| // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --function-signature --include-generated-funcs --replace-value-regex "__omp_offloading_[0-9a-z]+_[0-9a-z]+" "reduction_size[.].+[.]" "pl_cond[.].+[.|,]" --prefix-filecheck-ir-name _ |
| // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-ibm-linux-gnu -emit-llvm %s -fexceptions -fcxx-exceptions -o - | FileCheck %s --check-prefix=CHECK1 |
| // RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -std=c++11 -triple powerpc64le-ibm-linux-gnu -fexceptions -fcxx-exceptions -emit-pch -o %t %s |
| // RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-ibm-linux-gnu -fexceptions -fcxx-exceptions -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK1 |
| |
| // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp-simd -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-ibm-linux-gnu -emit-llvm %s -fexceptions -fcxx-exceptions -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}" |
| // RUN: %clang_cc1 -no-opaque-pointers -fopenmp-simd -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -std=c++11 -triple powerpc64le-ibm-linux-gnu -fexceptions -fcxx-exceptions -emit-pch -o %t %s |
| // RUN: %clang_cc1 -no-opaque-pointers -fopenmp-simd -fopenmp-version=45 -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-ibm-linux-gnu -fexceptions -fcxx-exceptions -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}" |
| |
| // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-ibm-linux-gnu -emit-llvm %s -fexceptions -fcxx-exceptions -o - | FileCheck %s --check-prefix=CHECK5 |
| // RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -std=c++11 -triple powerpc64le-ibm-linux-gnu -fexceptions -fcxx-exceptions -emit-pch -o %t %s |
| // RUN: %clang_cc1 -no-opaque-pointers -fopenmp -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-ibm-linux-gnu -fexceptions -fcxx-exceptions -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix=CHECK5 |
| |
| // RUN: %clang_cc1 -no-opaque-pointers -verify -fopenmp-simd -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-ibm-linux-gnu -emit-llvm %s -fexceptions -fcxx-exceptions -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}" |
| // RUN: %clang_cc1 -no-opaque-pointers -fopenmp-simd -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -std=c++11 -triple powerpc64le-ibm-linux-gnu -fexceptions -fcxx-exceptions -emit-pch -o %t %s |
| // RUN: %clang_cc1 -no-opaque-pointers -fopenmp-simd -fopenmp-targets=powerpc64le-ibm-linux-gnu -x c++ -triple powerpc64le-ibm-linux-gnu -fexceptions -fcxx-exceptions -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --implicit-check-not="{{__kmpc|__tgt}}" |
| |
| // expected-no-diagnostics |
| #ifndef HEADER |
| #define HEADER |
| |
| typedef __INTPTR_TYPE__ intptr_t; |
| |
| |
| void foo(); |
| |
| struct S { |
| intptr_t a, b, c; |
| S(intptr_t a) : a(a) {} |
| operator char() { return a; } |
| ~S() {} |
| }; |
| |
| template <typename T, int C> |
| int tmain() { |
| #pragma omp target |
| #pragma omp teams distribute parallel for num_threads(C) |
| for (int i = 0; i < 100; i++) |
| foo(); |
| #pragma omp target |
| #pragma omp teams distribute parallel for num_threads(T(23)) |
| for (int i = 0; i < 100; i++) |
| foo(); |
| return 0; |
| } |
| |
| int main() { |
| S s(0); |
| char a = s; |
| #pragma omp target |
| #pragma omp teams distribute parallel for num_threads(2) |
| for (int i = 0; i < 100; i++) { |
| foo(); |
| } |
| #pragma omp target |
| |
| #pragma omp teams distribute parallel for num_threads(a) |
| for (int i = 0; i < 100; i++) { |
| foo(); |
| } |
| return a + tmain<char, 5>() + tmain<S, 1>(); |
| } |
| |
| // tmain 5 |
| |
| // tmain 1 |
| |
| |
| |
| |
| |
| |
| |
| |
| #endif |
| // CHECK1-LABEL: define {{[^@]+}}@main |
| // CHECK1-SAME: () #[[ATTR0:[0-9]+]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[RETVAL:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[S:%.*]] = alloca [[STRUCT_S:%.*]], align 8 |
| // CHECK1-NEXT: [[A:%.*]] = alloca i8, align 1 |
| // CHECK1-NEXT: [[EXN_SLOT:%.*]] = alloca i8*, align 8 |
| // CHECK1-NEXT: [[EHSELECTOR_SLOT:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[A_CASTED:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTOFFLOAD_BASEPTRS:%.*]] = alloca [1 x i8*], align 8 |
| // CHECK1-NEXT: [[DOTOFFLOAD_PTRS:%.*]] = alloca [1 x i8*], align 8 |
| // CHECK1-NEXT: [[DOTOFFLOAD_MAPPERS:%.*]] = alloca [1 x i8*], align 8 |
| // CHECK1-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i8, align 1 |
| // CHECK1-NEXT: [[_TMP1:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[RETVAL]], align 4 |
| // CHECK1-NEXT: call void @_ZN1SC1El(%struct.S* noundef nonnull align 8 dereferenceable(24) [[S]], i64 noundef 0) |
| // CHECK1-NEXT: [[CALL:%.*]] = invoke noundef signext i8 @_ZN1ScvcEv(%struct.S* noundef nonnull align 8 dereferenceable(24) [[S]]) |
| // CHECK1-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[LPAD:%.*]] |
| // CHECK1: invoke.cont: |
| // CHECK1-NEXT: store i8 [[CALL]], i8* [[A]], align 1 |
| // CHECK1-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8 |
| // CHECK1-NEXT: [[TMP0:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 0 |
| // CHECK1-NEXT: store i32 1, i32* [[TMP0]], align 4 |
| // CHECK1-NEXT: [[TMP1:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 1 |
| // CHECK1-NEXT: store i32 0, i32* [[TMP1]], align 4 |
| // CHECK1-NEXT: [[TMP2:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 2 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP2]], align 8 |
| // CHECK1-NEXT: [[TMP3:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 3 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP3]], align 8 |
| // CHECK1-NEXT: [[TMP4:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 4 |
| // CHECK1-NEXT: store i64* null, i64** [[TMP4]], align 8 |
| // CHECK1-NEXT: [[TMP5:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 5 |
| // CHECK1-NEXT: store i64* null, i64** [[TMP5]], align 8 |
| // CHECK1-NEXT: [[TMP6:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 6 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP6]], align 8 |
| // CHECK1-NEXT: [[TMP7:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 7 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP7]], align 8 |
| // CHECK1-NEXT: [[TMP8:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 8 |
| // CHECK1-NEXT: store i64 100, i64* [[TMP8]], align 8 |
| // CHECK1-NEXT: [[TMP9:%.*]] = call i32 @__tgt_target_kernel(%struct.ident_t* @[[GLOB3:[0-9]+]], i64 -1, i32 0, i32 2, i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l50.region_id, %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]]) |
| // CHECK1-NEXT: [[TMP10:%.*]] = icmp ne i32 [[TMP9]], 0 |
| // CHECK1-NEXT: br i1 [[TMP10]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]] |
| // CHECK1: omp_offload.failed: |
| // CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l50() #[[ATTR6:[0-9]+]] |
| // CHECK1-NEXT: br label [[OMP_OFFLOAD_CONT]] |
| // CHECK1: lpad: |
| // CHECK1-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK1-NEXT: cleanup |
| // CHECK1-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK1-NEXT: store i8* [[TMP12]], i8** [[EXN_SLOT]], align 8 |
| // CHECK1-NEXT: [[TMP13:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 1 |
| // CHECK1-NEXT: store i32 [[TMP13]], i32* [[EHSELECTOR_SLOT]], align 4 |
| // CHECK1-NEXT: call void @_ZN1SD1Ev(%struct.S* noundef nonnull align 8 dereferenceable(24) [[S]]) #[[ATTR6]] |
| // CHECK1-NEXT: br label [[EH_RESUME:%.*]] |
| // CHECK1: omp_offload.cont: |
| // CHECK1-NEXT: [[TMP14:%.*]] = load i8, i8* [[A]], align 1 |
| // CHECK1-NEXT: [[CONV:%.*]] = bitcast i64* [[A_CASTED]] to i8* |
| // CHECK1-NEXT: store i8 [[TMP14]], i8* [[CONV]], align 1 |
| // CHECK1-NEXT: [[TMP15:%.*]] = load i64, i64* [[A_CASTED]], align 8 |
| // CHECK1-NEXT: [[TMP16:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0 |
| // CHECK1-NEXT: [[TMP17:%.*]] = bitcast i8** [[TMP16]] to i64* |
| // CHECK1-NEXT: store i64 [[TMP15]], i64* [[TMP17]], align 8 |
| // CHECK1-NEXT: [[TMP18:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[DOTOFFLOAD_PTRS]], i32 0, i32 0 |
| // CHECK1-NEXT: [[TMP19:%.*]] = bitcast i8** [[TMP18]] to i64* |
| // CHECK1-NEXT: store i64 [[TMP15]], i64* [[TMP19]], align 8 |
| // CHECK1-NEXT: [[TMP20:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[DOTOFFLOAD_MAPPERS]], i64 0, i64 0 |
| // CHECK1-NEXT: store i8* null, i8** [[TMP20]], align 8 |
| // CHECK1-NEXT: [[TMP21:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0 |
| // CHECK1-NEXT: [[TMP22:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[DOTOFFLOAD_PTRS]], i32 0, i32 0 |
| // CHECK1-NEXT: [[TMP23:%.*]] = load i8, i8* [[A]], align 1 |
| // CHECK1-NEXT: store i8 [[TMP23]], i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK1-NEXT: [[TMP24:%.*]] = load i8, i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK1-NEXT: [[TMP25:%.*]] = zext i8 [[TMP24]] to i32 |
| // CHECK1-NEXT: [[KERNEL_ARGS2:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8 |
| // CHECK1-NEXT: [[TMP26:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 0 |
| // CHECK1-NEXT: store i32 1, i32* [[TMP26]], align 4 |
| // CHECK1-NEXT: [[TMP27:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 1 |
| // CHECK1-NEXT: store i32 1, i32* [[TMP27]], align 4 |
| // CHECK1-NEXT: [[TMP28:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 2 |
| // CHECK1-NEXT: store i8** [[TMP21]], i8*** [[TMP28]], align 8 |
| // CHECK1-NEXT: [[TMP29:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 3 |
| // CHECK1-NEXT: store i8** [[TMP22]], i8*** [[TMP29]], align 8 |
| // CHECK1-NEXT: [[TMP30:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 4 |
| // CHECK1-NEXT: store i64* getelementptr inbounds ([1 x i64], [1 x i64]* @.offload_sizes, i32 0, i32 0), i64** [[TMP30]], align 8 |
| // CHECK1-NEXT: [[TMP31:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 5 |
| // CHECK1-NEXT: store i64* getelementptr inbounds ([1 x i64], [1 x i64]* @.offload_maptypes, i32 0, i32 0), i64** [[TMP31]], align 8 |
| // CHECK1-NEXT: [[TMP32:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 6 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP32]], align 8 |
| // CHECK1-NEXT: [[TMP33:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 7 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP33]], align 8 |
| // CHECK1-NEXT: [[TMP34:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 8 |
| // CHECK1-NEXT: store i64 100, i64* [[TMP34]], align 8 |
| // CHECK1-NEXT: [[TMP35:%.*]] = call i32 @__tgt_target_kernel(%struct.ident_t* @[[GLOB3]], i64 -1, i32 0, i32 [[TMP25]], i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l55.region_id, %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]]) |
| // CHECK1-NEXT: [[TMP36:%.*]] = icmp ne i32 [[TMP35]], 0 |
| // CHECK1-NEXT: br i1 [[TMP36]], label [[OMP_OFFLOAD_FAILED3:%.*]], label [[OMP_OFFLOAD_CONT4:%.*]] |
| // CHECK1: omp_offload.failed3: |
| // CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l55(i64 [[TMP15]]) #[[ATTR6]] |
| // CHECK1-NEXT: br label [[OMP_OFFLOAD_CONT4]] |
| // CHECK1: omp_offload.cont4: |
| // CHECK1-NEXT: [[TMP37:%.*]] = load i8, i8* [[A]], align 1 |
| // CHECK1-NEXT: [[CONV5:%.*]] = sext i8 [[TMP37]] to i32 |
| // CHECK1-NEXT: [[CALL7:%.*]] = invoke noundef signext i32 @_Z5tmainIcLi5EEiv() |
| // CHECK1-NEXT: to label [[INVOKE_CONT6:%.*]] unwind label [[LPAD]] |
| // CHECK1: invoke.cont6: |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 [[CONV5]], [[CALL7]] |
| // CHECK1-NEXT: [[CALL9:%.*]] = invoke noundef signext i32 @_Z5tmainI1SLi1EEiv() |
| // CHECK1-NEXT: to label [[INVOKE_CONT8:%.*]] unwind label [[LPAD]] |
| // CHECK1: invoke.cont8: |
| // CHECK1-NEXT: [[ADD10:%.*]] = add nsw i32 [[ADD]], [[CALL9]] |
| // CHECK1-NEXT: store i32 [[ADD10]], i32* [[RETVAL]], align 4 |
| // CHECK1-NEXT: call void @_ZN1SD1Ev(%struct.S* noundef nonnull align 8 dereferenceable(24) [[S]]) #[[ATTR6]] |
| // CHECK1-NEXT: [[TMP38:%.*]] = load i32, i32* [[RETVAL]], align 4 |
| // CHECK1-NEXT: ret i32 [[TMP38]] |
| // CHECK1: eh.resume: |
| // CHECK1-NEXT: [[EXN:%.*]] = load i8*, i8** [[EXN_SLOT]], align 8 |
| // CHECK1-NEXT: [[SEL:%.*]] = load i32, i32* [[EHSELECTOR_SLOT]], align 4 |
| // CHECK1-NEXT: [[LPAD_VAL:%.*]] = insertvalue { i8*, i32 } undef, i8* [[EXN]], 0 |
| // CHECK1-NEXT: [[LPAD_VAL11:%.*]] = insertvalue { i8*, i32 } [[LPAD_VAL]], i32 [[SEL]], 1 |
| // CHECK1-NEXT: resume { i8*, i32 } [[LPAD_VAL11]] |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@_ZN1SC1El |
| // CHECK1-SAME: (%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS:%.*]], i64 noundef [[A:%.*]]) unnamed_addr #[[ATTR1:[0-9]+]] comdat align 2 { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[THIS_ADDR:%.*]] = alloca %struct.S*, align 8 |
| // CHECK1-NEXT: [[A_ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: store %struct.S* [[THIS]], %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[A]], i64* [[A_ADDR]], align 8 |
| // CHECK1-NEXT: [[THIS1:%.*]] = load %struct.S*, %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i64, i64* [[A_ADDR]], align 8 |
| // CHECK1-NEXT: call void @_ZN1SC2El(%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS1]], i64 noundef [[TMP0]]) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@_ZN1ScvcEv |
| // CHECK1-SAME: (%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS:%.*]]) #[[ATTR2:[0-9]+]] comdat align 2 { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[THIS_ADDR:%.*]] = alloca %struct.S*, align 8 |
| // CHECK1-NEXT: store %struct.S* [[THIS]], %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK1-NEXT: [[THIS1:%.*]] = load %struct.S*, %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK1-NEXT: [[A:%.*]] = getelementptr inbounds [[STRUCT_S:%.*]], %struct.S* [[THIS1]], i32 0, i32 0 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i64, i64* [[A]], align 8 |
| // CHECK1-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i8 |
| // CHECK1-NEXT: ret i8 [[CONV]] |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l50 |
