { // CHECK: insertelement <16 x i8> %{{.*}}, i8 %{{.*}}, {{i32|i64}} 1 return _mm_insert_epi8(x, b, 1); } +TEST_CONSTEXPR(match_v16qi(_mm_insert_epi8(((__m128i)(__v16qi){ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15}), 101, 33), 0, 101, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15)); __m128i test_mm_insert_epi32(__m128i x, int b) { // CHECK-LABEL: test_mm_insert_epi32 // CHECK: insertelement <4 x i32> %{{.*}}, i32 %{{.*}}, {{i32|i64}} 1 return _mm_insert_epi32(x, b, 1); } +TEST_CONSTEXPR(match_v4si(_mm_insert_epi32(((__m128i)(__v4si){0, 1, 2, 3}), 5678, 18), 0, 1, 5678, 3)); #ifdef __x86_64__ __m128i test_mm_insert_epi64(__m128i x, long long b) { @@ -292,6 +298,7 @@ __m128i test_mm_insert_epi64(__m128i x, long long b) { // X64: insertelement <2 x i64> %{{.*}}, i64 %{{.*}}, {{i32|i64}} 1 return _mm_insert_epi64(x, b, 1); } +TEST_CONSTEXPR(match_v2di(_mm_insert_epi64(((__m128i)(__v2di){100, 200}), -999, 9), 100, -999)); #endif __m128 test_mm_insert_ps(__m128 x, __m128 y) { diff --git a/clang/test/CodeGenHLSL/resources/AppendStructuredBuffer-elementtype.hlsl b/clang/test/CodeGenHLSL/resources/AppendStructuredBuffer-elementtype.hlsl deleted file mode 100644 index 094006f..0000000 --- a/clang/test/CodeGenHLSL/resources/AppendStructuredBuffer-elementtype.hlsl +++ /dev/null @@ -1,54 +0,0 @@ -// RUN: %clang_cc1 -triple dxil-pc-shadermodel6.2-compute -finclude-default-header -fnative-half-type -emit-llvm -o - %s | FileCheck %s -check-prefixes=DXIL -// RUN: %clang_cc1 -triple spirv-unknown-vulkan1.3-compute -finclude-default-header -fnative-half-type -emit-llvm -o - %s | FileCheck %s -check-prefixes=SPV - -struct MyStruct { - float4 a; - int2 b; -}; - -// DXIL: %"class.hlsl::AppendStructuredBuffer" = type { target("dx.RawBuffer", i16, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.0" = type { target("dx.RawBuffer", i16, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.1" = type { target("dx.RawBuffer", i32, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.2" = type { target("dx.RawBuffer", i32, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.3" = type { target("dx.RawBuffer", i64, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.4" = type { target("dx.RawBuffer", i64, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.5" = type { target("dx.RawBuffer", half, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.6" = type { target("dx.RawBuffer", float, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.7" = type { target("dx.RawBuffer", double, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.8" = type { target("dx.RawBuffer", <4 x i16>, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.9" = type { target("dx.RawBuffer", <3 x i32>, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.10" = type { target("dx.RawBuffer", <2 x half>, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.11" = type { target("dx.RawBuffer", <3 x float>, 1, 0) -// DXIL: %"class.hlsl::AppendStructuredBuffer.12" = type { target("dx.RawBuffer", %struct.MyStruct, 1, 0) -// DXIL: %struct.MyStruct = type <{ <4 x float>, <2 x i32> }> -// DXIL: %"class.hlsl::AppendStructuredBuffer.13" = type { target("dx.RawBuffer", i32, 1, 0) -// SPV: %"class.hlsl::AppendStructuredBuffer.13" = type { target("spirv.VulkanBuffer", [0 x i32], 12, 1) -// DXIL: %"class.hlsl::AppendStructuredBuffer.14" = type { target("dx.RawBuffer", <4 x i32>, 1, 0) -// SPV: %"class.hlsl::AppendStructuredBuffer.14" = type { target("spirv.VulkanBuffer", [0 x <4 x i32>], 12, 1) - -AppendStructuredBuffer<int16_t> BufI16; -AppendStructuredBuffer<uint16_t> BufU16; -AppendStructuredBuffer<int> BufI32; -AppendStructuredBuffer<uint> BufU32; -AppendStructuredBuffer<int64_t> BufI64; -AppendStructuredBuffer<uint64_t> BufU64; -AppendStructuredBuffer<half> BufF16; -AppendStructuredBuffer<float> BufF32; -AppendStructuredBuffer<double> BufF64; -AppendStructuredBuffer< vector<int16_t, 4> > BufI16x4; -AppendStructuredBuffer< vector<uint, 3> > BufU32x3; -AppendStructuredBuffer<half2> BufF16x2; -AppendStructuredBuffer<float3> BufF32x3; -// TODO: AppendStructuredBuffer<snorm half> BufSNormF16; -// TODO: AppendStructuredBuffer<unorm half> BufUNormF16; -// TODO: AppendStructuredBuffer<snorm float> BufSNormF32; -// TODO: AppendStructuredBuffer<unorm float> BufUNormF32; -// TODO: