1 // Test host codegen. 2 // RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -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 --check-prefix OMP45 3 // RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s 4 // RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -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 --check-prefix OMP45 5 // RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -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 --check-prefix OMP45 6 // RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s 7 // RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -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 --check-prefix OMP45 8 9 // RUN: %clang_cc1 -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 --check-prefix OMP50 10 // RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s 11 // RUN: %clang_cc1 -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 --check-prefix OMP50 12 // RUN: %clang_cc1 -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 --check-prefix OMP50 13 // RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s 14 // RUN: %clang_cc1 -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 --check-prefix OMP50 15 16 // RUN: %clang_cc1 -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 17 // RUN: %clang_cc1 -fopenmp-simd -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -o %t %s 18 // RUN: %clang_cc1 -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 19 // RUN: %clang_cc1 -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 20 // RUN: %clang_cc1 -fopenmp-simd -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -o %t %s 21 // RUN: %clang_cc1 -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 22 // SIMD-ONLY0-NOT: {{__kmpc|__tgt}} 23 24 // Test target codegen - host bc file has to be created first. 25 // RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm-bc %s -o %t-ppc-host.bc 26 // RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck %s --check-prefix TCHECK --check-prefix TCHECK-64 27 // RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o %t %s 28 // RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix TCHECK --check-prefix TCHECK-64 29 // RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm-bc %s -o %t-x86-host.bc 30 // RUN: %clang_cc1 -verify -fopenmp -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --check-prefix TCHECK --check-prefix TCHECK-32 31 // RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o %t %s 32 // RUN: %clang_cc1 -fopenmp -fopenmp-version=45 -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix TCHECK --check-prefix TCHECK-32 33 34 // RUN: %clang_cc1 -verify -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm-bc %s -o %t-ppc-host.bc 35 // RUN: %clang_cc1 -verify -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck %s --check-prefix TCHECK --check-prefix TCHECK-64 36 // RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o %t %s 37 // RUN: %clang_cc1 -fopenmp -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix TCHECK --check-prefix TCHECK-64 38 // RUN: %clang_cc1 -verify -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm-bc %s -o %t-x86-host.bc 39 // RUN: %clang_cc1 -verify -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck %s --check-prefix TCHECK --check-prefix TCHECK-32 40 // RUN: %clang_cc1 -fopenmp -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o %t %s 41 // RUN: %clang_cc1 -fopenmp -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck %s --check-prefix TCHECK --check-prefix TCHECK-32 42 43 // RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm-bc %s -o %t-ppc-host.bc 44 // RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o - | FileCheck --check-prefix SIMD-ONLY1 %s 45 // RUN: %clang_cc1 -fopenmp-simd -x c++ -std=c++11 -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -o %t %s 46 // RUN: %clang_cc1 -fopenmp-simd -x c++ -triple powerpc64le-unknown-unknown -fopenmp-targets=powerpc64le-ibm-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-ppc-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck --check-prefix SIMD-ONLY1 %s 47 // RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm-bc %s -o %t-x86-host.bc 48 // RUN: %clang_cc1 -verify -fopenmp-simd -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-llvm %s -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o - | FileCheck --check-prefix SIMD-ONLY1 %s 49 // RUN: %clang_cc1 -fopenmp-simd -x c++ -std=c++11 -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -emit-pch -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -o %t %s 50 // RUN: %clang_cc1 -fopenmp-simd -x c++ -triple i386-unknown-unknown -fopenmp-targets=i386-pc-linux-gnu -std=c++11 -fopenmp-is-device -fopenmp-host-ir-file-path %t-x86-host.bc -include-pch %t -verify %s -emit-llvm -o - | FileCheck --check-prefix SIMD-ONLY1 %s 51 // SIMD-ONLY1-NOT: {{__kmpc|__tgt}} 52 53 // expected-no-diagnostics 54 #ifndef HEADER 55 #define HEADER 56 57 // CHECK-DAG: [[TT:%.+]] = type { i64, i8 } 58 // CHECK-DAG: [[S1:%.+]] = type { double } 59 // CHECK-DAG: [[S2:%.+]] = type { i32, i32, i32 } 60 // CHECK-DAG: [[ENTTY:%.+]] = type { i8*, i8*, i[[SZ:32|64]], i32, i32 } 61 62 // TCHECK: [[ENTTY:%.+]] = type { i8*, i8*, i{{32|64}}, i32, i32 } 63 64 // We have 9 target regions, but only 8 that actually will generate offloading 65 // code and have mapped arguments, and only 6 have all-constant map sizes. 66 67 // CHECK-DAG: [[SIZET:@.+]] = private unnamed_addr constant [2 x i64] [i64 0, i64 4] 68 // CHECK-DAG: [[MAPT:@.+]] = private unnamed_addr constant [2 x i64] [i64 544, i64 800] 69 // CHECK-DAG: [[SIZET2:@.+]] = private unnamed_addr constant [1 x i{{32|64}}] [i64 2] 70 // CHECK-DAG: [[MAPT2:@.+]] = private unnamed_addr constant [1 x i64] [i64 800] 71 // CHECK-DAG: [[SIZET3:@.+]] = private unnamed_addr constant [2 x i64] [i64 4, i64 2] 72 // CHECK-DAG: [[MAPT3:@.+]] = private unnamed_addr constant [2 x i64] [i64 800, i64 800] 73 // CHECK-DAG: [[MAPT4:@.+]] = private unnamed_addr constant [9 x i64] [i64 800, i64 547, i64 800, i64 547, i64 547, i64 800, i64 800, i64 547, i64 547] 74 // CHECK-DAG: [[SIZET5:@.