| // CHECK1-SAME: () #[[ATTR3:[0-9]+]] { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 0, void (i32*, i32*, ...)* bitcast (void (i32*, i32*)* @.omp_outlined. to void (i32*, i32*, ...)*)) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_outlined. |
| // CHECK1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR3]] { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: store i32 99, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4 |
| // CHECK1-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1:[0-9]+]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99 |
| // CHECK1-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK1: cond.true: |
| // CHECK1-NEXT: br label [[COND_END:%.*]] |
| // CHECK1: cond.false: |
| // CHECK1-NEXT: [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: br label [[COND_END]] |
| // CHECK1: cond.end: |
| // CHECK1-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ] |
| // CHECK1-NEXT: store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK1: omp.inner.for.cond: |
| // CHECK1-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]] |
| // CHECK1-NEXT: br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK1: omp.inner.for.body: |
| // CHECK1-NEXT: call void @__kmpc_push_num_threads(%struct.ident_t* @[[GLOB3]], i32 [[TMP1]], i32 2) |
| // CHECK1-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: [[TMP8:%.*]] = zext i32 [[TMP7]] to i64 |
| // CHECK1-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[TMP10:%.*]] = zext i32 [[TMP9]] to i64 |
| // CHECK1-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..1 to void (i32*, i32*, ...)*), i64 [[TMP8]], i64 [[TMP10]]) |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK1: omp.inner.for.inc: |
| // CHECK1-NEXT: [[TMP11:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP12:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP11]], [[TMP12]] |
| // CHECK1-NEXT: store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK1: omp.inner.for.end: |
| // CHECK1-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK1: omp.loop.exit: |
| // CHECK1-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]]) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_outlined..1 |
| // CHECK1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_LB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_UB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 99, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i32 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32 |
| // CHECK1-NEXT: store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4 |
| // CHECK1-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2:[0-9]+]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK1-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99 |
| // CHECK1-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK1: cond.true: |
| // CHECK1-NEXT: br label [[COND_END:%.*]] |
| // CHECK1: cond.false: |
| // CHECK1-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: br label [[COND_END]] |
| // CHECK1: cond.end: |
| // CHECK1-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ] |
| // CHECK1-NEXT: store i32 [[COND]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK1: omp.inner.for.cond: |
| // CHECK1-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]] |
| // CHECK1-NEXT: br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK1: omp.inner.for.body: |
| // CHECK1-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1 |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 0, [[MUL]] |
| // CHECK1-NEXT: store i32 [[ADD]], i32* [[I]], align 4 |
| // CHECK1-NEXT: invoke void @_Z3foov() |
| // CHECK1-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK1: invoke.cont: |
| // CHECK1-NEXT: br label [[OMP_BODY_CONTINUE:%.*]] |
| // CHECK1: omp.body.continue: |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK1: omp.inner.for.inc: |
| // CHECK1-NEXT: [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1 |
| // CHECK1-NEXT: store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK1: omp.inner.for.end: |
| // CHECK1-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK1: omp.loop.exit: |
| // CHECK1-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]]) |
| // CHECK1-NEXT: ret void |
| // CHECK1: terminate.lpad: |
| // CHECK1-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK1-NEXT: catch i8* null |
| // CHECK1-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK1-NEXT: call void @__clang_call_terminate(i8* [[TMP12]]) #[[ATTR9:[0-9]+]] |
| // CHECK1-NEXT: unreachable |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@__clang_call_terminate |
| // CHECK1-SAME: (i8* [[TMP0:%.*]]) #[[ATTR5:[0-9]+]] comdat { |
| // CHECK1-NEXT: [[TMP2:%.*]] = call i8* @__cxa_begin_catch(i8* [[TMP0]]) #[[ATTR6]] |
| // CHECK1-NEXT: call void @_ZSt9terminatev() #[[ATTR9]] |
| // CHECK1-NEXT: unreachable |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l55 |
| // CHECK1-SAME: (i64 noundef [[A:%.*]]) #[[ATTR3]] { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[A_ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i8, align 1 |
| // CHECK1-NEXT: [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: store i64 [[A]], i64* [[A_ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV:%.*]] = bitcast i64* [[A_ADDR]] to i8* |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i8, i8* [[CONV]], align 1 |
| // CHECK1-NEXT: store i8 [[TMP0]], i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i8, i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK1-NEXT: [[CONV1:%.*]] = bitcast i64* [[DOTCAPTURE_EXPR__CASTED]] to i8* |
| // CHECK1-NEXT: store i8 [[TMP1]], i8* [[CONV1]], align 1 |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i64, i64* [[DOTCAPTURE_EXPR__CASTED]], align 8 |
| // CHECK1-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 1, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64)* @.omp_outlined..2 to void (i32*, i32*, ...)*), i64 [[TMP2]]) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_outlined..2 |
| // CHECK1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR3]] { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTCAPTURE_EXPR_]], i64* [[DOTCAPTURE_EXPR__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV:%.*]] = bitcast i64* [[DOTCAPTURE_EXPR__ADDR]] to i8* |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: store i32 99, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4 |
| // CHECK1-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99 |
| // CHECK1-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK1: cond.true: |
| // CHECK1-NEXT: br label [[COND_END:%.*]] |
| // CHECK1: cond.false: |
| // CHECK1-NEXT: [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: br label [[COND_END]] |
| // CHECK1: cond.end: |
| // CHECK1-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ] |
| // CHECK1-NEXT: store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK1: omp.inner.for.cond: |
| // CHECK1-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]] |
| // CHECK1-NEXT: br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK1: omp.inner.for.body: |
| // CHECK1-NEXT: [[TMP7:%.*]] = load i8, i8* [[CONV]], align 1 |
| // CHECK1-NEXT: [[TMP8:%.*]] = sext i8 [[TMP7]] to i32 |
| // CHECK1-NEXT: call void @__kmpc_push_num_threads(%struct.ident_t* @[[GLOB3]], i32 [[TMP1]], i32 [[TMP8]]) |
| // CHECK1-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: [[TMP10:%.*]] = zext i32 [[TMP9]] to i64 |
| // CHECK1-NEXT: [[TMP11:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[TMP12:%.*]] = zext i32 [[TMP11]] to i64 |
| // CHECK1-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..3 to void (i32*, i32*, ...)*), i64 [[TMP10]], i64 [[TMP12]]) |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK1: omp.inner.for.inc: |
| // CHECK1-NEXT: [[TMP13:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP14:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP13]], [[TMP14]] |
| // CHECK1-NEXT: store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK1: omp.inner.for.end: |
| // CHECK1-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK1: omp.loop.exit: |
| // CHECK1-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]]) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_outlined..3 |
| // CHECK1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_LB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_UB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 99, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i32 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32 |
| // CHECK1-NEXT: store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4 |
| // CHECK1-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK1-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99 |
| // CHECK1-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK1: cond.true: |
| // CHECK1-NEXT: br label [[COND_END:%.*]] |
| // CHECK1: cond.false: |
| // CHECK1-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: br label [[COND_END]] |
| // CHECK1: cond.end: |
| // CHECK1-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ] |
| // CHECK1-NEXT: store i32 [[COND]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK1: omp.inner.for.cond: |
| // CHECK1-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]] |
| // CHECK1-NEXT: br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK1: omp.inner.for.body: |
| // CHECK1-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1 |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 0, [[MUL]] |
| // CHECK1-NEXT: store i32 [[ADD]], i32* [[I]], align 4 |
| // CHECK1-NEXT: invoke void @_Z3foov() |
| // CHECK1-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK1: invoke.cont: |
| // CHECK1-NEXT: br label [[OMP_BODY_CONTINUE:%.*]] |
| // CHECK1: omp.body.continue: |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK1: omp.inner.for.inc: |
| // CHECK1-NEXT: [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1 |
| // CHECK1-NEXT: store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK1: omp.inner.for.end: |
| // CHECK1-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK1: omp.loop.exit: |
| // CHECK1-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]]) |
| // CHECK1-NEXT: ret void |
| // CHECK1: terminate.lpad: |
| // CHECK1-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK1-NEXT: catch i8* null |
| // CHECK1-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK1-NEXT: call void @__clang_call_terminate(i8* [[TMP12]]) #[[ATTR9]] |
| // CHECK1-NEXT: unreachable |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@_Z5tmainIcLi5EEiv |
| // CHECK1-SAME: () #[[ATTR2]] comdat { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[_TMP1:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8 |
| // CHECK1-NEXT: [[TMP0:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 0 |
| // CHECK1-NEXT: store i32 1, i32* [[TMP0]], align 4 |
| // CHECK1-NEXT: [[TMP1:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 1 |
| // CHECK1-NEXT: store i32 0, i32* [[TMP1]], align 4 |
| // CHECK1-NEXT: [[TMP2:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 2 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP2]], align 8 |
| // CHECK1-NEXT: [[TMP3:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 3 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP3]], align 8 |
| // CHECK1-NEXT: [[TMP4:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 4 |
| // CHECK1-NEXT: store i64* null, i64** [[TMP4]], align 8 |
| // CHECK1-NEXT: [[TMP5:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 5 |
| // CHECK1-NEXT: store i64* null, i64** [[TMP5]], align 8 |
| // CHECK1-NEXT: [[TMP6:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 6 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP6]], align 8 |
| // CHECK1-NEXT: [[TMP7:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 7 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP7]], align 8 |
| // CHECK1-NEXT: [[TMP8:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 8 |
| // CHECK1-NEXT: store i64 100, i64* [[TMP8]], align 8 |
| // CHECK1-NEXT: [[TMP9:%.*]] = call i32 @__tgt_target_kernel(%struct.ident_t* @[[GLOB3]], i64 -1, i32 0, i32 5, i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIcLi5EEiv_l36.region_id, %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]]) |
| // CHECK1-NEXT: [[TMP10:%.*]] = icmp ne i32 [[TMP9]], 0 |
| // CHECK1-NEXT: br i1 [[TMP10]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]] |
| // CHECK1: omp_offload.failed: |
| // CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIcLi5EEiv_l36() #[[ATTR6]] |
| // CHECK1-NEXT: br label [[OMP_OFFLOAD_CONT]] |
| // CHECK1: omp_offload.cont: |
| // CHECK1-NEXT: [[KERNEL_ARGS2:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8 |
| // CHECK1-NEXT: [[TMP11:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 0 |
| // CHECK1-NEXT: store i32 1, i32* [[TMP11]], align 4 |
| // CHECK1-NEXT: [[TMP12:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 1 |
| // CHECK1-NEXT: store i32 0, i32* [[TMP12]], align 4 |
| // CHECK1-NEXT: [[TMP13:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 2 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP13]], align 8 |
| // CHECK1-NEXT: [[TMP14:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 3 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP14]], align 8 |
| // CHECK1-NEXT: [[TMP15:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 4 |
| // CHECK1-NEXT: store i64* null, i64** [[TMP15]], align 8 |
| // CHECK1-NEXT: [[TMP16:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 5 |
| // CHECK1-NEXT: store i64* null, i64** [[TMP16]], align 8 |
| // CHECK1-NEXT: [[TMP17:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 6 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP17]], align 8 |
| // CHECK1-NEXT: [[TMP18:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 7 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP18]], align 8 |
| // CHECK1-NEXT: [[TMP19:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 8 |
| // CHECK1-NEXT: store i64 100, i64* [[TMP19]], align 8 |