AppendStructuredBuffer<snorm double> BufSNormF64; -// TODO: AppendStructuredBuffer<unorm double> BufUNormF64; -AppendStructuredBuffer<MyStruct> BufMyStruct; -AppendStructuredBuffer<bool> BufBool; -AppendStructuredBuffer<bool4> BufBoolVec; - -[numthreads(1,1,1)] -void main(int GI : SV_GroupIndex) { -} diff --git a/clang/test/CodeGenHLSL/resources/ConsumeStructuredBuffer-elementtype.hlsl b/clang/test/CodeGenHLSL/resources/ConsumeStructuredBuffer-elementtype.hlsl deleted file mode 100644 index 632fd91..0000000 --- a/clang/test/CodeGenHLSL/resources/ConsumeStructuredBuffer-elementtype.hlsl +++ /dev/null @@ -1,54 +0,0 @@ -// RUN: %clang_cc1 -triple dxil-pc-shadermodel6.2-compute -finclude-default-header -fnative-half-type -emit-llvm -o - %s | FileCheck %s -check-prefixes=DXIL -// RUN: %clang_cc1 -triple spirv-unknown-vulkan1.3-compute -finclude-default-header -fnative-half-type -emit-llvm -o - %s | FileCheck %s -check-prefixes=SPV - -struct MyStruct { - float4 a; - int2 b; -}; - -// DXIL: %"class.hlsl::ConsumeStructuredBuffer" = type { target("dx.RawBuffer", i16, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.0" = type { target("dx.RawBuffer", i16, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.1" = type { target("dx.RawBuffer", i32, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.2" = type { target("dx.RawBuffer", i32, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.3" = type { target("dx.RawBuffer", i64, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.4" = type { target("dx.RawBuffer", i64, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.5" = type { target("dx.RawBuffer", half, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.6" = type { target("dx.RawBuffer", float, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.7" = type { target("dx.RawBuffer", double, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.8" = type { target("dx.RawBuffer", <4 x i16>, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.9" = type { target("dx.RawBuffer", <3 x i32>, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.10" = type { target("dx.RawBuffer", <2 x half>, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.11" = type { target("dx.RawBuffer", <3 x float>, 1, 0) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.12" = type { target("dx.RawBuffer", %struct.MyStruct, 1, 0) -// DXIL: %struct.MyStruct = type <{ <4 x float>, <2 x i32> }> -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.13" = type { target("dx.RawBuffer", i32, 1, 0) -// SPV: %"class.hlsl::ConsumeStructuredBuffer.13" = type { target("spirv.VulkanBuffer", [0 x i32], 12, 1) -// DXIL: %"class.hlsl::ConsumeStructuredBuffer.14" = type { target("dx.RawBuffer", <4 x i32>, 1, 0) -// SPV: %"class.hlsl::ConsumeStructuredBuffer.14" = type { target("spirv.VulkanBuffer", [0 x <4 x i32>], 12, 1) - -ConsumeStructuredBuffer<int16_t> BufI16; -ConsumeStructuredBuffer<uint16_t> BufU16; -ConsumeStructuredBuffer<int> BufI32; -ConsumeStructuredBuffer<uint> BufU32; -ConsumeStructuredBuffer<int64_t> BufI64; -ConsumeStructuredBuffer<uint64_t> BufU64; -ConsumeStructuredBuffer<half> BufF16; -ConsumeStructuredBuffer<float> BufF32; -ConsumeStructuredBuffer<double> BufF64; -ConsumeStructuredBuffer< vector<int16_t, 4> > BufI16x4; -ConsumeStructuredBuffer< vector<uint, 3> > BufU32x3; -ConsumeStructuredBuffer<half2> BufF16x2; -ConsumeStructuredBuffer<float3> BufF32x3; -// TODO: ConsumeStructuredBuffer<snorm half> BufSNormF16; -// TODO: ConsumeStructuredBuffer<unorm half> BufUNormF16; -// TODO: ConsumeStructuredBuffer<snorm float> BufSNormF32; -// TODO: ConsumeStructuredBuffer<unorm float> BufUNormF32; -// TODO: ConsumeStructuredBuffer<snorm double> BufSNormF64; -// TODO: ConsumeStructuredBuffer<unorm double> BufUNormF64; -ConsumeStructuredBuffer<MyStruct> BufMyStruct; -ConsumeStructuredBuffer<bool> BufBool; -ConsumeStructuredBuffer<bool4> BufBoolVec; - -[numthreads(1,1,1)] -void main(int GI : SV_GroupIndex) { -} diff --git a/clang/test/CodeGenHLSL/resources/RWStructuredBuffer-elementtype.hlsl b/clang/test/CodeGenHLSL/resources/RWStructuredBuffer-elementtype.hlsl deleted file mode 100644 index 9f0a5b7..0000000 --- a/clang/test/CodeGenHLSL/resources/RWStructuredBuffer-elementtype.hlsl +++ /dev/null @@ -1,74 +0,0 @@ -// RUN: %clang_cc1 -triple dxil-pc-shadermodel6.2-compute -finclude-default-header -fnative-half-type -emit-llvm -o - %s | FileCheck %s -check-prefixes=CHECK -// RUN: %clang_cc1 -triple spirv-unknown-vulkan1.3-compute -finclude-default-header -fnative-half-type -emit-llvm -o - %s | FileCheck %s -check-prefixes=SPV - -// CHECK: %"class.hlsl::RWStructuredBuffer" = type { target("dx.RawBuffer", i16, 1, 0), target("dx.RawBuffer", i16, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer" = type { target("spirv.VulkanBuffer", [0 x i16], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.0" = type { target("dx.RawBuffer", i16, 1, 0), target("dx.RawBuffer", i16, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.0" = type { target("spirv.VulkanBuffer", [0 x i16], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.1" = type { target("dx.RawBuffer", i32, 1, 0), target("dx.RawBuffer", i32, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.1" = type { target("spirv.VulkanBuffer", [0 x i32], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.2" = type { target("dx.RawBuffer", i32, 1, 0), target("dx.RawBuffer", i32, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.2" = type { target("spirv.VulkanBuffer", [0 x i32], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.3" = type { target("dx.RawBuffer", i64, 1, 0), target("dx.RawBuffer", i64, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.3" = type { target("spirv.VulkanBuffer", [0 x i64], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.4" = type { target("dx.RawBuffer", i64, 1, 0), target("dx.RawBuffer", i64, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.4" = type { target("spirv.VulkanBuffer", [0 x i64], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.5" = type { target("dx.RawBuffer", half, 1, 0), target("dx.RawBuffer", half, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.5" = type { target("spirv.VulkanBuffer", [0 x half], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.6" = type { target("dx.RawBuffer", float, 1, 0), target("dx.RawBuffer", float, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.6" = type { target("spirv.VulkanBuffer", [0 x float], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.7" = type { target("dx.RawBuffer", double, 1, 0), target("dx.RawBuffer", double, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.7" = type { target("spirv.VulkanBuffer", [0 x double], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.8" = type { target("dx.RawBuffer", <4 x i16>, 1, 0), target("dx.RawBuffer", <4 x i16>, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.8" = type { target("spirv.VulkanBuffer", [0 x <4 x i16>], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.9" = type { target("dx.RawBuffer", <3 x i32>, 1, 0), target("dx.RawBuffer", <3 x i32>, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.9" = type { target("spirv.VulkanBuffer", [0 x <3 x i32>], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.10" = type { target("dx.RawBuffer", <2 x half>, 1, 0), target("dx.RawBuffer", <2 x half>, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.10" = type { target("spirv.VulkanBuffer", [0 x <2 x half>], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.11" = type { target("dx.RawBuffer", <3 x float>, 1, 0), target("dx.RawBuffer", <3 x float>, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.11" = type { target("spirv.VulkanBuffer", [0 x <3 x float>], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.12" = type { target("dx.RawBuffer", i32, 1, 0), target("dx.RawBuffer", i32, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.12" = type { target("spirv.VulkanBuffer", [0 x i32], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } -// CHECK: %"class.hlsl::RWStructuredBuffer.13" = type { target("dx.RawBuffer", <4 x i32>, 1, 0), target("dx.RawBuffer", <4 x i32>, 1, 0) } -// SPV: %"class.hlsl::RWStructuredBuffer.13" = type { target("spirv.VulkanBuffer", [0 x <4 x i32>], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } - -RWStructuredBuffer<int16_t> BufI16; -RWStructuredBuffer<uint16_t> BufU16; -RWStructuredBuffer<int> BufI32; -RWStructuredBuffer<uint> BufU32; -RWStructuredBuffer<int64_t> BufI64; -RWStructuredBuffer<uint64_t> BufU64; -RWStructuredBuffer<half> BufF16; -RWStructuredBuffer<float> BufF32; -RWStructuredBuffer<double> BufF64; -RWStructuredBuffer< vector<int16_t, 