+]] = private unnamed_addr constant [3 x i64] [i64 4, i64 2, i64 40] 75 // CHECK-DAG: [[MAPT5:@.+]] = private unnamed_addr constant [3 x i64] [i64 800, i64 800, i64 547] 76 // CHECK-DAG: [[SIZET6:@.+]] = private unnamed_addr constant [4 x i64] [i64 4, i64 2, i64 1, i64 40] 77 // CHECK-DAG: [[MAPT6:@.+]] = private unnamed_addr constant [4 x i64] [i64 800, i64 800, i64 800, i64 547] 78 // CHECK-DAG: [[MAPT7:@.+]] = private unnamed_addr constant [6 x i64] [i64 32, i64 281474976711171, i64 800, i64 800, i64 800, i64 547] 79 // CHECK-DAG: [[SIZET9:@.+]] = private unnamed_addr constant [1 x i64] [i64 12] 80 // CHECK-DAG: [[MAPT10:@.+]] = private unnamed_addr constant [1 x i64] [i64 35] 81 // CHECK-DAG: @{{.*}} = weak constant i8 0 82 // CHECK-DAG: @{{.*}} = weak constant i8 0 83 // CHECK-DAG: @{{.*}} = weak constant i8 0 84 // CHECK-DAG: @{{.*}} = weak constant i8 0 85 // CHECK-DAG: @{{.*}} = weak constant i8 0 86 // CHECK-DAG: @{{.*}} = weak constant i8 0 87 // CHECK-DAG: @{{.*}} = weak constant i8 0 88 // CHECK-DAG: @{{.*}} = weak constant i8 0 89 90 // TCHECK: @{{.+}} = weak constant [[ENTTY]] 91 // TCHECK: @{{.+}} = weak constant [[ENTTY]] 92 // TCHECK: @{{.+}} = weak constant [[ENTTY]] 93 // TCHECK: @{{.+}} = weak constant [[ENTTY]] 94 // TCHECK: @{{.+}} = weak constant [[ENTTY]] 95 // TCHECK: @{{.+}} = weak constant [[ENTTY]] 96 // TCHECK: @{{.+}} = weak constant [[ENTTY]] 97 // TCHECK: @{{.+}} = weak constant [[ENTTY]] 98 // TCHECK: @{{.+}} = weak constant [[ENTTY]] 99 // TCHECK: @{{.+}} = weak constant [[ENTTY]] 100 // TCHECK-NOT: @{{.+}} = weak constant [[ENTTY]] 101 102 // Check target registration is registered as a Ctor. 103 // CHECK: appending global [1 x { i32, void ()*, i8* }] [{ i32, void ()*, i8* } { i32 0, void ()* @.omp_offloading.requires_reg, i8* null }] 104 105 106 template<typename tx, typename ty> 107 struct TT{ 108 tx X; 109 ty Y; 110 }; 111 112 int global; 113 extern int global; 114 115 // CHECK: define {{.*}}[[FOO:@.+]]( 116 int foo(int n) { 117 int a = 0; 118 short aa = 0; 119 float b[10]; 120 float bn[n]; 121 double c[5][10]; 122 double cn[5][n]; 123 TT<long long, char> d; 124 static long *plocal; 125 126 // CHECK: [[ADD:%.+]] = add nsw i32 127 // CHECK: store i32 [[ADD]], i32* [[DEVICE_CAP:%.+]], 128 // CHECK: [[DEV:%.+]] = load i32, i32* [[DEVICE_CAP]], 129 // CHECK: [[DEVICE:%.+]] = sext i32 [[DEV]] to i64 130 // CHECK: [[RET:%.+]] = call i32 @__tgt_target_mapper(i64 [[DEVICE]], i8* @{{[^,]+}}, i32 0, i8** null, i8** null, i64* null, i64* null, i8** null) 131 // CHECK-NEXT: [[ERROR:%.+]] = icmp ne i32 [[RET]], 0 132 // CHECK-NEXT: br i1 [[ERROR]], label %[[FAIL:[^,]+]], label %[[END:[^,]+]] 133 // CHECK: [[FAIL]] 134 // CHECK: call void [[HVT0:@.+]]() 135 // CHECK-NEXT: br label %[[END]] 136 // CHECK: [[END]] 137 #pragma omp target device(global + a) 138 { 139 } 140 141 // CHECK-DAG: [[ADD:%.+]] = add nsw i32 142 // CHECK-DAG: store i32 [[ADD]], i32* [[DEVICE_CAP:%.+]], 143 // CHECK-DAG: [[DEV:%.+]] = load i32, i32* [[DEVICE_CAP]], 144 // CHECK-DAG: [[DEVICE:%.+]] = sext i32 [[DEV]] to i64 145 // CHECK-DAG: [[RET:%.