| // CHECK1-NEXT: [[TMP20:%.*]] = call i32 @__tgt_target_kernel(%struct.ident_t* @[[GLOB3]], i64 -1, i32 0, i32 23, i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIcLi5EEiv_l40.region_id, %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]]) |
| // CHECK1-NEXT: [[TMP21:%.*]] = icmp ne i32 [[TMP20]], 0 |
| // CHECK1-NEXT: br i1 [[TMP21]], label [[OMP_OFFLOAD_FAILED3:%.*]], label [[OMP_OFFLOAD_CONT4:%.*]] |
| // CHECK1: omp_offload.failed3: |
| // CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIcLi5EEiv_l40() #[[ATTR6]] |
| // CHECK1-NEXT: br label [[OMP_OFFLOAD_CONT4]] |
| // CHECK1: omp_offload.cont4: |
| // CHECK1-NEXT: ret i32 0 |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@_Z5tmainI1SLi1EEiv |
| // CHECK1-SAME: () #[[ATTR2]] comdat personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i8, align 1 |
| // CHECK1-NEXT: [[REF_TMP:%.*]] = alloca [[STRUCT_S:%.*]], align 8 |
| // CHECK1-NEXT: [[_TMP1:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8 |
| // CHECK1-NEXT: [[TMP0:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 0 |
| // CHECK1-NEXT: store i32 1, i32* [[TMP0]], align 4 |
| // CHECK1-NEXT: [[TMP1:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 1 |
| // CHECK1-NEXT: store i32 0, i32* [[TMP1]], align 4 |
| // CHECK1-NEXT: [[TMP2:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 2 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP2]], align 8 |
| // CHECK1-NEXT: [[TMP3:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 3 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP3]], align 8 |
| // CHECK1-NEXT: [[TMP4:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 4 |
| // CHECK1-NEXT: store i64* null, i64** [[TMP4]], align 8 |
| // CHECK1-NEXT: [[TMP5:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 5 |
| // CHECK1-NEXT: store i64* null, i64** [[TMP5]], align 8 |
| // CHECK1-NEXT: [[TMP6:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 6 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP6]], align 8 |
| // CHECK1-NEXT: [[TMP7:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 7 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP7]], align 8 |
| // CHECK1-NEXT: [[TMP8:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 8 |
| // CHECK1-NEXT: store i64 100, i64* [[TMP8]], align 8 |
| // CHECK1-NEXT: [[TMP9:%.*]] = call i32 @__tgt_target_kernel(%struct.ident_t* @[[GLOB3]], i64 -1, i32 0, i32 1, i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainI1SLi1EEiv_l36.region_id, %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]]) |
| // CHECK1-NEXT: [[TMP10:%.*]] = icmp ne i32 [[TMP9]], 0 |
| // CHECK1-NEXT: br i1 [[TMP10]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]] |
| // CHECK1: omp_offload.failed: |
| // CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainI1SLi1EEiv_l36() #[[ATTR6]] |
| // CHECK1-NEXT: br label [[OMP_OFFLOAD_CONT]] |
| // CHECK1: omp_offload.cont: |
| // CHECK1-NEXT: invoke void @_ZN1SC1El(%struct.S* noundef nonnull align 8 dereferenceable(24) [[REF_TMP]], i64 noundef 23) |
| // CHECK1-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK1: invoke.cont: |
| // CHECK1-NEXT: [[CALL:%.*]] = call noundef signext i8 @_ZN1ScvcEv(%struct.S* noundef nonnull align 8 dereferenceable(24) [[REF_TMP]]) |
| // CHECK1-NEXT: call void @_ZN1SD1Ev(%struct.S* noundef nonnull align 8 dereferenceable(24) [[REF_TMP]]) #[[ATTR6]] |
| // CHECK1-NEXT: store i8 [[CALL]], i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK1-NEXT: [[TMP11:%.*]] = load i8, i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK1-NEXT: [[TMP12:%.*]] = zext i8 [[TMP11]] to i32 |
| // CHECK1-NEXT: [[KERNEL_ARGS2:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8 |
| // CHECK1-NEXT: [[TMP13:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 0 |
| // CHECK1-NEXT: store i32 1, i32* [[TMP13]], align 4 |
| // CHECK1-NEXT: [[TMP14:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 1 |
| // CHECK1-NEXT: store i32 0, i32* [[TMP14]], align 4 |
| // CHECK1-NEXT: [[TMP15:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 2 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP15]], align 8 |
| // CHECK1-NEXT: [[TMP16:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 3 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP16]], align 8 |
| // CHECK1-NEXT: [[TMP17:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 4 |
| // CHECK1-NEXT: store i64* null, i64** [[TMP17]], align 8 |
| // CHECK1-NEXT: [[TMP18:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 5 |
| // CHECK1-NEXT: store i64* null, i64** [[TMP18]], align 8 |
| // CHECK1-NEXT: [[TMP19:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 6 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP19]], align 8 |
| // CHECK1-NEXT: [[TMP20:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 7 |
| // CHECK1-NEXT: store i8** null, i8*** [[TMP20]], align 8 |
| // CHECK1-NEXT: [[TMP21:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 8 |
| // CHECK1-NEXT: store i64 100, i64* [[TMP21]], align 8 |
| // CHECK1-NEXT: [[TMP22:%.*]] = call i32 @__tgt_target_kernel(%struct.ident_t* @[[GLOB3]], i64 -1, i32 0, i32 [[TMP12]], i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainI1SLi1EEiv_l40.region_id, %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]]) |
| // CHECK1-NEXT: [[TMP23:%.*]] = icmp ne i32 [[TMP22]], 0 |
| // CHECK1-NEXT: br i1 [[TMP23]], label [[OMP_OFFLOAD_FAILED3:%.*]], label [[OMP_OFFLOAD_CONT4:%.*]] |
| // CHECK1: omp_offload.failed3: |
| // CHECK1-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainI1SLi1EEiv_l40() #[[ATTR6]] |
| // CHECK1-NEXT: br label [[OMP_OFFLOAD_CONT4]] |
| // CHECK1: omp_offload.cont4: |
| // CHECK1-NEXT: ret i32 0 |
| // CHECK1: terminate.lpad: |
| // CHECK1-NEXT: [[TMP24:%.*]] = landingpad { i8*, i32 } |
| // CHECK1-NEXT: catch i8* null |
| // CHECK1-NEXT: [[TMP25:%.*]] = extractvalue { i8*, i32 } [[TMP24]], 0 |
| // CHECK1-NEXT: call void @__clang_call_terminate(i8* [[TMP25]]) #[[ATTR9]] |
| // CHECK1-NEXT: unreachable |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@_ZN1SD1Ev |
| // CHECK1-SAME: (%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS:%.*]]) unnamed_addr #[[ATTR7:[0-9]+]] comdat align 2 { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[THIS_ADDR:%.*]] = alloca %struct.S*, align 8 |
| // CHECK1-NEXT: store %struct.S* [[THIS]], %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK1-NEXT: [[THIS1:%.*]] = load %struct.S*, %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK1-NEXT: call void @_ZN1SD2Ev(%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS1]]) #[[ATTR6]] |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@_ZN1SC2El |
| // CHECK1-SAME: (%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS:%.*]], i64 noundef [[A:%.*]]) unnamed_addr #[[ATTR7]] comdat align 2 { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[THIS_ADDR:%.*]] = alloca %struct.S*, align 8 |
| // CHECK1-NEXT: [[A_ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: store %struct.S* [[THIS]], %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[A]], i64* [[A_ADDR]], align 8 |
| // CHECK1-NEXT: [[THIS1:%.*]] = load %struct.S*, %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK1-NEXT: [[A2:%.*]] = getelementptr inbounds [[STRUCT_S:%.*]], %struct.S* [[THIS1]], i32 0, i32 0 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i64, i64* [[A_ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[TMP0]], i64* [[A2]], align 8 |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@_ZN1SD2Ev |
| // CHECK1-SAME: (%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS:%.*]]) unnamed_addr #[[ATTR7]] comdat align 2 { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[THIS_ADDR:%.*]] = alloca %struct.S*, align 8 |
| // CHECK1-NEXT: store %struct.S* [[THIS]], %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK1-NEXT: [[THIS1:%.*]] = load %struct.S*, %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIcLi5EEiv_l36 |
| // CHECK1-SAME: () #[[ATTR3]] { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 0, void (i32*, i32*, ...)* bitcast (void (i32*, i32*)* @.omp_outlined..4 to void (i32*, i32*, ...)*)) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_outlined..4 |
| // CHECK1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR3]] { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: store i32 99, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4 |
| // CHECK1-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99 |
| // CHECK1-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK1: cond.true: |
| // CHECK1-NEXT: br label [[COND_END:%.*]] |
| // CHECK1: cond.false: |
| // CHECK1-NEXT: [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: br label [[COND_END]] |
| // CHECK1: cond.end: |
| // CHECK1-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ] |
| // CHECK1-NEXT: store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK1: omp.inner.for.cond: |
| // CHECK1-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]] |
| // CHECK1-NEXT: br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK1: omp.inner.for.body: |
| // CHECK1-NEXT: call void @__kmpc_push_num_threads(%struct.ident_t* @[[GLOB3]], i32 [[TMP1]], i32 5) |
| // CHECK1-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: [[TMP8:%.*]] = zext i32 [[TMP7]] to i64 |
| // CHECK1-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[TMP10:%.*]] = zext i32 [[TMP9]] to i64 |
| // CHECK1-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..5 to void (i32*, i32*, ...)*), i64 [[TMP8]], i64 [[TMP10]]) |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK1: omp.inner.for.inc: |
| // CHECK1-NEXT: [[TMP11:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP12:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP11]], [[TMP12]] |
| // CHECK1-NEXT: store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK1: omp.inner.for.end: |
| // CHECK1-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK1: omp.loop.exit: |
| // CHECK1-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]]) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_outlined..5 |
| // CHECK1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_LB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_UB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 99, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i32 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32 |
| // CHECK1-NEXT: store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4 |
| // CHECK1-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK1-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99 |
| // CHECK1-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK1: cond.true: |
| // CHECK1-NEXT: br label [[COND_END:%.*]] |
| // CHECK1: cond.false: |
| // CHECK1-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: br label [[COND_END]] |
| // CHECK1: cond.end: |
| // CHECK1-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ] |
| // CHECK1-NEXT: store i32 [[COND]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK1: omp.inner.for.cond: |
| // CHECK1-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]] |
| // CHECK1-NEXT: br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK1: omp.inner.for.body: |
| // CHECK1-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1 |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 0, [[MUL]] |
| // CHECK1-NEXT: store i32 [[ADD]], i32* [[I]], align 4 |
| // CHECK1-NEXT: invoke void @_Z3foov() |
| // CHECK1-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK1: invoke.cont: |
| // CHECK1-NEXT: br label [[OMP_BODY_CONTINUE:%.*]] |
| // CHECK1: omp.body.continue: |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK1: omp.inner.for.inc: |
| // CHECK1-NEXT: [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1 |
| // CHECK1-NEXT: store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK1: omp.inner.for.end: |
| // CHECK1-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK1: omp.loop.exit: |
| // CHECK1-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]]) |
| // CHECK1-NEXT: ret void |
| // CHECK1: terminate.lpad: |
| // CHECK1-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK1-NEXT: catch i8* null |
| // CHECK1-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK1-NEXT: call void @__clang_call_terminate(i8* [[TMP12]]) #[[ATTR9]] |
| // CHECK1-NEXT: unreachable |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIcLi5EEiv_l40 |
| // CHECK1-SAME: () #[[ATTR3]] { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 0, void (i32*, i32*, ...)* bitcast (void (i32*, i32*)* @.omp_outlined..6 to void (i32*, i32*, ...)*)) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_outlined..6 |
| // CHECK1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR3]] { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: store i32 99, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4 |
| // CHECK1-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99 |
| // CHECK1-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK1: cond.true: |
| // CHECK1-NEXT: br label [[COND_END:%.*]] |
| // CHECK1: cond.false: |
| // CHECK1-NEXT: [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: br label [[COND_END]] |
| // CHECK1: cond.end: |
| // CHECK1-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ] |
| // CHECK1-NEXT: store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK1: omp.inner.for.cond: |
| // CHECK1-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]] |
| // CHECK1-NEXT: br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK1: omp.inner.for.body: |
| // CHECK1-NEXT: call void @__kmpc_push_num_threads(%struct.ident_t* @[[GLOB3]], i32 [[TMP1]], i32 23) |
| // CHECK1-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: [[TMP8:%.*]] = zext i32 [[TMP7]] to i64 |
| // CHECK1-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[TMP10:%.*]] = zext i32 [[TMP9]] to i64 |
| // CHECK1-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..7 to void (i32*, i32*, ...)*), i64 [[TMP8]], i64 [[TMP10]]) |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK1: omp.inner.for.inc: |