4> > BufI16x4; -RWStructuredBuffer< vector<uint, 3> > BufU32x3; -RWStructuredBuffer<half2> BufF16x2; -RWStructuredBuffer<float3> BufF32x3; -RWStructuredBuffer<bool> BufBool; -RWStructuredBuffer<bool4> BufBoolVec; -// TODO: RWStructuredBuffer<snorm half> BufSNormF16; -// TODO: RWStructuredBuffer<unorm half> BufUNormF16; -// TODO: RWStructuredBuffer<snorm float> BufSNormF32; -// TODO: RWStructuredBuffer<unorm float> BufUNormF32; -// TODO: RWStructuredBuffer<snorm double> BufSNormF64; -// TODO: RWStructuredBuffer<unorm double> BufUNormF64; - -[numthreads(1,1,1)] -void main(int GI : SV_GroupIndex) { - BufI16[GI] = 0; - BufU16[GI] = 0; - BufI32[GI] = 0; - BufU32[GI] = 0; - BufI64[GI] = 0; - BufU64[GI] = 0; - BufF16[GI] = 0; - BufF32[GI] = 0; - BufF64[GI] = 0; - BufI16x4[GI] = 0; - BufU32x3[GI] = 0; - BufF16x2[GI] = 0; - BufF32x3[GI] = 0; - BufBool[GI] = false; - BufBool[GI] = false; -} diff --git a/clang/test/CodeGenHLSL/resources/StructuredBuffer-elementtype.hlsl b/clang/test/CodeGenHLSL/resources/StructuredBuffer-elementtype.hlsl deleted file mode 100644 index 00216df..0000000 --- a/clang/test/CodeGenHLSL/resources/StructuredBuffer-elementtype.hlsl +++ /dev/null @@ -1,61 +0,0 @@ -// RUN: %clang_cc1 -triple dxil-pc-shadermodel6.2-compute -finclude-default-header -fnative-half-type -emit-llvm -o - %s | FileCheck %s -check-prefixes=CHECK -// RUN: %clang_cc1 -triple spirv-unknown-vulkan1.3-compute -finclude-default-header -fnative-half-type -emit-llvm -o - %s | FileCheck %s -check-prefixes=SPV - -// CHECK: %"class.hlsl::StructuredBuffer" = type { target("dx.RawBuffer", i16, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.0" = type { target("dx.RawBuffer", i16, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.1" = type { target("dx.RawBuffer", i32, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.2" = type { target("dx.RawBuffer", i32, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.3" = type { target("dx.RawBuffer", i64, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.4" = type { target("dx.RawBuffer", i64, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.5" = type { target("dx.RawBuffer", half, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.6" = type { target("dx.RawBuffer", float, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.7" = type { target("dx.RawBuffer", double, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.8" = type { target("dx.RawBuffer", <4 x i16>, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.9" = type { target("dx.RawBuffer", <3 x i32>, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.10" = type { target("dx.RawBuffer", <2 x half>, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.11" = type { target("dx.RawBuffer", <3 x float>, 0, 0) } -// CHECK: %"class.hlsl::StructuredBuffer.12" = type { target("dx.RawBuffer", i32, 0, 0) } -// SPV: %"class.hlsl::StructuredBuffer.12" = type { target("spirv.VulkanBuffer", [0 x i32], 12, 0) -// CHECK: %"class.hlsl::StructuredBuffer.13" = type { target("dx.RawBuffer", <4 x i32>, 0, 0) } -// SPV: %"class.hlsl::StructuredBuffer.13" = type { target("spirv.VulkanBuffer", [0 x <4 x i32>], 12, 0) - -StructuredBuffer<int16_t> BufI16; -StructuredBuffer<uint16_t> BufU16; -StructuredBuffer<int> BufI32; -StructuredBuffer<uint> BufU32; -StructuredBuffer<int64_t> BufI64; -StructuredBuffer<uint64_t> BufU64; -StructuredBuffer<half> BufF16; -StructuredBuffer<float> BufF32; -StructuredBuffer<double> BufF64; -StructuredBuffer< vector<int16_t, 4> > BufI16x4; -StructuredBuffer< vector<uint, 3> > BufU32x3; -StructuredBuffer<half2> BufF16x2; -StructuredBuffer<float3> BufF32x3; -StructuredBuffer<bool> BufBool; -StructuredBuffer<bool4> BufBoolVec; -// TODO: StructuredBuffer<snorm half> BufSNormF16; -// TODO: StructuredBuffer<unorm half> BufUNormF16; -// TODO: StructuredBuffer<snorm float> BufSNormF32; -// TODO: StructuredBuffer<unorm float> BufUNormF32; -// TODO: StructuredBuffer<snorm double> BufSNormF64; -// TODO: StructuredBuffer<unorm double> BufUNormF64; - -[numthreads(1,1,1)] -void main(int GI : SV_GroupIndex) { - int16_t v1 = BufI16[GI]; - uint16_t v2 = BufU16[GI]; - int v3 = BufI32[GI]; - uint v4 = BufU32[GI]; - int64_t v5 = BufI64[GI]; - uint64_t v6 = BufU64[GI]; - half v7 = BufF16[GI]; - float v8 = BufF32[GI]; - double v9 = BufF64[GI]; - vector<int16_t,4> v10 = BufI16x4[GI]; - vector<int, 3> v11 = BufU32x3[GI]; - half2 v12 = BufF16x2[GI]; - float3 v13 = BufF32x3[GI]; - bool v14 = BufBool[GI]; - bool4 v15 = BufBoolVec[GI]; -} diff --git a/clang/test/CodeGenHLSL/resources/StructuredBuffers-elementtype.hlsl b/clang/test/CodeGenHLSL/resources/StructuredBuffers-elementtype.hlsl new file mode 100644 index 0000000..2b286bd --- /dev/null +++ b/clang/test/CodeGenHLSL/resources/StructuredBuffers-elementtype.hlsl @@ -0,0 +1,113 @@ +// RUN: %clang_cc1 -triple dxil-pc-shadermodel6.2-compute -finclude-default-header -fnative-half-type \ +// RUN: -emit-llvm -o - -DRESOURCE=StructuredBuffer %s | FileCheck %s -DRESOURCE=StructuredBuffer -check-prefixes=DXIL-RO + +// RUN: %clang_cc1 -triple spirv-unknown-vulkan1.3-compute -finclude-default-header -fnative-half-type \ +// RUN: -emit-llvm -o - -DRESOURCE=StructuredBuffer %s | FileCheck %s -DRESOURCE=StructuredBuffer -check-prefixes=SPV-RO + +// RUN: %clang_cc1 -triple dxil-pc-shadermodel6.2-compute -finclude-default-header -fnative-half-type \ +// RUN: -emit-llvm -o - -DRESOURCE=RWStructuredBuffer %s | FileCheck %s -DRESOURCE=RWStructuredBuffer -check-prefixes=DXIL-RW + +// RUN: %clang_cc1 -triple spirv-unknown-vulkan1.3-compute -finclude-default-header -fnative-half-type \ +// RUN: -emit-llvm -o - -DRESOURCE=RWStructuredBuffer %s | FileCheck %s -DRESOURCE=RWStructuredBuffer -check-prefixes=SPV-RW + +// RUN: %clang_cc1 -triple dxil-pc-shadermodel6.2-compute -finclude-default-header -fnative-half-type \ +// RUN: -emit-llvm -o - -DRESOURCE=AppendStructuredBuffer %s | FileCheck %s -DRESOURCE=AppendStructuredBuffer -check-prefixes=DXIL-RW + +// RUN: %clang_cc1 -triple spirv-unknown-vulkan1.3-compute -finclude-default-header -fnative-half-type \ +// RUN: -emit-llvm -o - -DRESOURCE=AppendStructuredBuffer %s | FileCheck %s -DRESOURCE=AppendStructuredBuffer -check-prefixes=SPV-RW + +// RUN: %clang_cc1 -triple dxil-pc-shadermodel6.2-compute -finclude-default-header -fnative-half-type \ +// RUN: -emit-llvm -o - -DRESOURCE=ConsumeStructuredBuffer %s | FileCheck %s -DRESOURCE=ConsumeStructuredBuffer -check-prefixes=DXIL-RW + +// RUN: %clang_cc1 -triple spirv-unknown-vulkan1.3-compute -finclude-default-header -fnative-half-type \ +// RUN: -emit-llvm -o - -DRESOURCE=ConsumeStructuredBuffer %s | FileCheck %s -DRESOURCE=ConsumeStructuredBuffer -check-prefixes=SPV-RW + +// DXIL-RO: %"class.hlsl::[[RESOURCE]]" = type { target("dx.RawBuffer", i16, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].0" = type { target("dx.RawBuffer", i16, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].1" = type { target("dx.RawBuffer", i32, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].2" = type { target("dx.RawBuffer", i32, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].3" = type { target("dx.RawBuffer", i64, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].4" = type { target("dx.RawBuffer", i64, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].5" = type { target("dx.RawBuffer", half, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].6" = type { target("dx.RawBuffer", float, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].7" = type { target("dx.RawBuffer", double, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].8" = type { target("dx.RawBuffer", <4 x i16>, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].9" = type { target("dx.RawBuffer", <3 x i32>, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].10" = type { target("dx.RawBuffer", <2 x half>, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].11" = type { target("dx.RawBuffer", <3 x float>, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].12" = type { target("dx.RawBuffer", i32, 0, 0) } +// DXIL-RO: %"class.hlsl::[[RESOURCE]].13" = type { target("dx.RawBuffer", <4 x i32>, 0, 0) } + +// DXIL-RW: %"class.hlsl::[[RESOURCE]]" = type { target("dx.RawBuffer", i16, 1, 0), target("dx.RawBuffer", i16, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].0" = type { target("dx.RawBuffer", i16, 1, 0), target("dx.RawBuffer", i16, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].1" = type { target("dx.RawBuffer", i32, 1, 0), target("dx.RawBuffer", i32, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].2" = type { target("dx.RawBuffer", i32, 1, 