+]] = call i32 @__tgt_target_nowait_mapper(i64 [[DEVICE]], i8* @{{[^,]+}}, i32 2, i8** [[BPR:%[^,]+]], i8** [[PR:%[^,]+]], i64* getelementptr inbounds ([2 x i64], [2 x i64]* [[SIZET]], i32 0, i32 0), i64* getelementptr inbounds ([2 x i64], [2 x i64]* [[MAPT]], i32 0, i32 0), i8** null) 146 // CHECK-DAG: [[BPR]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[BP:%[^,]+]], i32 0, i32 0 147 // CHECK-DAG: [[PR]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[P:%[^,]+]], i32 0, i32 0 148 149 // CHECK-DAG: [[BPADDR0:%.+]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[BP]], i32 0, i32 0 150 // CHECK-DAG: [[PADDR0:%.+]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[P]], i32 0, i32 0 151 // CHECK-DAG: [[CBPADDR0:%.+]] = bitcast i8** [[BPADDR0]] to i[[SZ]]** 152 // CHECK-DAG: [[CPADDR0:%.+]] = bitcast i8** [[PADDR0]] to i[[SZ]]** 153 // CHECK-DAG: store i[[SZ]]* [[BP0:%[^,]+]], i[[SZ]]** [[CBPADDR0]] 154 // CHECK-DAG: store i[[SZ]]* [[BP0]], i[[SZ]]** [[CPADDR0]] 155 156 // CHECK-DAG: [[BPADDR1:%.+]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[BP]], i32 0, i32 1 157 // CHECK-DAG: [[PADDR1:%.+]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[P]], i32 0, i32 1 158 // CHECK-DAG: [[CBPADDR1:%.+]] = bitcast i8** [[BPADDR1]] to i[[SZ]]* 159 // CHECK-DAG: [[CPADDR1:%.+]] = bitcast i8** [[PADDR1]] to i[[SZ]]* 160 // CHECK-DAG: store i[[SZ]] [[BP1:%[^,]+]], i[[SZ]]* [[CBPADDR1]] 161 // CHECK-DAG: store i[[SZ]] [[BP1]], i[[SZ]]* [[CPADDR1]] 162 // CHECK: [[ERROR:%.+]] = icmp ne i32 [[RET]], 0 163 // CHECK-NEXT: br i1 [[ERROR]], label %[[FAIL:[^,]+]], label %[[END:[^,]+]] 164 // CHECK: [[FAIL]] 165 // CHECK: call void [[HVT0_:@.+]](i[[SZ]]* [[BP0]], i[[SZ]] [[BP1]]) 166 // CHECK-NEXT: br label %[[END]] 167 // CHECK: [[END]] 168 #pragma omp target device(global + a) nowait 169 { 170 static int local1; 171 *plocal = global; 172 local1 = global; 173 } 174 175 // CHECK: call void [[HVT1:@.+]](i[[SZ]] {{[^,]+}}) 176 #pragma omp target if(0) firstprivate(global) 177 { 178 global += 1; 179 } 180 181 // CHECK-DAG: [[RET:%.+]] = call i32 @__tgt_target_mapper(i64 -1, i8* @{{[^,]+}}, i32 1, i8** [[BP:%[^,]+]], i8** [[P:%[^,]+]], i64* getelementptr inbounds ([1 x i64], [1 x i64]* [[SIZET2]], i32 0, i32 0), i64* getelementptr inbounds ([1 x i64], [1 x i64]* [[MAPT2]], i32 0, i32 0), i8** null) 182 // CHECK-DAG: [[BP]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[BPR:%[^,]+]], i32 0, i32 0 183 // CHECK-DAG: [[P]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[PR:%[^,]+]], i32 0, i32 0 184 // CHECK-DAG: [[BPADDR0:%.+]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[BPR]], i32 0, i32 [[IDX0:[0-9]+]] 185 // CHECK-DAG: [[PADDR0:%.+]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[PR]], i32 0, i32 [[IDX0]] 186 // CHECK-DAG: [[CBPADDR0:%.+]] = bitcast i8** [[BPADDR0]] to i[[SZ]]* 187 // CHECK-DAG: [[CPADDR0:%.+]] = bitcast i8** [[PADDR0]] to i[[SZ]]* 188 // CHECK-DAG: store i[[SZ]] [[BP0:%[^,]+]], i[[SZ]]* [[CBPADDR0]] 189 // CHECK-DAG: store i[[SZ]] [[P0:%[^,]+]], i[[SZ]]* [[CPADDR0]] 190 191 // CHECK: [[ERROR:%.