1 // RUN: %clang_cc1 -DCHECK -verify -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-64 2 // RUN: %clang_cc1 -DCHECK -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s 3 // RUN: %clang_cc1 -DCHECK -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-64 4 // RUN: %clang_cc1 -DCHECK -verify -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-32 5 // RUN: %clang_cc1 -DCHECK -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s 6 // RUN: %clang_cc1 -DCHECK -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix CHECK --check-prefix CHECK-32 7 8 // RUN: %clang_cc1 -DCHECK -verify -fopenmp-simd -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck --check-prefix SIMD-ONLY0 %s 9 // RUN: %clang_cc1 -DCHECK -fopenmp-simd -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s 10 // RUN: %clang_cc1 -DCHECK -fopenmp-simd -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck --check-prefix SIMD-ONLY0 %s 11 // RUN: %clang_cc1 -DCHECK -verify -fopenmp-simd -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -o - | FileCheck --check-prefix SIMD-ONLY0 %s 12 // RUN: %clang_cc1 -DCHECK -fopenmp-simd -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s 13 // RUN: %clang_cc1 -DCHECK -fopenmp-simd -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck --check-prefix SIMD-ONLY0 %s 14 // SIMD-ONLY0-NOT: {{__kmpc|__tgt}} 15 16 // RUN: %clang_cc1 -DLAMBDA -verify -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-64 17 // RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s 18 // RUN: %clang_cc1 -DLAMBDA -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix LAMBDA --check-prefix LAMBDA-64 19 20 // RUN: %clang_cc1 -DLAMBDA -verify -fopenmp-simd -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -o - | FileCheck --check-prefix SIMD-ONLY1 %s 21 // RUN: %clang_cc1 -DLAMBDA -fopenmp-simd -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s 22 // RUN: %clang_cc1 -DLAMBDA -fopenmp-simd -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -include-pch %t -verify %s -emit-llvm -o - | FileCheck --check-prefix SIMD-ONLY1 %s 23 // SIMD-ONLY1-NOT: {{__kmpc|__tgt}} 24 25 // expected-no-diagnostics 26 #ifndef HEADER 27 #define HEADER 28 29 int x; 30 #pragma omp threadprivate(x) 31 32 template <typename T> 33 T tmain() { 34 int a[2]; 35 #pragma omp target 36 #pragma omp teams distribute parallel for copyin(x) 37 for (int i = 0; i < 2; ++i) { 38 a[i] = x; 39 } 40 return T(); 41 } 42 43 int main() { 44 int a[2]; 45 #ifdef LAMBDA 46 // LAMBDA-LABEL: @main 47 // LAMBDA: call void [[OUTER_LAMBDA:@.+]]( 48 [&]() { 49 // LAMBDA: define{{.*}} internal{{.*}} void [[OUTER_LAMBDA]]( 50 // LAMBDA: call i32 @__tgt_target_teams_mapper(%struct.ident_t* @{{.+}}, 51 // LAMBDA: call void @[[LOFFL1:.+]]( 52 // LAMBDA: ret 53 #pragma omp target 54 #pragma omp teams distribute parallel for copyin(x) 55 for (int i = 0; i < 2; ++i) { 56 // LAMBDA: define{{.