| // CHECK1-NEXT: [[TMP11:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP12:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP11]], [[TMP12]] |
| // CHECK1-NEXT: store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK1: omp.inner.for.end: |
| // CHECK1-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK1: omp.loop.exit: |
| // CHECK1-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]]) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_outlined..7 |
| // CHECK1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_LB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_UB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 99, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i32 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32 |
| // CHECK1-NEXT: store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4 |
| // CHECK1-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK1-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99 |
| // CHECK1-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK1: cond.true: |
| // CHECK1-NEXT: br label [[COND_END:%.*]] |
| // CHECK1: cond.false: |
| // CHECK1-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: br label [[COND_END]] |
| // CHECK1: cond.end: |
| // CHECK1-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ] |
| // CHECK1-NEXT: store i32 [[COND]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK1: omp.inner.for.cond: |
| // CHECK1-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]] |
| // CHECK1-NEXT: br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK1: omp.inner.for.body: |
| // CHECK1-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1 |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 0, [[MUL]] |
| // CHECK1-NEXT: store i32 [[ADD]], i32* [[I]], align 4 |
| // CHECK1-NEXT: invoke void @_Z3foov() |
| // CHECK1-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK1: invoke.cont: |
| // CHECK1-NEXT: br label [[OMP_BODY_CONTINUE:%.*]] |
| // CHECK1: omp.body.continue: |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK1: omp.inner.for.inc: |
| // CHECK1-NEXT: [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1 |
| // CHECK1-NEXT: store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK1: omp.inner.for.end: |
| // CHECK1-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK1: omp.loop.exit: |
| // CHECK1-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]]) |
| // CHECK1-NEXT: ret void |
| // CHECK1: terminate.lpad: |
| // CHECK1-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK1-NEXT: catch i8* null |
| // CHECK1-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK1-NEXT: call void @__clang_call_terminate(i8* [[TMP12]]) #[[ATTR9]] |
| // CHECK1-NEXT: unreachable |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainI1SLi1EEiv_l36 |
| // CHECK1-SAME: () #[[ATTR3]] { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 0, void (i32*, i32*, ...)* bitcast (void (i32*, i32*)* @.omp_outlined..8 to void (i32*, i32*, ...)*)) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_outlined..8 |
| // CHECK1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR3]] { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: store i32 99, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4 |
| // CHECK1-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99 |
| // CHECK1-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK1: cond.true: |
| // CHECK1-NEXT: br label [[COND_END:%.*]] |
| // CHECK1: cond.false: |
| // CHECK1-NEXT: [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: br label [[COND_END]] |
| // CHECK1: cond.end: |
| // CHECK1-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ] |
| // CHECK1-NEXT: store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK1: omp.inner.for.cond: |
| // CHECK1-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]] |
| // CHECK1-NEXT: br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK1: omp.inner.for.body: |
| // CHECK1-NEXT: call void @__kmpc_push_num_threads(%struct.ident_t* @[[GLOB3]], i32 [[TMP1]], i32 1) |
| // CHECK1-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: [[TMP8:%.*]] = zext i32 [[TMP7]] to i64 |
| // CHECK1-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[TMP10:%.*]] = zext i32 [[TMP9]] to i64 |
| // CHECK1-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..9 to void (i32*, i32*, ...)*), i64 [[TMP8]], i64 [[TMP10]]) |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK1: omp.inner.for.inc: |
| // CHECK1-NEXT: [[TMP11:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP12:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP11]], [[TMP12]] |
| // CHECK1-NEXT: store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK1: omp.inner.for.end: |
| // CHECK1-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK1: omp.loop.exit: |
| // CHECK1-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]]) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_outlined..9 |
| // CHECK1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_LB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_UB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 99, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i32 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32 |
| // CHECK1-NEXT: store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4 |
| // CHECK1-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK1-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99 |
| // CHECK1-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK1: cond.true: |
| // CHECK1-NEXT: br label [[COND_END:%.*]] |
| // CHECK1: cond.false: |
| // CHECK1-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: br label [[COND_END]] |
| // CHECK1: cond.end: |
| // CHECK1-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ] |
| // CHECK1-NEXT: store i32 [[COND]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK1: omp.inner.for.cond: |
| // CHECK1-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]] |
| // CHECK1-NEXT: br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK1: omp.inner.for.body: |
| // CHECK1-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1 |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 0, [[MUL]] |
| // CHECK1-NEXT: store i32 [[ADD]], i32* [[I]], align 4 |
| // CHECK1-NEXT: invoke void @_Z3foov() |
| // CHECK1-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK1: invoke.cont: |
| // CHECK1-NEXT: br label [[OMP_BODY_CONTINUE:%.*]] |
| // CHECK1: omp.body.continue: |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK1: omp.inner.for.inc: |
| // CHECK1-NEXT: [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1 |
| // CHECK1-NEXT: store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK1: omp.inner.for.end: |
| // CHECK1-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK1: omp.loop.exit: |
| // CHECK1-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]]) |
| // CHECK1-NEXT: ret void |
| // CHECK1: terminate.lpad: |
| // CHECK1-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK1-NEXT: catch i8* null |
| // CHECK1-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK1-NEXT: call void @__clang_call_terminate(i8* [[TMP12]]) #[[ATTR9]] |
| // CHECK1-NEXT: unreachable |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainI1SLi1EEiv_l40 |
| // CHECK1-SAME: () #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i8, align 1 |
| // CHECK1-NEXT: [[REF_TMP:%.*]] = alloca [[STRUCT_S:%.*]], align 8 |
| // CHECK1-NEXT: [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: invoke void @_ZN1SC1El(%struct.S* noundef nonnull align 8 dereferenceable(24) [[REF_TMP]], i64 noundef 23) |
| // CHECK1-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK1: invoke.cont: |
| // CHECK1-NEXT: [[CALL:%.*]] = call noundef signext i8 @_ZN1ScvcEv(%struct.S* noundef nonnull align 8 dereferenceable(24) [[REF_TMP]]) |
| // CHECK1-NEXT: call void @_ZN1SD1Ev(%struct.S* noundef nonnull align 8 dereferenceable(24) [[REF_TMP]]) #[[ATTR6]] |
| // CHECK1-NEXT: store i8 [[CALL]], i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i8, i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK1-NEXT: [[CONV:%.*]] = bitcast i64* [[DOTCAPTURE_EXPR__CASTED]] to i8* |
| // CHECK1-NEXT: store i8 [[TMP0]], i8* [[CONV]], align 1 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTCAPTURE_EXPR__CASTED]], align 8 |
| // CHECK1-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 1, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64)* @.omp_outlined..10 to void (i32*, i32*, ...)*), i64 [[TMP1]]) |
| // CHECK1-NEXT: ret void |
| // CHECK1: terminate.lpad: |
| // CHECK1-NEXT: [[TMP2:%.*]] = landingpad { i8*, i32 } |
| // CHECK1-NEXT: catch i8* null |
| // CHECK1-NEXT: [[TMP3:%.*]] = extractvalue { i8*, i32 } [[TMP2]], 0 |
| // CHECK1-NEXT: call void @__clang_call_terminate(i8* [[TMP3]]) #[[ATTR9]] |
| // CHECK1-NEXT: unreachable |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_outlined..10 |
| // CHECK1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR3]] { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTCAPTURE_EXPR_]], i64* [[DOTCAPTURE_EXPR__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV:%.*]] = bitcast i64* [[DOTCAPTURE_EXPR__ADDR]] to i8* |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: store i32 99, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4 |
| // CHECK1-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99 |
| // CHECK1-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK1: cond.true: |
| // CHECK1-NEXT: br label [[COND_END:%.*]] |
| // CHECK1: cond.false: |
| // CHECK1-NEXT: [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: br label [[COND_END]] |
| // CHECK1: cond.end: |
| // CHECK1-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ] |
| // CHECK1-NEXT: store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK1: omp.inner.for.cond: |
| // CHECK1-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]] |
| // CHECK1-NEXT: br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK1: omp.inner.for.body: |
| // CHECK1-NEXT: [[TMP7:%.*]] = load i8, i8* [[CONV]], align 1 |
| // CHECK1-NEXT: [[TMP8:%.*]] = sext i8 [[TMP7]] to i32 |
| // CHECK1-NEXT: call void @__kmpc_push_num_threads(%struct.ident_t* @[[GLOB3]], i32 [[TMP1]], i32 [[TMP8]]) |
| // CHECK1-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK1-NEXT: [[TMP10:%.*]] = zext i32 [[TMP9]] to i64 |
| // CHECK1-NEXT: [[TMP11:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK1-NEXT: [[TMP12:%.*]] = zext i32 [[TMP11]] to i64 |
| // CHECK1-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..11 to void (i32*, i32*, ...)*), i64 [[TMP10]], i64 [[TMP12]]) |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK1: omp.inner.for.inc: |
| // CHECK1-NEXT: [[TMP13:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP14:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP13]], [[TMP14]] |
| // CHECK1-NEXT: store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK1: omp.inner.for.end: |
| // CHECK1-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK1: omp.loop.exit: |
| // CHECK1-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]]) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_outlined..11 |
| // CHECK1-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK1-NEXT: [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK1-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_LB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_UB:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK1-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK1-NEXT: store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 99, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i32 |
| // CHECK1-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK1-NEXT: [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32 |
| // CHECK1-NEXT: store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK1-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK1-NEXT: [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK1-NEXT: [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4 |
| // CHECK1-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK1-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99 |
| // CHECK1-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK1: cond.true: |
| // CHECK1-NEXT: br label [[COND_END:%.*]] |
| // CHECK1: cond.false: |
| // CHECK1-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: br label [[COND_END]] |
| // CHECK1: cond.end: |
| // CHECK1-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ] |
| // CHECK1-NEXT: store i32 [[COND]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4 |
| // CHECK1-NEXT: store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK1: omp.inner.for.cond: |
| // CHECK1-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK1-NEXT: [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]] |
| // CHECK1-NEXT: br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK1: omp.inner.for.body: |
| // CHECK1-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1 |
| // CHECK1-NEXT: [[ADD:%.*]] = add nsw i32 0, [[MUL]] |
| // CHECK1-NEXT: store i32 [[ADD]], i32* [[I]], align 4 |
| // CHECK1-NEXT: invoke void @_Z3foov() |
| // CHECK1-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK1: invoke.cont: |
| // CHECK1-NEXT: br label [[OMP_BODY_CONTINUE:%.*]] |
| // CHECK1: omp.body.continue: |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK1: omp.inner.for.inc: |
| // CHECK1-NEXT: [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1 |
| // CHECK1-NEXT: store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK1-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK1: omp.inner.for.end: |
| // CHECK1-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK1: omp.loop.exit: |
| // CHECK1-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]]) |
| // CHECK1-NEXT: ret void |
| // CHECK1: terminate.lpad: |
| // CHECK1-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK1-NEXT: catch i8* null |
| // CHECK1-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK1-NEXT: call void @__clang_call_terminate(i8* [[TMP12]]) #[[ATTR9]] |
| // CHECK1-NEXT: unreachable |