0), target("dx.RawBuffer", i32, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].3" = type { target("dx.RawBuffer", i64, 1, 0), target("dx.RawBuffer", i64, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].4" = type { target("dx.RawBuffer", i64, 1, 0), target("dx.RawBuffer", i64, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].5" = type { target("dx.RawBuffer", half, 1, 0), target("dx.RawBuffer", half, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].6" = type { target("dx.RawBuffer", float, 1, 0), target("dx.RawBuffer", float, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].7" = type { target("dx.RawBuffer", double, 1, 0), target("dx.RawBuffer", double, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].8" = type { target("dx.RawBuffer", <4 x i16>, 1, 0), target("dx.RawBuffer", <4 x i16>, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].9" = type { target("dx.RawBuffer", <3 x i32>, 1, 0), target("dx.RawBuffer", <3 x i32>, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].10" = type { target("dx.RawBuffer", <2 x half>, 1, 0), target("dx.RawBuffer", <2 x half>, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].11" = type { target("dx.RawBuffer", <3 x float>, 1, 0), target("dx.RawBuffer", <3 x float>, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].12" = type { target("dx.RawBuffer", i32, 1, 0), target("dx.RawBuffer", i32, 1, 0) } +// DXIL-RW: %"class.hlsl::[[RESOURCE]].13" = type { target("dx.RawBuffer", <4 x i32>, 1, 0), target("dx.RawBuffer", <4 x i32>, 1, 0) } + +// SPV-RO: %"class.hlsl::[[RESOURCE]]" = type { target("spirv.VulkanBuffer", [0 x i16], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].0" = type { target("spirv.VulkanBuffer", [0 x i16], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].1" = type { target("spirv.VulkanBuffer", [0 x i32], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].2" = type { target("spirv.VulkanBuffer", [0 x i32], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].3" = type { target("spirv.VulkanBuffer", [0 x i64], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].4" = type { target("spirv.VulkanBuffer", [0 x i64], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].5" = type { target("spirv.VulkanBuffer", [0 x half], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].6" = type { target("spirv.VulkanBuffer", [0 x float], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].7" = type { target("spirv.VulkanBuffer", [0 x double], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].8" = type { target("spirv.VulkanBuffer", [0 x <4 x i16>], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].9" = type { target("spirv.VulkanBuffer", [0 x <3 x i32>], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].10" = type { target("spirv.VulkanBuffer", [0 x <2 x half>], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].11" = type { target("spirv.VulkanBuffer", [0 x <3 x float>], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].12" = type { target("spirv.VulkanBuffer", [0 x i32], 12, 0) } +// SPV-RO: %"class.hlsl::[[RESOURCE]].13" = type { target("spirv.VulkanBuffer", [0 x <4 x i32>], 12, 0) } + +// SPV-RW: %"class.hlsl::[[RESOURCE]]" = type { target("spirv.VulkanBuffer", [0 x i16], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].0" = type { target("spirv.VulkanBuffer", [0 x i16], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].1" = type { target("spirv.VulkanBuffer", [0 x i32], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].2" = type { target("spirv.VulkanBuffer", [0 x i32], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].3" = type { target("spirv.VulkanBuffer", [0 x i64], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].4" = type { target("spirv.VulkanBuffer", [0 x i64], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].5" = type { target("spirv.VulkanBuffer", [0 x half], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].6" = type { target("spirv.VulkanBuffer", [0 x float], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].7" = type { target("spirv.VulkanBuffer", [0 x double], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].8" = type { target("spirv.VulkanBuffer", [0 x <4 x i16>], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].9" = type { target("spirv.VulkanBuffer", [0 x <3 x i32>], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].10" = type { target("spirv.VulkanBuffer", [0 x <2 x half>], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].11" = type { target("spirv.VulkanBuffer", [0 x <3 x float>], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].12" = type { target("spirv.VulkanBuffer", [0 x i32], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } +// SPV-RW: %"class.hlsl::[[RESOURCE]].13" = type { target("spirv.VulkanBuffer", [0 x <4 x i32>], 12, 1), target("spirv.VulkanBuffer", i32, 12, 1) } + +RESOURCE<int16_t> BufI16; +RESOURCE<uint16_t> BufU16; +RESOURCE<int> BufI32; +RESOURCE<uint> BufU32; +RESOURCE<int64_t> BufI64; +RESOURCE<uint64_t> BufU64; +RESOURCE<half> BufF16; +RESOURCE<float> BufF32; +RESOURCE<double> BufF64; +RESOURCE< vector<int16_t, 4> > BufI16x4; +RESOURCE< vector<uint, 3> > BufU32x3; +RESOURCE<half2> BufF16x2; +RESOURCE<float3> BufF32x3; +RESOURCE<bool> BufBool; +RESOURCE<bool4> BufBoolVec; +// TODO: RESOURCE<snorm half> BufSNormF16; +// TODO: RESOURCE<unorm half> BufUNormF16; +// TODO: RESOURCE<snorm float> BufSNormF32; +// TODO: RESOURCE<unorm float> BufUNormF32; +// TODO: RESOURCE<snorm double> BufSNormF64; +// TODO: RESOURCE<unorm double> BufUNormF64; + +[numthreads(1,1,1)] +void main() { +} diff --git a/clang/test/Driver/dxc_frs.hlsl b/clang/test/Driver/dxc_frs.hlsl index 767cab6..ffc3886 100644 --- a/clang/test/Driver/dxc_frs.hlsl +++ b/clang/test/Driver/dxc_frs.hlsl @@ -1,10 +1,9 @@ -// RUN: %clang_dxc -T cs_6_0 /Fo %t.dxo /Frs %t.rs.dxo -### %s 2>&1 | FileCheck %s +// RUN: %clang_dxc -Vd -T cs_6_0 /Fo %t.dxo /Frs %t.rs.dxo -### %s 2>&1 | FileCheck %s // Test to demonstrate extracting the root signature to the specified // output file with /Frs. // CHECK: "{{.*}}llvm-objcopy{{(.exe)?}}" "{{.*}}.obj" "{{.*}}.dxo" "--extract-section=RTS0={{.*}}.rs.dxo" - [shader("compute"), RootSignature("")] [numthreads(1,1,1)] void EmptyEntry() {} diff --git a/clang/test/Driver/dxc_rootsignature_target.hlsl b/clang/test/Driver/dxc_rootsignature_target.hlsl index 08cd1ab..bb48063 100644 --- a/clang/test/Driver/dxc_rootsignature_target.hlsl +++ b/clang/test/Driver/dxc_rootsignature_target.hlsl @@ -1,4 +1,4 @@ -// RUN: %clang_dxc -E EntryRS -T rootsig_1_1 /Fo %t.dxo -### %s 2>&1 | FileCheck %s --check-prefix=CMDS +// RUN: %clang_dxc -Vd -E EntryRS -T rootsig_1_1 /Fo %t.dxo -### %s 2>&1 | FileCheck %s --check-prefix=CMDS // CMDS: "{{.*}}clang{{.*}}" "-cc1" // CMDS-SAME: "-triple" "dxilv1.1-unknown-shadermodel1.1-rootsignature" diff --git a/clang/test/Driver/hip-options.hip b/clang/test/Driver/hip-options.hip index 6206020..09f1ffa 100644 --- a/clang/test/Driver/hip-options.hip +++ b/clang/test/Driver/hip-options.hip @@ -254,3 +254,9 @@ // RUN: --offload-arch=gfx1100 --offload-new-driver --offload-jobs=0x4 %s 2>&1 | \ // RUN: FileCheck -check-prefix=INVJOBS %s // INVJOBS: clang: error: invalid integral value '0x4' in '--offload-jobs=0x4' + +// RUN: %clang -### -Werror --target=x86_64-unknown-linux-gnu -nogpuinc -nogpulib \ +// RUN: --offload-arch=gfx1100 --offload-new-driver --offload-jobs=jobserver %s 2>&1 | \ +// RUN: FileCheck -check-prefix=JOBSV %s +// JOBSV: clang-linker-wrapper{{.*}} "--wrapper-jobs=jobserver" + diff --git a/clang/test/Driver/linker-wrapper.c b/clang/test/Driver/linker-wrapper.c index c060dae..1c0fb96 100644 --- a/clang/test/Driver/linker-wrapper.c +++ b/clang/test/Driver/linker-wrapper.c @@ -114,6 +114,8 @@ __attribute__((visibility("protected"), used)) int x; // RUN: -fembed-offload-object=%t.out // RUN: clang-linker-wrapper --dry-run --host-triple=x86_64-unknown-linux-gnu --wrapper-jobs=4 \ // RUN: --linker-path=/usr/bin/ld %t.o -o a.out 2>&1 | FileCheck %s --check-prefix=CUDA-PAR +// RUN: clang-linker-wrapper --dry-run --host-triple=x86_64-unknown-linux-gnu --wrapper-jobs=jobserver \ +// RUN: --linker-path=/usr/bin/ld %t.o -o a.out 2>&1 | FileCheck %s --check-prefix=CUDA-PAR // CUDA-PAR: fatbinary{{.*}}-64 --create {{.