*}} internal{{.*}} void @[[LOFFL1]]( 57 // LAMBDA: call void {{.+}} @__kmpc_fork_teams({{.+}}, i32 {{.+}}, {{.+}} @[[LOUTL1:.+]] to {{.+}}) 58 // LAMBDA: ret void 59 60 // LAMBDA: define internal void @[[LOUTL1]]( 61 // Skip global, bound tid and loop vars 62 // LAMBDA: {{.+}} = alloca i32*, 63 // LAMBDA: {{.+}} = alloca i32*, 64 // LAMBDA: alloca i32, 65 // LAMBDA: alloca i32, 66 // LAMBDA: alloca i32, 67 // LAMBDA: alloca i32, 68 // LAMBDA: alloca i32, 69 a[i] = x; 70 71 // LAMBDA: call void @__kmpc_for_static_init_4( 72 // LAMBDA: call void {{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}} @[[LPAR_OUTL:.+]] to 73 // LAMBDA: call void @__kmpc_for_static_fini( 74 // LAMBDA: ret void 75 76 // LAMBDA: define internal void @[[LPAR_OUTL]]({{.+}}) 77 // Skip global, bound tid and loop vars 78 // LAMBDA: {{.+}} = alloca i32*, 79 // LAMBDA: {{.+}} = alloca i32*, 80 // LAMBDA: alloca i{{[0-9]+}}, 81 // LAMBDA: alloca i{{[0-9]+}}, 82 // LAMBDA: {{%.+}} = alloca [2 x i{{[0-9]+}}]*, 83 // LAMBDA: [[X_ADDR:%.+]] = alloca i{{[0-9]+}}*, 84 // LAMBDA: alloca i32, 85 // LAMBDA: alloca i32, 86 // LAMBDA: alloca i32, 87 // LAMBDA: alloca i32, 88 // LAMBDA: alloca i32, 89 90 // LAMBDA: [[X_REF:%.+]] = load {{.+}}, {{.+}} [[X_ADDR]], 91 92 // LAMBDA: call void @__kmpc_for_static_init_4( 93 // LAMBDA: [[X_VAL:%.+]] = load {{.+}}, {{.+}} [[X_REF]], 94 // LAMBDA: store {{.+}} [[X_VAL]], 95 // LAMBDA: call void [[INNER_LAMBDA:@.+]]( 96 // LAMBDA: call void @__kmpc_for_static_fini( 97 // LAMBDA: ret void 98 [&]() { 99 // LAMBDA: define {{.+}} void [[INNER_LAMBDA]]({{.+}} [[ARG_PTR:%.+]]) 100 // LAMBDA: store %{{.+}}* [[ARG_PTR]], %{{.+}}** [[ARG_PTR_REF:%.+]], 101 a[i] = x; 102 }(); 103 } 104 }(); 105 return 0; 106 #else 107 #pragma omp target 108 #pragma omp teams distribute parallel for copyin(x) 109 for (int i = 0; i < 2; ++i) { 110 a[i] = x; 111 } 112 return tmain<int>(); 113 //return 0; 114 #endif 115 } 116 117 // CHECK: define {{.*}}i{{[0-9]+}} @main() 118 // CHECK: call i32 @__tgt_target_teams_mapper(%struct.ident_t* @{{.+}}, 119 // CHECK: call void @[[OFFL1:.+]](i{{64|32}} %{{.+}}) 120 // CHECK: {{%.+}} = call{{.*}} i32 @[[TMAIN_INT:.+]]() 121 // CHECK: ret 122 123 // CHECK: define{{.*}} void @[[OFFL1]]({{.+}}) 124 // CHECK: call void {{.+}} @__kmpc_fork_teams({{.+}}, i32 {{.+}}, {{.+}} @[[OUTL1:.+]] to {{.+}}) 125 // CHECK: ret void 126 127 // CHECK: define internal void @[[OUTL1]]({{.+}}) 128 // Skip global, bound tid and loop vars 129 // CHECK: {{.+}} = alloca i32*, 130 // CHECK: {{.+}} = alloca i32*, 131 // CHECK: {{.+}} = alloca i32, 132 // CHECK: {{.+}} = alloca i32, 133 // CHECK: {{.+}} = alloca i32, 134 // CHECK: {{.+}} = alloca i32, 135 // CHECK: {{.