| // |
| // |
| // CHECK1-LABEL: define {{[^@]+}}@.omp_offloading.requires_reg |
| // CHECK1-SAME: () #[[ATTR8:[0-9]+]] section ".text.startup" { |
| // CHECK1-NEXT: entry: |
| // CHECK1-NEXT: call void @__tgt_register_requires(i64 1) |
| // CHECK1-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@main |
| // CHECK5-SAME: () #[[ATTR0:[0-9]+]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[RETVAL:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[S:%.*]] = alloca [[STRUCT_S:%.*]], align 8 |
| // CHECK5-NEXT: [[A:%.*]] = alloca i8, align 1 |
| // CHECK5-NEXT: [[EXN_SLOT:%.*]] = alloca i8*, align 8 |
| // CHECK5-NEXT: [[EHSELECTOR_SLOT:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[A_CASTED:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTOFFLOAD_BASEPTRS:%.*]] = alloca [1 x i8*], align 8 |
| // CHECK5-NEXT: [[DOTOFFLOAD_PTRS:%.*]] = alloca [1 x i8*], align 8 |
| // CHECK5-NEXT: [[DOTOFFLOAD_MAPPERS:%.*]] = alloca [1 x i8*], align 8 |
| // CHECK5-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i8, align 1 |
| // CHECK5-NEXT: [[_TMP1:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[RETVAL]], align 4 |
| // CHECK5-NEXT: call void @_ZN1SC1El(%struct.S* noundef nonnull align 8 dereferenceable(24) [[S]], i64 noundef 0) |
| // CHECK5-NEXT: [[CALL:%.*]] = invoke noundef signext i8 @_ZN1ScvcEv(%struct.S* noundef nonnull align 8 dereferenceable(24) [[S]]) |
| // CHECK5-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[LPAD:%.*]] |
| // CHECK5: invoke.cont: |
| // CHECK5-NEXT: store i8 [[CALL]], i8* [[A]], align 1 |
| // CHECK5-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8 |
| // CHECK5-NEXT: [[TMP0:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 0 |
| // CHECK5-NEXT: store i32 1, i32* [[TMP0]], align 4 |
| // CHECK5-NEXT: [[TMP1:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 1 |
| // CHECK5-NEXT: store i32 0, i32* [[TMP1]], align 4 |
| // CHECK5-NEXT: [[TMP2:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 2 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP2]], align 8 |
| // CHECK5-NEXT: [[TMP3:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 3 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP3]], align 8 |
| // CHECK5-NEXT: [[TMP4:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 4 |
| // CHECK5-NEXT: store i64* null, i64** [[TMP4]], align 8 |
| // CHECK5-NEXT: [[TMP5:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 5 |
| // CHECK5-NEXT: store i64* null, i64** [[TMP5]], align 8 |
| // CHECK5-NEXT: [[TMP6:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 6 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP6]], align 8 |
| // CHECK5-NEXT: [[TMP7:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 7 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP7]], align 8 |
| // CHECK5-NEXT: [[TMP8:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 8 |
| // CHECK5-NEXT: store i64 100, i64* [[TMP8]], align 8 |
| // CHECK5-NEXT: [[TMP9:%.*]] = call i32 @__tgt_target_kernel(%struct.ident_t* @[[GLOB3:[0-9]+]], i64 -1, i32 0, i32 2, i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l50.region_id, %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]]) |
| // CHECK5-NEXT: [[TMP10:%.*]] = icmp ne i32 [[TMP9]], 0 |
| // CHECK5-NEXT: br i1 [[TMP10]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]] |
| // CHECK5: omp_offload.failed: |
| // CHECK5-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l50() #[[ATTR6:[0-9]+]] |
| // CHECK5-NEXT: br label [[OMP_OFFLOAD_CONT]] |
| // CHECK5: lpad: |
| // CHECK5-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK5-NEXT: cleanup |
| // CHECK5-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK5-NEXT: store i8* [[TMP12]], i8** [[EXN_SLOT]], align 8 |
| // CHECK5-NEXT: [[TMP13:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 1 |
| // CHECK5-NEXT: store i32 [[TMP13]], i32* [[EHSELECTOR_SLOT]], align 4 |
| // CHECK5-NEXT: call void @_ZN1SD1Ev(%struct.S* noundef nonnull align 8 dereferenceable(24) [[S]]) #[[ATTR6]] |
| // CHECK5-NEXT: br label [[EH_RESUME:%.*]] |
| // CHECK5: omp_offload.cont: |
| // CHECK5-NEXT: [[TMP14:%.*]] = load i8, i8* [[A]], align 1 |
| // CHECK5-NEXT: [[CONV:%.*]] = bitcast i64* [[A_CASTED]] to i8* |
| // CHECK5-NEXT: store i8 [[TMP14]], i8* [[CONV]], align 1 |
| // CHECK5-NEXT: [[TMP15:%.*]] = load i64, i64* [[A_CASTED]], align 8 |
| // CHECK5-NEXT: [[TMP16:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0 |
| // CHECK5-NEXT: [[TMP17:%.*]] = bitcast i8** [[TMP16]] to i64* |
| // CHECK5-NEXT: store i64 [[TMP15]], i64* [[TMP17]], align 8 |
| // CHECK5-NEXT: [[TMP18:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[DOTOFFLOAD_PTRS]], i32 0, i32 0 |
| // CHECK5-NEXT: [[TMP19:%.*]] = bitcast i8** [[TMP18]] to i64* |
| // CHECK5-NEXT: store i64 [[TMP15]], i64* [[TMP19]], align 8 |
| // CHECK5-NEXT: [[TMP20:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[DOTOFFLOAD_MAPPERS]], i64 0, i64 0 |
| // CHECK5-NEXT: store i8* null, i8** [[TMP20]], align 8 |
| // CHECK5-NEXT: [[TMP21:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[DOTOFFLOAD_BASEPTRS]], i32 0, i32 0 |
| // CHECK5-NEXT: [[TMP22:%.*]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[DOTOFFLOAD_PTRS]], i32 0, i32 0 |
| // CHECK5-NEXT: [[TMP23:%.*]] = load i8, i8* [[A]], align 1 |
| // CHECK5-NEXT: store i8 [[TMP23]], i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK5-NEXT: [[TMP24:%.*]] = load i8, i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK5-NEXT: [[TMP25:%.*]] = zext i8 [[TMP24]] to i32 |
| // CHECK5-NEXT: [[KERNEL_ARGS2:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8 |
| // CHECK5-NEXT: [[TMP26:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 0 |
| // CHECK5-NEXT: store i32 1, i32* [[TMP26]], align 4 |
| // CHECK5-NEXT: [[TMP27:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 1 |
| // CHECK5-NEXT: store i32 1, i32* [[TMP27]], align 4 |
| // CHECK5-NEXT: [[TMP28:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 2 |
| // CHECK5-NEXT: store i8** [[TMP21]], i8*** [[TMP28]], align 8 |
| // CHECK5-NEXT: [[TMP29:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 3 |
| // CHECK5-NEXT: store i8** [[TMP22]], i8*** [[TMP29]], align 8 |
| // CHECK5-NEXT: [[TMP30:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 4 |
| // CHECK5-NEXT: store i64* getelementptr inbounds ([1 x i64], [1 x i64]* @.offload_sizes, i32 0, i32 0), i64** [[TMP30]], align 8 |
| // CHECK5-NEXT: [[TMP31:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 5 |
| // CHECK5-NEXT: store i64* getelementptr inbounds ([1 x i64], [1 x i64]* @.offload_maptypes, i32 0, i32 0), i64** [[TMP31]], align 8 |
| // CHECK5-NEXT: [[TMP32:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 6 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP32]], align 8 |
| // CHECK5-NEXT: [[TMP33:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 7 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP33]], align 8 |
| // CHECK5-NEXT: [[TMP34:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 8 |
| // CHECK5-NEXT: store i64 100, i64* [[TMP34]], align 8 |
| // CHECK5-NEXT: [[TMP35:%.*]] = call i32 @__tgt_target_kernel(%struct.ident_t* @[[GLOB3]], i64 -1, i32 0, i32 [[TMP25]], i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l55.region_id, %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]]) |
| // CHECK5-NEXT: [[TMP36:%.*]] = icmp ne i32 [[TMP35]], 0 |
| // CHECK5-NEXT: br i1 [[TMP36]], label [[OMP_OFFLOAD_FAILED3:%.*]], label [[OMP_OFFLOAD_CONT4:%.*]] |
| // CHECK5: omp_offload.failed3: |
| // CHECK5-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l55(i64 [[TMP15]]) #[[ATTR6]] |
| // CHECK5-NEXT: br label [[OMP_OFFLOAD_CONT4]] |
| // CHECK5: omp_offload.cont4: |
| // CHECK5-NEXT: [[TMP37:%.*]] = load i8, i8* [[A]], align 1 |
| // CHECK5-NEXT: [[CONV5:%.*]] = sext i8 [[TMP37]] to i32 |
| // CHECK5-NEXT: [[CALL7:%.*]] = invoke noundef signext i32 @_Z5tmainIcLi5EEiv() |
| // CHECK5-NEXT: to label [[INVOKE_CONT6:%.*]] unwind label [[LPAD]] |
| // CHECK5: invoke.cont6: |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 [[CONV5]], [[CALL7]] |
| // CHECK5-NEXT: [[CALL9:%.*]] = invoke noundef signext i32 @_Z5tmainI1SLi1EEiv() |
| // CHECK5-NEXT: to label [[INVOKE_CONT8:%.*]] unwind label [[LPAD]] |
| // CHECK5: invoke.cont8: |
| // CHECK5-NEXT: [[ADD10:%.*]] = add nsw i32 [[ADD]], [[CALL9]] |
| // CHECK5-NEXT: store i32 [[ADD10]], i32* [[RETVAL]], align 4 |
| // CHECK5-NEXT: call void @_ZN1SD1Ev(%struct.S* noundef nonnull align 8 dereferenceable(24) [[S]]) #[[ATTR6]] |
| // CHECK5-NEXT: [[TMP38:%.*]] = load i32, i32* [[RETVAL]], align 4 |
| // CHECK5-NEXT: ret i32 [[TMP38]] |
| // CHECK5: eh.resume: |
| // CHECK5-NEXT: [[EXN:%.*]] = load i8*, i8** [[EXN_SLOT]], align 8 |
| // CHECK5-NEXT: [[SEL:%.*]] = load i32, i32* [[EHSELECTOR_SLOT]], align 4 |
| // CHECK5-NEXT: [[LPAD_VAL:%.*]] = insertvalue { i8*, i32 } undef, i8* [[EXN]], 0 |
| // CHECK5-NEXT: [[LPAD_VAL11:%.*]] = insertvalue { i8*, i32 } [[LPAD_VAL]], i32 [[SEL]], 1 |
| // CHECK5-NEXT: resume { i8*, i32 } [[LPAD_VAL11]] |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@_ZN1SC1El |
| // CHECK5-SAME: (%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS:%.*]], i64 noundef [[A:%.*]]) unnamed_addr #[[ATTR1:[0-9]+]] comdat align 2 { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[THIS_ADDR:%.*]] = alloca %struct.S*, align 8 |
| // CHECK5-NEXT: [[A_ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: store %struct.S* [[THIS]], %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[A]], i64* [[A_ADDR]], align 8 |
| // CHECK5-NEXT: [[THIS1:%.*]] = load %struct.S*, %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i64, i64* [[A_ADDR]], align 8 |
| // CHECK5-NEXT: call void @_ZN1SC2El(%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS1]], i64 noundef [[TMP0]]) |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@_ZN1ScvcEv |
| // CHECK5-SAME: (%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS:%.*]]) #[[ATTR2:[0-9]+]] comdat align 2 { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[THIS_ADDR:%.*]] = alloca %struct.S*, align 8 |
| // CHECK5-NEXT: store %struct.S* [[THIS]], %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK5-NEXT: [[THIS1:%.*]] = load %struct.S*, %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK5-NEXT: [[A:%.*]] = getelementptr inbounds [[STRUCT_S:%.*]], %struct.S* [[THIS1]], i32 0, i32 0 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i64, i64* [[A]], align 8 |
| // CHECK5-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i8 |
| // CHECK5-NEXT: ret i8 [[CONV]] |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l50 |
| // CHECK5-SAME: () #[[ATTR3:[0-9]+]] { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 0, void (i32*, i32*, ...)* bitcast (void (i32*, i32*)* @.omp_outlined. to void (i32*, i32*, ...)*)) |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_outlined. |
| // CHECK5-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR3]] { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: store i32 99, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4 |
| // CHECK5-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1:[0-9]+]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99 |
| // CHECK5-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK5: cond.true: |
| // CHECK5-NEXT: br label [[COND_END:%.*]] |
| // CHECK5: cond.false: |
| // CHECK5-NEXT: [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: br label [[COND_END]] |
| // CHECK5: cond.end: |
| // CHECK5-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ] |
| // CHECK5-NEXT: store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK5: omp.inner.for.cond: |
| // CHECK5-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]] |
| // CHECK5-NEXT: br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK5: omp.inner.for.body: |
| // CHECK5-NEXT: call void @__kmpc_push_num_threads(%struct.ident_t* @[[GLOB3]], i32 [[TMP1]], i32 2) |
| // CHECK5-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: [[TMP8:%.*]] = zext i32 [[TMP7]] to i64 |
| // CHECK5-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[TMP10:%.*]] = zext i32 [[TMP9]] to i64 |
| // CHECK5-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..1 to void (i32*, i32*, ...)*), i64 [[TMP8]], i64 [[TMP10]]) |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK5: omp.inner.for.inc: |
| // CHECK5-NEXT: [[TMP11:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP12:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP11]], [[TMP12]] |
| // CHECK5-NEXT: store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK5: omp.inner.for.end: |
| // CHECK5-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK5: omp.loop.exit: |
| // CHECK5-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]]) |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_outlined..1 |
| // CHECK5-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_LB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_UB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 99, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i32 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32 |
| // CHECK5-NEXT: store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4 |
| // CHECK5-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2:[0-9]+]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK5-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99 |
| // CHECK5-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK5: cond.true: |
| // CHECK5-NEXT: br label [[COND_END:%.*]] |
| // CHECK5: cond.false: |
| // CHECK5-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: br label [[COND_END]] |
| // CHECK5: cond.end: |
| // CHECK5-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ] |
| // CHECK5-NEXT: store i32 [[COND]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK5: omp.inner.for.cond: |
| // CHECK5-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]] |
| // CHECK5-NEXT: br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK5: omp.inner.for.body: |