*}}.fatbin diff --git a/clang/tools/clang-linker-wrapper/ClangLinkerWrapper.cpp b/clang/tools/clang-linker-wrapper/ClangLinkerWrapper.cpp index 1419b8c..4d5b956 100644 --- a/clang/tools/clang-linker-wrapper/ClangLinkerWrapper.cpp +++ b/clang/tools/clang-linker-wrapper/ClangLinkerWrapper.cpp @@ -1295,12 +1295,18 @@ int main(int Argc, char **Argv) { parallel::strategy = hardware_concurrency(1); if (auto *Arg = Args.getLastArg(OPT_wrapper_jobs)) { - unsigned Threads = 0; - if (!llvm::to_integer(Arg->getValue(), Threads) || Threads == 0) - reportError(createStringError("%s: expected a positive integer, got '%s'", - Arg->getSpelling().data(), - Arg->getValue())); - parallel::strategy = hardware_concurrency(Threads); + StringRef Val = Arg->getValue(); + if (Val.equals_insensitive("jobserver")) + parallel::strategy = jobserver_concurrency(); + else { + unsigned Threads = 0; + if (!llvm::to_integer(Val, Threads) || Threads == 0) + reportError(createStringError( + "%s: expected a positive integer or 'jobserver', got '%s'", + Arg->getSpelling().data(), Val.data())); + else + parallel::strategy = hardware_concurrency(Threads); + } } if (Args.hasArg(OPT_wrapper_time_trace_eq)) { diff --git a/clang/tools/clang-linker-wrapper/LinkerWrapperOpts.td b/clang/tools/clang-linker-wrapper/LinkerWrapperOpts.td index fa73e02..87f911c 100644 --- a/clang/tools/clang-linker-wrapper/LinkerWrapperOpts.td +++ b/clang/tools/clang-linker-wrapper/LinkerWrapperOpts.td @@ -53,7 +53,8 @@ def wrapper_time_trace_granularity : Joined<["--"], "wrapper-time-trace-granular def wrapper_jobs : Joined<["--"], "wrapper-jobs=">, Flags<[WrapperOnlyOption]>, MetaVarName<"<number>">, - HelpText<"Sets the number of parallel jobs to use for device linking">; + HelpText<"Sets the number of parallel jobs for device linking. Can be a " + "positive integer or 'jobserver'.">; def override_image : Joined<["--"], "override-image=">, Flags<[WrapperOnlyOption]>, MetaVarName<"<kind=file>">, diff --git a/clang/unittests/Format/FormatTestObjC.cpp b/clang/unittests/Format/FormatTestObjC.cpp index f7f73db..700d7cf8 100644 --- a/clang/unittests/Format/FormatTestObjC.cpp +++ b/clang/unittests/Format/FormatTestObjC.cpp @@ -763,6 +763,15 @@ TEST_F(FormatTestObjC, FormatObjCMethodExpr) { " backing:NSBackingStoreBuffered\n" " defer:NO]);\n" "}"); + Style.ColumnLimit = 63; + verifyFormat( + "- (void)test {\n" + " if ([object\n" + " respondsToSelector:@selector(\n" + " selectorName:param1:param2:)])\n" + " return;\n" + "}"); + Style.ColumnLimit = PreviousColumnLimit; verifyFormat("[contentsContainer replaceSubview:[subviews objectAtIndex:0]\n" " with:contentsNativeView];"); diff --git a/clang/unittests/Format/TokenAnnotatorTest.cpp b/clang/unittests/Format/TokenAnnotatorTest.cpp index 4a8f27f..c21b118 100644 --- a/clang/unittests/Format/TokenAnnotatorTest.cpp +++ b/clang/unittests/Format/TokenAnnotatorTest.cpp @@ -1929,6 +1929,37 @@ TEST_F(TokenAnnotatorTest, UnderstandsObjCMethodExpr) { ASSERT_EQ(Tokens.size(), 20u) << Tokens; EXPECT_TOKEN(Tokens[9], tok::l_square, TT_ObjCMethodExpr); EXPECT_TOKEN(Tokens[15], tok::greater, TT_BinaryOperator); + + Tokens = annotate("a = @selector(name:);"); + ASSERT_EQ(Tokens.size(), 10u) << Tokens; + EXPECT_TOKEN(Tokens[4], tok::l_paren, TT_ObjCSelector); + EXPECT_TOKEN(Tokens[6], tok::colon, TT_ObjCSelector); + EXPECT_TOKEN(Tokens[7], tok::r_paren, TT_ObjCSelector); + + Tokens = + annotate("[object respondsToSelector:@selector(name:param1:param2:)\n" + " respondsToSelector:@selector(name:param1:param2:)];"); + ASSERT_EQ(Tokens.size(), 29u) << Tokens; + EXPECT_TOKEN(Tokens[0], tok::l_square, TT_ObjCMethodExpr); + EXPECT_TOKEN(Tokens[3], tok::colon, TT_ObjCMethodExpr); + EXPECT_TOKEN(Tokens[6], tok::l_paren, TT_ObjCSelector); + EXPECT_TOKEN(Tokens[8], tok::colon, TT_ObjCSelector); + EXPECT_TOKEN(Tokens[10], tok::colon, TT_ObjCSelector); + EXPECT_TOKEN(Tokens[12], tok::colon, TT_ObjCSelector); + EXPECT_TOKEN(Tokens[13], tok::r_paren, TT_ObjCSelector); + EXPECT_TOKEN(Tokens[15], tok::colon, TT_ObjCMethodExpr); + EXPECT_TOKEN(Tokens[18], tok::l_paren, TT_ObjCSelector); + EXPECT_TOKEN(Tokens[20], tok::colon, TT_ObjCSelector); + EXPECT_TOKEN(Tokens[22], tok::colon, TT_ObjCSelector); + EXPECT_TOKEN(Tokens[24], tok::colon, TT_ObjCSelector); + EXPECT_TOKEN(Tokens[25], tok::r_paren, TT_ObjCSelector); + EXPECT_TOKEN(Tokens[26], tok::r_square, TT_ObjCMethodExpr); + + Tokens = annotate("[a b:c];"); + ASSERT_EQ(Tokens.size(), 8u) << Tokens; + EXPECT_TOKEN(Tokens[0], tok::l_square, TT_ObjCMethodExpr); + EXPECT_TOKEN(Tokens[3], tok::colon, TT_ObjCMethodExpr); + EXPECT_TOKEN(Tokens[5], tok::r_square, TT_ObjCMethodExpr); } TEST_F(TokenAnnotatorTest, UnderstandsObjCMethodDecl) { |