+}} = alloca i32, 136 137 // CHECK: call void @__kmpc_for_static_init_4( 138 // CHECK: call void {{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}} @[[PAR_OUTL1:.+]] to 139 // CHECK: call void @__kmpc_for_static_fini( 140 // CHECK: ret void 141 142 // CHECK: define internal void @[[PAR_OUTL1]]({{.+}}) 143 // Skip global, bound tid and loop vars 144 // CHECK: {{.+}} = alloca i32*, 145 // CHECK: {{.+}} = alloca i32*, 146 // CHECK: {{.+}} = alloca i{{[0-9]+}}, 147 // CHECK: {{.+}} = alloca i{{[0-9]+}}, 148 // CHECK: {{%.+}} = alloca [2 x i{{[0-9]+}}]*, 149 // CHECK: [[X_ADDR:%.+]] = alloca i{{[0-9]+}}*, 150 // CHECK: {{.+}} = alloca i32, 151 // CHECK: {{.+}} = alloca i32, 152 // CHECK: {{.+}} = alloca i32, 153 // CHECK: {{.+}} = alloca i32, 154 // CHECK: {{.+}} = alloca i32, 155 156 // CHECK: [[X_REF:%.+]] = load {{.+}}, {{.+}} [[X_ADDR]], 157 // CHECK: call void @__kmpc_for_static_init_4( 158 // CHECK: [[X_VAL:%.+]] = load {{.+}}, {{.+}} [[X_REF]], 159 // CHECK: store {{.+}} [[X_VAL]], 160 // CHECK: call void @__kmpc_for_static_fini( 161 // CHECK: ret void 162 163 // CHECK: define{{.*}} i{{[0-9]+}} @[[TMAIN_INT]]() 164 // CHECK: call i32 @__tgt_target_teams_mapper(%struct.ident_t* @{{.+}}, 165 // CHECK: call void @[[TOFFL1:.+]]( 166 // CHECK: ret 167 168 // CHECK: define {{.*}}void @[[TOFFL1]]( 169 // CHECK: call void {{.+}} @__kmpc_fork_teams({{.+}}, i32 {{.+}}, {{.+}} @[[TOUTL1:.+]] to {{.+}}) 170 // CHECK: ret void 171 172 // CHECK: define internal void @[[TOUTL1]]({{.+}}) 173 // Skip global, bound tid and loop vars 174 // CHECK: {{.+}} = alloca i32*, 175 // CHECK: {{.+}} = alloca i32*, 176 // CHECK: alloca i{{[0-9]+}}, 177 // CHECK: alloca i{{[0-9]+}}, 178 // CHECK: alloca i{{[0-9]+}}, 179 // CHECK: alloca i{{[0-9]+}}, 180 // CHECK: alloca i{{[0-9]+}}, 181 182 // CHECK: call void @__kmpc_for_static_init_4( 183 // CHECK: call void {{.+}} @__kmpc_fork_call({{.+}}, {{.+}}, {{.+}} @[[TPAR_OUTL1:.+]] to 184 // CHECK: call void @__kmpc_for_static_fini( 185 // CHECK: ret void 186 187 // CHECK: define internal void @[[TPAR_OUTL1]]({{.+}}) 188 // Skip global, bound tid and loop vars 189 // CHECK: {{.+}} = alloca i32*, 190 // CHECK: {{.+}} = alloca i32*, 191 // prev lb and ub 192 // CHECK: alloca i{{[0-9]+}}, 193 // CHECK: alloca i{{[0-9]+}}, 194 195 // CHECK: {{%.+}} = alloca [2 x i{{[0-9]+}}]*, 196 // CHECK: [[X_ADDR:%.+]] = alloca i{{[0-9]+}}*, 197 // iter variables 198 // CHECK: alloca i{{[0-9]+}}, 199 // CHECK: alloca i{{[0-9]+}}, 200 // CHECK: alloca i{{[0-9]+}}, 201 // CHECK: alloca i{{[0-9]+}}, 202 // CHECK: alloca i{{[0-9]+}}, 203 204 // CHECK: [[X_REF:%.+]] = load {{.+}}, {{.+}} [[X_ADDR]], 205 // CHECK: call void @__kmpc_for_static_init_4( 206 // CHECK: [[X_VAL:%.+]] = load {{.+}}, {{.+}} [[X_REF]], 207 // CHECK: store {{.+}} [[X_VAL]], 208 // CHECK: call void @__kmpc_for_static_fini( 209 // CHECK: ret void 210 211 212 #endif 213