| // CHECK5-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1 |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 0, [[MUL]] |
| // CHECK5-NEXT: store i32 [[ADD]], i32* [[I]], align 4 |
| // CHECK5-NEXT: invoke void @_Z3foov() |
| // CHECK5-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK5: invoke.cont: |
| // CHECK5-NEXT: br label [[OMP_BODY_CONTINUE:%.*]] |
| // CHECK5: omp.body.continue: |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK5: omp.inner.for.inc: |
| // CHECK5-NEXT: [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1 |
| // CHECK5-NEXT: store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK5: omp.inner.for.end: |
| // CHECK5-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK5: omp.loop.exit: |
| // CHECK5-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]]) |
| // CHECK5-NEXT: ret void |
| // CHECK5: terminate.lpad: |
| // CHECK5-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK5-NEXT: catch i8* null |
| // CHECK5-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK5-NEXT: call void @__clang_call_terminate(i8* [[TMP12]]) #[[ATTR9:[0-9]+]] |
| // CHECK5-NEXT: unreachable |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@__clang_call_terminate |
| // CHECK5-SAME: (i8* [[TMP0:%.*]]) #[[ATTR5:[0-9]+]] comdat { |
| // CHECK5-NEXT: [[TMP2:%.*]] = call i8* @__cxa_begin_catch(i8* [[TMP0]]) #[[ATTR6]] |
| // CHECK5-NEXT: call void @_ZSt9terminatev() #[[ATTR9]] |
| // CHECK5-NEXT: unreachable |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}_main_l55 |
| // CHECK5-SAME: (i64 noundef [[A:%.*]]) #[[ATTR3]] { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[A_ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i8, align 1 |
| // CHECK5-NEXT: [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: store i64 [[A]], i64* [[A_ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV:%.*]] = bitcast i64* [[A_ADDR]] to i8* |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i8, i8* [[CONV]], align 1 |
| // CHECK5-NEXT: store i8 [[TMP0]], i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i8, i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK5-NEXT: [[CONV1:%.*]] = bitcast i64* [[DOTCAPTURE_EXPR__CASTED]] to i8* |
| // CHECK5-NEXT: store i8 [[TMP1]], i8* [[CONV1]], align 1 |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i64, i64* [[DOTCAPTURE_EXPR__CASTED]], align 8 |
| // CHECK5-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 1, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64)* @.omp_outlined..2 to void (i32*, i32*, ...)*), i64 [[TMP2]]) |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_outlined..2 |
| // CHECK5-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR3]] { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTCAPTURE_EXPR_]], i64* [[DOTCAPTURE_EXPR__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV:%.*]] = bitcast i64* [[DOTCAPTURE_EXPR__ADDR]] to i8* |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: store i32 99, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4 |
| // CHECK5-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99 |
| // CHECK5-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK5: cond.true: |
| // CHECK5-NEXT: br label [[COND_END:%.*]] |
| // CHECK5: cond.false: |
| // CHECK5-NEXT: [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: br label [[COND_END]] |
| // CHECK5: cond.end: |
| // CHECK5-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ] |
| // CHECK5-NEXT: store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK5: omp.inner.for.cond: |
| // CHECK5-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]] |
| // CHECK5-NEXT: br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK5: omp.inner.for.body: |
| // CHECK5-NEXT: [[TMP7:%.*]] = load i8, i8* [[CONV]], align 1 |
| // CHECK5-NEXT: [[TMP8:%.*]] = sext i8 [[TMP7]] to i32 |
| // CHECK5-NEXT: call void @__kmpc_push_num_threads(%struct.ident_t* @[[GLOB3]], i32 [[TMP1]], i32 [[TMP8]]) |
| // CHECK5-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: [[TMP10:%.*]] = zext i32 [[TMP9]] to i64 |
| // CHECK5-NEXT: [[TMP11:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[TMP12:%.*]] = zext i32 [[TMP11]] to i64 |
| // CHECK5-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..3 to void (i32*, i32*, ...)*), i64 [[TMP10]], i64 [[TMP12]]) |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK5: omp.inner.for.inc: |
| // CHECK5-NEXT: [[TMP13:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP14:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP13]], [[TMP14]] |
| // CHECK5-NEXT: store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK5: omp.inner.for.end: |
| // CHECK5-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK5: omp.loop.exit: |
| // CHECK5-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]]) |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_outlined..3 |
| // CHECK5-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_LB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_UB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 99, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i32 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32 |
| // CHECK5-NEXT: store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4 |
| // CHECK5-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK5-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99 |
| // CHECK5-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK5: cond.true: |
| // CHECK5-NEXT: br label [[COND_END:%.*]] |
| // CHECK5: cond.false: |
| // CHECK5-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: br label [[COND_END]] |
| // CHECK5: cond.end: |
| // CHECK5-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ] |
| // CHECK5-NEXT: store i32 [[COND]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK5: omp.inner.for.cond: |
| // CHECK5-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]] |
| // CHECK5-NEXT: br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK5: omp.inner.for.body: |
| // CHECK5-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1 |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 0, [[MUL]] |
| // CHECK5-NEXT: store i32 [[ADD]], i32* [[I]], align 4 |
| // CHECK5-NEXT: invoke void @_Z3foov() |
| // CHECK5-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK5: invoke.cont: |
| // CHECK5-NEXT: br label [[OMP_BODY_CONTINUE:%.*]] |
| // CHECK5: omp.body.continue: |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK5: omp.inner.for.inc: |
| // CHECK5-NEXT: [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1 |
| // CHECK5-NEXT: store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK5: omp.inner.for.end: |
| // CHECK5-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK5: omp.loop.exit: |
| // CHECK5-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]]) |
| // CHECK5-NEXT: ret void |
| // CHECK5: terminate.lpad: |
| // CHECK5-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK5-NEXT: catch i8* null |
| // CHECK5-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK5-NEXT: call void @__clang_call_terminate(i8* [[TMP12]]) #[[ATTR9]] |
| // CHECK5-NEXT: unreachable |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@_Z5tmainIcLi5EEiv |
| // CHECK5-SAME: () #[[ATTR2]] comdat { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[_TMP1:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8 |
| // CHECK5-NEXT: [[TMP0:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 0 |
| // CHECK5-NEXT: store i32 1, i32* [[TMP0]], align 4 |
| // CHECK5-NEXT: [[TMP1:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 1 |
| // CHECK5-NEXT: store i32 0, i32* [[TMP1]], align 4 |
| // CHECK5-NEXT: [[TMP2:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 2 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP2]], align 8 |
| // CHECK5-NEXT: [[TMP3:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 3 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP3]], align 8 |
| // CHECK5-NEXT: [[TMP4:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 4 |
| // CHECK5-NEXT: store i64* null, i64** [[TMP4]], align 8 |
| // CHECK5-NEXT: [[TMP5:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 5 |
| // CHECK5-NEXT: store i64* null, i64** [[TMP5]], align 8 |
| // CHECK5-NEXT: [[TMP6:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 6 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP6]], align 8 |
| // CHECK5-NEXT: [[TMP7:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 7 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP7]], align 8 |
| // CHECK5-NEXT: [[TMP8:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 8 |
| // CHECK5-NEXT: store i64 100, i64* [[TMP8]], align 8 |
| // CHECK5-NEXT: [[TMP9:%.*]] = call i32 @__tgt_target_kernel(%struct.ident_t* @[[GLOB3]], i64 -1, i32 0, i32 5, i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIcLi5EEiv_l36.region_id, %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]]) |
| // CHECK5-NEXT: [[TMP10:%.*]] = icmp ne i32 [[TMP9]], 0 |
| // CHECK5-NEXT: br i1 [[TMP10]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]] |
| // CHECK5: omp_offload.failed: |
| // CHECK5-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIcLi5EEiv_l36() #[[ATTR6]] |
| // CHECK5-NEXT: br label [[OMP_OFFLOAD_CONT]] |
| // CHECK5: omp_offload.cont: |
| // CHECK5-NEXT: [[KERNEL_ARGS2:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8 |
| // CHECK5-NEXT: [[TMP11:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 0 |
| // CHECK5-NEXT: store i32 1, i32* [[TMP11]], align 4 |
| // CHECK5-NEXT: [[TMP12:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 1 |
| // CHECK5-NEXT: store i32 0, i32* [[TMP12]], align 4 |
| // CHECK5-NEXT: [[TMP13:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 2 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP13]], align 8 |
| // CHECK5-NEXT: [[TMP14:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 3 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP14]], align 8 |
| // CHECK5-NEXT: [[TMP15:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 4 |
| // CHECK5-NEXT: store i64* null, i64** [[TMP15]], align 8 |
| // CHECK5-NEXT: [[TMP16:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 5 |
| // CHECK5-NEXT: store i64* null, i64** [[TMP16]], align 8 |
| // CHECK5-NEXT: [[TMP17:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 6 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP17]], align 8 |
| // CHECK5-NEXT: [[TMP18:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 7 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP18]], align 8 |
| // CHECK5-NEXT: [[TMP19:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 8 |
| // CHECK5-NEXT: store i64 100, i64* [[TMP19]], align 8 |
| // CHECK5-NEXT: [[TMP20:%.*]] = call i32 @__tgt_target_kernel(%struct.ident_t* @[[GLOB3]], i64 -1, i32 0, i32 23, i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIcLi5EEiv_l40.region_id, %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]]) |
| // CHECK5-NEXT: [[TMP21:%.*]] = icmp ne i32 [[TMP20]], 0 |
| // CHECK5-NEXT: br i1 [[TMP21]], label [[OMP_OFFLOAD_FAILED3:%.*]], label [[OMP_OFFLOAD_CONT4:%.*]] |
| // CHECK5: omp_offload.failed3: |
| // CHECK5-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIcLi5EEiv_l40() #[[ATTR6]] |
| // CHECK5-NEXT: br label [[OMP_OFFLOAD_CONT4]] |
| // CHECK5: omp_offload.cont4: |
| // CHECK5-NEXT: ret i32 0 |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@_Z5tmainI1SLi1EEiv |
| // CHECK5-SAME: () #[[ATTR2]] comdat personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i8, align 1 |
| // CHECK5-NEXT: [[REF_TMP:%.*]] = alloca [[STRUCT_S:%.*]], align 8 |
| // CHECK5-NEXT: [[_TMP1:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[KERNEL_ARGS:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS:%.*]], align 8 |
| // CHECK5-NEXT: [[TMP0:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 0 |
| // CHECK5-NEXT: store i32 1, i32* [[TMP0]], align 4 |
| // CHECK5-NEXT: [[TMP1:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 1 |
| // CHECK5-NEXT: store i32 0, i32* [[TMP1]], align 4 |
| // CHECK5-NEXT: [[TMP2:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 2 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP2]], align 8 |
| // CHECK5-NEXT: [[TMP3:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 3 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP3]], align 8 |
| // CHECK5-NEXT: [[TMP4:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 4 |
| // CHECK5-NEXT: store i64* null, i64** [[TMP4]], align 8 |
| // CHECK5-NEXT: [[TMP5:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 5 |
| // CHECK5-NEXT: store i64* null, i64** [[TMP5]], align 8 |
| // CHECK5-NEXT: [[TMP6:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 6 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP6]], align 8 |
| // CHECK5-NEXT: [[TMP7:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 7 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP7]], align 8 |
| // CHECK5-NEXT: [[TMP8:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]], i32 0, i32 8 |
| // CHECK5-NEXT: store i64 100, i64* [[TMP8]], align 8 |
| // CHECK5-NEXT: [[TMP9:%.*]] = call i32 @__tgt_target_kernel(%struct.ident_t* @[[GLOB3]], i64 -1, i32 0, i32 1, i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainI1SLi1EEiv_l36.region_id, %struct.__tgt_kernel_arguments* [[KERNEL_ARGS]]) |
| // CHECK5-NEXT: [[TMP10:%.*]] = icmp ne i32 [[TMP9]], 0 |
| // CHECK5-NEXT: br i1 [[TMP10]], label [[OMP_OFFLOAD_FAILED:%.*]], label [[OMP_OFFLOAD_CONT:%.*]] |
| // CHECK5: omp_offload.failed: |
| // CHECK5-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainI1SLi1EEiv_l36() #[[ATTR6]] |
| // CHECK5-NEXT: br label [[OMP_OFFLOAD_CONT]] |
| // CHECK5: omp_offload.cont: |
| // CHECK5-NEXT: invoke void @_ZN1SC1El(%struct.S* noundef nonnull align 8 dereferenceable(24) [[REF_TMP]], i64 noundef 23) |
| // CHECK5-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK5: invoke.cont: |
| // CHECK5-NEXT: [[CALL:%.*]] = call noundef signext i8 @_ZN1ScvcEv(%struct.S* noundef nonnull align 8 dereferenceable(24) [[REF_TMP]]) |
| // CHECK5-NEXT: call void @_ZN1SD1Ev(%struct.S* noundef nonnull align 8 dereferenceable(24) [[REF_TMP]]) #[[ATTR6]] |
| // CHECK5-NEXT: store i8 [[CALL]], i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK5-NEXT: [[TMP11:%.*]] = load i8, i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK5-NEXT: [[TMP12:%.*]] = zext i8 [[TMP11]] to i32 |
| // CHECK5-NEXT: [[KERNEL_ARGS2:%.*]] = alloca [[STRUCT___TGT_KERNEL_ARGUMENTS]], align 8 |
| // CHECK5-NEXT: [[TMP13:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 0 |
| // CHECK5-NEXT: store i32 1, i32* [[TMP13]], align 4 |
| // CHECK5-NEXT: [[TMP14:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 1 |
| // CHECK5-NEXT: store i32 0, i32* [[TMP14]], align 4 |
| // CHECK5-NEXT: [[TMP15:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 2 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP15]], align 8 |
| // CHECK5-NEXT: [[TMP16:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 3 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP16]], align 8 |
| // CHECK5-NEXT: [[TMP17:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 4 |
| // CHECK5-NEXT: store i64* null, i64** [[TMP17]], align 8 |
| // CHECK5-NEXT: [[TMP18:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 5 |
| // CHECK5-NEXT: store i64* null, i64** [[TMP18]], align 8 |
| // CHECK5-NEXT: [[TMP19:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 6 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP19]], align 8 |
| // CHECK5-NEXT: [[TMP20:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 7 |
| // CHECK5-NEXT: store i8** null, i8*** [[TMP20]], align 8 |
| // CHECK5-NEXT: [[TMP21:%.*]] = getelementptr inbounds [[STRUCT___TGT_KERNEL_ARGUMENTS]], %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]], i32 0, i32 8 |
| // CHECK5-NEXT: store i64 100, i64* [[TMP21]], align 8 |
| // CHECK5-NEXT: [[TMP22:%.*]] = call i32 @__tgt_target_kernel(%struct.ident_t* @[[GLOB3]], i64 -1, i32 0, i32 [[TMP12]], i8* @.{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainI1SLi1EEiv_l40.region_id, %struct.__tgt_kernel_arguments* [[KERNEL_ARGS2]]) |
| // CHECK5-NEXT: [[TMP23:%.*]] = icmp ne i32 [[TMP22]], 0 |
| // CHECK5-NEXT: br i1 [[TMP23]], label [[OMP_OFFLOAD_FAILED3:%.*]], label [[OMP_OFFLOAD_CONT4:%.*]] |
| // CHECK5: omp_offload.failed3: |
| // CHECK5-NEXT: call void @{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainI1SLi1EEiv_l40() #[[ATTR6]] |
| // CHECK5-NEXT: br label [[OMP_OFFLOAD_CONT4]] |
| // CHECK5: omp_offload.cont4: |
| // CHECK5-NEXT: ret i32 0 |
| // CHECK5: terminate.lpad: |
| // CHECK5-NEXT: [[TMP24:%.*]] = landingpad { i8*, i32 } |
| // CHECK5-NEXT: catch i8* null |
| // CHECK5-NEXT: [[TMP25:%.*]] = extractvalue { i8*, i32 } [[TMP24]], 0 |
| // CHECK5-NEXT: call void @__clang_call_terminate(i8* [[TMP25]]) #[[ATTR9]] |
| // CHECK5-NEXT: unreachable |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@_ZN1SD1Ev |
| // CHECK5-SAME: (%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS:%.*]]) unnamed_addr #[[ATTR7:[0-9]+]] comdat align 2 { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[THIS_ADDR:%.*]] = alloca %struct.S*, align 8 |
| // CHECK5-NEXT: store %struct.S* [[THIS]], %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK5-NEXT: [[THIS1:%.*]] = load %struct.S*, %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK5-NEXT: call void @_ZN1SD2Ev(%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS1]]) #[[ATTR6]] |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@_ZN1SC2El |
| // CHECK5-SAME: (%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS:%.*]], i64 noundef [[A:%.*]]) unnamed_addr #[[ATTR7]] comdat align 2 { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[THIS_ADDR:%.*]] = alloca %struct.S*, align 8 |
| // CHECK5-NEXT: [[A_ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: store %struct.S* [[THIS]], %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[A]], i64* [[A_ADDR]], align 8 |
| // CHECK5-NEXT: [[THIS1:%.*]] = load %struct.S*, %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK5-NEXT: [[A2:%.*]] = getelementptr inbounds [[STRUCT_S:%.*]], %struct.S* [[THIS1]], i32 0, i32 0 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i64, i64* [[A_ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[TMP0]], i64* [[A2]], align 8 |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIcLi5EEiv_l36 |
| // CHECK5-SAME: () #[[ATTR3]] { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 0, void (i32*, i32*, ...)* bitcast (void (i32*, i32*)* @.omp_outlined..4 to void (i32*, i32*, ...)*)) |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_outlined..4 |
| // CHECK5-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR3]] { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: store i32 99, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4 |
| // CHECK5-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99 |
| // CHECK5-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK5: cond.true: |
| // CHECK5-NEXT: br label [[COND_END:%.*]] |
| // CHECK5: cond.false: |
| // CHECK5-NEXT: [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: br label [[COND_END]] |
| // CHECK5: cond.end: |
| // CHECK5-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ] |
| // CHECK5-NEXT: store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK5: omp.inner.for.cond: |
| // CHECK5-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]] |
| // CHECK5-NEXT: br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK5: omp.inner.for.body: |
| // CHECK5-NEXT: call void @__kmpc_push_num_threads(%struct.ident_t* @[[GLOB3]], i32 [[TMP1]], i32 5) |
| // CHECK5-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: [[TMP8:%.*]] = zext i32 [[TMP7]] to i64 |
| // CHECK5-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[TMP10:%.*]] = zext i32 [[TMP9]] to i64 |
| // CHECK5-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..5 to void (i32*, i32*, ...)*), i64 [[TMP8]], i64 [[TMP10]]) |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK5: omp.inner.for.inc: |
| // CHECK5-NEXT: [[TMP11:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP12:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP11]], [[TMP12]] |
| // CHECK5-NEXT: store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK5: omp.inner.for.end: |
| // CHECK5-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK5: omp.loop.exit: |
| // CHECK5-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]]) |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_outlined..5 |
| // CHECK5-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_LB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_UB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 99, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i32 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32 |
| // CHECK5-NEXT: store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4 |
| // CHECK5-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK5-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99 |
| // CHECK5-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK5: cond.true: |
| // CHECK5-NEXT: br label [[COND_END:%.*]] |
| // CHECK5: cond.false: |
| // CHECK5-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: br label [[COND_END]] |
| // CHECK5: cond.end: |
| // CHECK5-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ] |
| // CHECK5-NEXT: store i32 [[COND]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK5: omp.inner.for.cond: |
| // CHECK5-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]] |
| // CHECK5-NEXT: br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK5: omp.inner.for.body: |
| // CHECK5-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1 |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 0, [[MUL]] |
| // CHECK5-NEXT: store i32 [[ADD]], i32* [[I]], align 4 |
| // CHECK5-NEXT: invoke void @_Z3foov() |
| // CHECK5-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK5: invoke.cont: |
| // CHECK5-NEXT: br label [[OMP_BODY_CONTINUE:%.*]] |
| // CHECK5: omp.body.continue: |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK5: omp.inner.for.inc: |
| // CHECK5-NEXT: [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1 |
| // CHECK5-NEXT: store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK5: omp.inner.for.end: |
| // CHECK5-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK5: omp.loop.exit: |
| // CHECK5-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]]) |
| // CHECK5-NEXT: ret void |
| // CHECK5: terminate.lpad: |
| // CHECK5-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK5-NEXT: catch i8* null |
| // CHECK5-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK5-NEXT: call void @__clang_call_terminate(i8* [[TMP12]]) #[[ATTR9]] |
| // CHECK5-NEXT: unreachable |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainIcLi5EEiv_l40 |
| // CHECK5-SAME: () #[[ATTR3]] { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 0, void (i32*, i32*, ...)* bitcast (void (i32*, i32*)* @.omp_outlined..6 to void (i32*, i32*, ...)*)) |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_outlined..6 |
| // CHECK5-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR3]] { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: store i32 99, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4 |
| // CHECK5-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99 |
| // CHECK5-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK5: cond.true: |
| // CHECK5-NEXT: br label [[COND_END:%.*]] |
| // CHECK5: cond.false: |
| // CHECK5-NEXT: [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: br label [[COND_END]] |
| // CHECK5: cond.end: |
| // CHECK5-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ] |
| // CHECK5-NEXT: store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK5: omp.inner.for.cond: |
| // CHECK5-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]] |
| // CHECK5-NEXT: br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK5: omp.inner.for.body: |
| // CHECK5-NEXT: call void @__kmpc_push_num_threads(%struct.ident_t* @[[GLOB3]], i32 [[TMP1]], i32 23) |
| // CHECK5-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: [[TMP8:%.*]] = zext i32 [[TMP7]] to i64 |
| // CHECK5-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[TMP10:%.*]] = zext i32 [[TMP9]] to i64 |
| // CHECK5-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..7 to void (i32*, i32*, ...)*), i64 [[TMP8]], i64 [[TMP10]]) |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK5: omp.inner.for.inc: |
| // CHECK5-NEXT: [[TMP11:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP12:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP11]], [[TMP12]] |
| // CHECK5-NEXT: store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK5: omp.inner.for.end: |
| // CHECK5-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK5: omp.loop.exit: |
| // CHECK5-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]]) |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_outlined..7 |
| // CHECK5-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_LB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_UB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 99, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i32 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32 |
| // CHECK5-NEXT: store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4 |
| // CHECK5-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK5-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99 |
| // CHECK5-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK5: cond.true: |
| // CHECK5-NEXT: br label [[COND_END:%.*]] |
| // CHECK5: cond.false: |
| // CHECK5-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: br label [[COND_END]] |
| // CHECK5: cond.end: |
| // CHECK5-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ] |
| // CHECK5-NEXT: store i32 [[COND]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK5: omp.inner.for.cond: |
| // CHECK5-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]] |
| // CHECK5-NEXT: br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK5: omp.inner.for.body: |
| // CHECK5-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1 |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 0, [[MUL]] |
| // CHECK5-NEXT: store i32 [[ADD]], i32* [[I]], align 4 |
| // CHECK5-NEXT: invoke void @_Z3foov() |
| // CHECK5-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK5: invoke.cont: |
| // CHECK5-NEXT: br label [[OMP_BODY_CONTINUE:%.*]] |
| // CHECK5: omp.body.continue: |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK5: omp.inner.for.inc: |
| // CHECK5-NEXT: [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1 |
| // CHECK5-NEXT: store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK5: omp.inner.for.end: |
| // CHECK5-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK5: omp.loop.exit: |
| // CHECK5-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]]) |
| // CHECK5-NEXT: ret void |
| // CHECK5: terminate.lpad: |
| // CHECK5-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK5-NEXT: catch i8* null |
| // CHECK5-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK5-NEXT: call void @__clang_call_terminate(i8* [[TMP12]]) #[[ATTR9]] |
| // CHECK5-NEXT: unreachable |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainI1SLi1EEiv_l36 |
| // CHECK5-SAME: () #[[ATTR3]] { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 0, void (i32*, i32*, ...)* bitcast (void (i32*, i32*)* @.omp_outlined..8 to void (i32*, i32*, ...)*)) |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_outlined..8 |
| // CHECK5-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]]) #[[ATTR3]] { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: store i32 99, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4 |
| // CHECK5-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99 |
| // CHECK5-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK5: cond.true: |
| // CHECK5-NEXT: br label [[COND_END:%.*]] |
| // CHECK5: cond.false: |
| // CHECK5-NEXT: [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: br label [[COND_END]] |
| // CHECK5: cond.end: |
| // CHECK5-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ] |
| // CHECK5-NEXT: store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK5: omp.inner.for.cond: |
| // CHECK5-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]] |
| // CHECK5-NEXT: br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK5: omp.inner.for.body: |
| // CHECK5-NEXT: call void @__kmpc_push_num_threads(%struct.ident_t* @[[GLOB3]], i32 [[TMP1]], i32 1) |
| // CHECK5-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: [[TMP8:%.*]] = zext i32 [[TMP7]] to i64 |
| // CHECK5-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[TMP10:%.*]] = zext i32 [[TMP9]] to i64 |
| // CHECK5-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..9 to void (i32*, i32*, ...)*), i64 [[TMP8]], i64 [[TMP10]]) |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK5: omp.inner.for.inc: |
| // CHECK5-NEXT: [[TMP11:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP12:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP11]], [[TMP12]] |
| // CHECK5-NEXT: store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK5: omp.inner.for.end: |
| // CHECK5-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK5: omp.loop.exit: |
| // CHECK5-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]]) |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_outlined..9 |
| // CHECK5-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_LB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_UB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 99, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i32 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32 |
| // CHECK5-NEXT: store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4 |
| // CHECK5-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK5-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99 |
| // CHECK5-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK5: cond.true: |
| // CHECK5-NEXT: br label [[COND_END:%.*]] |
| // CHECK5: cond.false: |
| // CHECK5-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: br label [[COND_END]] |
| // CHECK5: cond.end: |
| // CHECK5-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ] |
| // CHECK5-NEXT: store i32 [[COND]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK5: omp.inner.for.cond: |
| // CHECK5-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]] |
| // CHECK5-NEXT: br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK5: omp.inner.for.body: |
| // CHECK5-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1 |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 0, [[MUL]] |
| // CHECK5-NEXT: store i32 [[ADD]], i32* [[I]], align 4 |
| // CHECK5-NEXT: invoke void @_Z3foov() |
| // CHECK5-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK5: invoke.cont: |
| // CHECK5-NEXT: br label [[OMP_BODY_CONTINUE:%.*]] |
| // CHECK5: omp.body.continue: |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK5: omp.inner.for.inc: |
| // CHECK5-NEXT: [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1 |
| // CHECK5-NEXT: store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK5: omp.inner.for.end: |
| // CHECK5-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK5: omp.loop.exit: |
| // CHECK5-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]]) |
| // CHECK5-NEXT: ret void |
| // CHECK5: terminate.lpad: |
| // CHECK5-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK5-NEXT: catch i8* null |
| // CHECK5-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK5-NEXT: call void @__clang_call_terminate(i8* [[TMP12]]) #[[ATTR9]] |
| // CHECK5-NEXT: unreachable |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@{{__omp_offloading_[0-9a-z]+_[0-9a-z]+}}__Z5tmainI1SLi1EEiv_l40 |
| // CHECK5-SAME: () #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTCAPTURE_EXPR_:%.*]] = alloca i8, align 1 |
| // CHECK5-NEXT: [[REF_TMP:%.*]] = alloca [[STRUCT_S:%.*]], align 8 |
| // CHECK5-NEXT: [[DOTCAPTURE_EXPR__CASTED:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: invoke void @_ZN1SC1El(%struct.S* noundef nonnull align 8 dereferenceable(24) [[REF_TMP]], i64 noundef 23) |
| // CHECK5-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK5: invoke.cont: |
| // CHECK5-NEXT: [[CALL:%.*]] = call noundef signext i8 @_ZN1ScvcEv(%struct.S* noundef nonnull align 8 dereferenceable(24) [[REF_TMP]]) |
| // CHECK5-NEXT: call void @_ZN1SD1Ev(%struct.S* noundef nonnull align 8 dereferenceable(24) [[REF_TMP]]) #[[ATTR6]] |
| // CHECK5-NEXT: store i8 [[CALL]], i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i8, i8* [[DOTCAPTURE_EXPR_]], align 1 |
| // CHECK5-NEXT: [[CONV:%.*]] = bitcast i64* [[DOTCAPTURE_EXPR__CASTED]] to i8* |
| // CHECK5-NEXT: store i8 [[TMP0]], i8* [[CONV]], align 1 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTCAPTURE_EXPR__CASTED]], align 8 |
| // CHECK5-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_teams(%struct.ident_t* @[[GLOB3]], i32 1, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64)* @.omp_outlined..10 to void (i32*, i32*, ...)*), i64 [[TMP1]]) |
| // CHECK5-NEXT: ret void |
| // CHECK5: terminate.lpad: |
| // CHECK5-NEXT: [[TMP2:%.*]] = landingpad { i8*, i32 } |
| // CHECK5-NEXT: catch i8* null |
| // CHECK5-NEXT: [[TMP3:%.*]] = extractvalue { i8*, i32 } [[TMP2]], 0 |
| // CHECK5-NEXT: call void @__clang_call_terminate(i8* [[TMP3]]) #[[ATTR9]] |
| // CHECK5-NEXT: unreachable |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_outlined..10 |
| // CHECK5-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTCAPTURE_EXPR_:%.*]]) #[[ATTR3]] { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTCAPTURE_EXPR__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_COMB_LB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_COMB_UB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTCAPTURE_EXPR_]], i64* [[DOTCAPTURE_EXPR__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV:%.*]] = bitcast i64* [[DOTCAPTURE_EXPR__ADDR]] to i8* |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: store i32 99, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i32, i32* [[TMP0]], align 4 |
| // CHECK5-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]], i32 92, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_COMB_LB]], i32* [[DOTOMP_COMB_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP2]], 99 |
| // CHECK5-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK5: cond.true: |
| // CHECK5-NEXT: br label [[COND_END:%.*]] |
| // CHECK5: cond.false: |
| // CHECK5-NEXT: [[TMP3:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: br label [[COND_END]] |
| // CHECK5: cond.end: |
| // CHECK5-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP3]], [[COND_FALSE]] ] |
| // CHECK5-NEXT: store i32 [[COND]], i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[TMP4]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK5: omp.inner.for.cond: |
| // CHECK5-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[CMP1:%.*]] = icmp sle i32 [[TMP5]], [[TMP6]] |
| // CHECK5-NEXT: br i1 [[CMP1]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK5: omp.inner.for.body: |
| // CHECK5-NEXT: [[TMP7:%.*]] = load i8, i8* [[CONV]], align 1 |
| // CHECK5-NEXT: [[TMP8:%.*]] = sext i8 [[TMP7]] to i32 |
| // CHECK5-NEXT: call void @__kmpc_push_num_threads(%struct.ident_t* @[[GLOB3]], i32 [[TMP1]], i32 [[TMP8]]) |
| // CHECK5-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_COMB_LB]], align 4 |
| // CHECK5-NEXT: [[TMP10:%.*]] = zext i32 [[TMP9]] to i64 |
| // CHECK5-NEXT: [[TMP11:%.*]] = load i32, i32* [[DOTOMP_COMB_UB]], align 4 |
| // CHECK5-NEXT: [[TMP12:%.*]] = zext i32 [[TMP11]] to i64 |
| // CHECK5-NEXT: call void (%struct.ident_t*, i32, void (i32*, i32*, ...)*, ...) @__kmpc_fork_call(%struct.ident_t* @[[GLOB3]], i32 2, void (i32*, i32*, ...)* bitcast (void (i32*, i32*, i64, i64)* @.omp_outlined..11 to void (i32*, i32*, ...)*), i64 [[TMP10]], i64 [[TMP12]]) |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK5: omp.inner.for.inc: |
| // CHECK5-NEXT: [[TMP13:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP14:%.*]] = load i32, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 [[TMP13]], [[TMP14]] |
| // CHECK5-NEXT: store i32 [[ADD]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK5: omp.inner.for.end: |
| // CHECK5-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK5: omp.loop.exit: |
| // CHECK5-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP1]]) |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_outlined..11 |
| // CHECK5-SAME: (i32* noalias noundef [[DOTGLOBAL_TID_:%.*]], i32* noalias noundef [[DOTBOUND_TID_:%.*]], i64 noundef [[DOTPREVIOUS_LB_:%.*]], i64 noundef [[DOTPREVIOUS_UB_:%.*]]) #[[ATTR3]] personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[DOTGLOBAL_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTBOUND_TID__ADDR:%.*]] = alloca i32*, align 8 |
| // CHECK5-NEXT: [[DOTPREVIOUS_LB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTPREVIOUS_UB__ADDR:%.*]] = alloca i64, align 8 |
| // CHECK5-NEXT: [[DOTOMP_IV:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[TMP:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_LB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_UB:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_STRIDE:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[DOTOMP_IS_LAST:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: [[I:%.*]] = alloca i32, align 4 |
| // CHECK5-NEXT: store i32* [[DOTGLOBAL_TID_]], i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i32* [[DOTBOUND_TID_]], i32** [[DOTBOUND_TID__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTPREVIOUS_LB_]], i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK5-NEXT: store i64 [[DOTPREVIOUS_UB_]], i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 99, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[TMP0:%.*]] = load i64, i64* [[DOTPREVIOUS_LB__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV:%.*]] = trunc i64 [[TMP0]] to i32 |
| // CHECK5-NEXT: [[TMP1:%.*]] = load i64, i64* [[DOTPREVIOUS_UB__ADDR]], align 8 |
| // CHECK5-NEXT: [[CONV1:%.*]] = trunc i64 [[TMP1]] to i32 |
| // CHECK5-NEXT: store i32 [[CONV]], i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[CONV1]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: store i32 1, i32* [[DOTOMP_STRIDE]], align 4 |
| // CHECK5-NEXT: store i32 0, i32* [[DOTOMP_IS_LAST]], align 4 |
| // CHECK5-NEXT: [[TMP2:%.*]] = load i32*, i32** [[DOTGLOBAL_TID__ADDR]], align 8 |
| // CHECK5-NEXT: [[TMP3:%.*]] = load i32, i32* [[TMP2]], align 4 |
| // CHECK5-NEXT: call void @__kmpc_for_static_init_4(%struct.ident_t* @[[GLOB2]], i32 [[TMP3]], i32 34, i32* [[DOTOMP_IS_LAST]], i32* [[DOTOMP_LB]], i32* [[DOTOMP_UB]], i32* [[DOTOMP_STRIDE]], i32 1, i32 1) |
| // CHECK5-NEXT: [[TMP4:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[CMP:%.*]] = icmp sgt i32 [[TMP4]], 99 |
| // CHECK5-NEXT: br i1 [[CMP]], label [[COND_TRUE:%.*]], label [[COND_FALSE:%.*]] |
| // CHECK5: cond.true: |
| // CHECK5-NEXT: br label [[COND_END:%.*]] |
| // CHECK5: cond.false: |
| // CHECK5-NEXT: [[TMP5:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: br label [[COND_END]] |
| // CHECK5: cond.end: |
| // CHECK5-NEXT: [[COND:%.*]] = phi i32 [ 99, [[COND_TRUE]] ], [ [[TMP5]], [[COND_FALSE]] ] |
| // CHECK5-NEXT: store i32 [[COND]], i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[TMP6:%.*]] = load i32, i32* [[DOTOMP_LB]], align 4 |
| // CHECK5-NEXT: store i32 [[TMP6]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND:%.*]] |
| // CHECK5: omp.inner.for.cond: |
| // CHECK5-NEXT: [[TMP7:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[TMP8:%.*]] = load i32, i32* [[DOTOMP_UB]], align 4 |
| // CHECK5-NEXT: [[CMP2:%.*]] = icmp sle i32 [[TMP7]], [[TMP8]] |
| // CHECK5-NEXT: br i1 [[CMP2]], label [[OMP_INNER_FOR_BODY:%.*]], label [[OMP_INNER_FOR_END:%.*]] |
| // CHECK5: omp.inner.for.body: |
| // CHECK5-NEXT: [[TMP9:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[MUL:%.*]] = mul nsw i32 [[TMP9]], 1 |
| // CHECK5-NEXT: [[ADD:%.*]] = add nsw i32 0, [[MUL]] |
| // CHECK5-NEXT: store i32 [[ADD]], i32* [[I]], align 4 |
| // CHECK5-NEXT: invoke void @_Z3foov() |
| // CHECK5-NEXT: to label [[INVOKE_CONT:%.*]] unwind label [[TERMINATE_LPAD:%.*]] |
| // CHECK5: invoke.cont: |
| // CHECK5-NEXT: br label [[OMP_BODY_CONTINUE:%.*]] |
| // CHECK5: omp.body.continue: |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_INC:%.*]] |
| // CHECK5: omp.inner.for.inc: |
| // CHECK5-NEXT: [[TMP10:%.*]] = load i32, i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: [[ADD3:%.*]] = add nsw i32 [[TMP10]], 1 |
| // CHECK5-NEXT: store i32 [[ADD3]], i32* [[DOTOMP_IV]], align 4 |
| // CHECK5-NEXT: br label [[OMP_INNER_FOR_COND]] |
| // CHECK5: omp.inner.for.end: |
| // CHECK5-NEXT: br label [[OMP_LOOP_EXIT:%.*]] |
| // CHECK5: omp.loop.exit: |
| // CHECK5-NEXT: call void @__kmpc_for_static_fini(%struct.ident_t* @[[GLOB1]], i32 [[TMP3]]) |
| // CHECK5-NEXT: ret void |
| // CHECK5: terminate.lpad: |
| // CHECK5-NEXT: [[TMP11:%.*]] = landingpad { i8*, i32 } |
| // CHECK5-NEXT: catch i8* null |
| // CHECK5-NEXT: [[TMP12:%.*]] = extractvalue { i8*, i32 } [[TMP11]], 0 |
| // CHECK5-NEXT: call void @__clang_call_terminate(i8* [[TMP12]]) #[[ATTR9]] |
| // CHECK5-NEXT: unreachable |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@_ZN1SD2Ev |
| // CHECK5-SAME: (%struct.S* noundef nonnull align 8 dereferenceable(24) [[THIS:%.*]]) unnamed_addr #[[ATTR7]] comdat align 2 { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: [[THIS_ADDR:%.*]] = alloca %struct.S*, align 8 |
| // CHECK5-NEXT: store %struct.S* [[THIS]], %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK5-NEXT: [[THIS1:%.*]] = load %struct.S*, %struct.S** [[THIS_ADDR]], align 8 |
| // CHECK5-NEXT: ret void |
| // |
| // |
| // CHECK5-LABEL: define {{[^@]+}}@.omp_offloading.requires_reg |
| // CHECK5-SAME: () #[[ATTR8:[0-9]+]] section ".text.startup" { |
| // CHECK5-NEXT: entry: |
| // CHECK5-NEXT: call void @__tgt_register_requires(i64 1) |
| // CHECK5-NEXT: ret void |
| // |