+]] = icmp ne i32 [[RET]], 0 192 // CHECK-NEXT: br i1 [[ERROR]], label %[[FAIL:[^,]+]], label %[[END:[^,]+]] 193 // CHECK: [[FAIL]] 194 // CHECK: call void [[HVT2:@.+]](i[[SZ]] {{[^,]+}}) 195 // CHECK-NEXT: br label %[[END]] 196 // CHECK: [[END]] 197 #pragma omp target if(1) 198 { 199 aa += 1; 200 } 201 202 // CHECK: [[IF:%.+]] = icmp sgt i32 {{[^,]+}}, 10 203 // CHECK: br i1 [[IF]], label %[[IFTHEN:[^,]+]], label %[[IFELSE:[^,]+]] 204 // CHECK: [[IFTHEN]] 205 // CHECK-DAG: [[RET:%.+]] = call i32 @__tgt_target_mapper(i64 -1, i8* @{{[^,]+}}, i32 2, i8** [[BPR:%[^,]+]], i8** [[PR:%[^,]+]], i64* getelementptr inbounds ([2 x i64], [2 x i64]* [[SIZET3]], i32 0, i32 0), i64* getelementptr inbounds ([2 x i64], [2 x i64]* [[MAPT3]], i32 0, i32 0), i8** null) 206 // CHECK-DAG: [[BPR]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[BP:%[^,]+]], i32 0, i32 0 207 // CHECK-DAG: [[PR]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[P:%[^,]+]], i32 0, i32 0 208 209 // CHECK-DAG: [[BPADDR0:%.+]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[BP]], i32 0, i32 0 210 // CHECK-DAG: [[PADDR0:%.+]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[P]], i32 0, i32 0 211 // CHECK-DAG: [[CBPADDR0:%.+]] = bitcast i8** [[BPADDR0]] to i[[SZ]]* 212 // CHECK-DAG: [[CPADDR0:%.+]] = bitcast i8** [[PADDR0]] to i[[SZ]]* 213 // CHECK-DAG: store i[[SZ]] [[BP0:%[^,]+]], i[[SZ]]* [[CBPADDR0]] 214 // CHECK-DAG: store i[[SZ]] [[P0:%[^,]+]], i[[SZ]]* [[CPADDR0]] 215 216 // CHECK-DAG: [[BPADDR1:%.+]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[BP]], i32 0, i32 1 217 // CHECK-DAG: [[PADDR1:%.+]] = getelementptr inbounds [2 x i8*], [2 x i8*]* [[P]], i32 0, i32 1 218 // CHECK-DAG: [[CBPADDR1:%.+]] = bitcast i8** [[BPADDR1]] to i[[SZ]]* 219 // CHECK-DAG: [[CPADDR1:%.+]] = bitcast i8** [[PADDR1]] to i[[SZ]]* 220 // CHECK-DAG: store i[[SZ]] [[BP1:%[^,]+]], i[[SZ]]* [[CBPADDR1]] 221 // CHECK-DAG: store i[[SZ]] [[P1:%[^,]+]], i[[SZ]]* [[CPADDR1]] 222 // CHECK: [[ERROR:%.+]] = icmp ne i32 [[RET]], 0 223 // CHECK-NEXT: br i1 [[ERROR]], label %[[FAIL:.+]], label %[[END:[^,]+]] 224 // CHECK: [[FAIL]] 225 // CHECK: call void [[HVT3:@.+]]({{[^,]+}}, {{[^,]+}}) 226 // CHECK-NEXT: br label %[[END]] 227 // CHECK: [[END]] 228 // CHECK-NEXT: br label %[[IFEND:.+]] 229 // CHECK: [[IFELSE]] 230 // CHECK: call void [[HVT3]]({{[^,]+}}, {{[^,]+}}) 231 // CHECK-NEXT: br label %[[IFEND]] 232 233 // CHECK: [[IFEND]] 234 #pragma omp target if(n>10) 235 { 236 a += 1; 237 aa += 1; 238 } 239 240 // We capture 3 VLA sizes in this target region 241 // CHECK-64: [[A_VAL:%.+]] = load i32, i32* %{{.+}}, 242 // CHECK-64: [[A_ADDR:%.+]] = bitcast i[[SZ]]* [[A_CADDR:%.+]] to i32* 243 // CHECK-64: store i32 [[A_VAL]], i32* [[A_ADDR]], 244 // CHECK-64: [[A_CVAL:%.+]] = load i[[SZ]], i[[SZ]]* [[A_CADDR]], 245 246 // CHECK-32: [[A_VAL:%.+]] = load i32, i32* %{{.+}}, 247 // CHECK-32: store i32 [[A_VAL]], i32* [[A_CADDR:%.+]], 248 // CHECK-32: [[A_CVAL:%.+]] = load i[[SZ]], i[[SZ]]* [[A_CADDR]], 249 250 // CHECK: [[IF:%.+]] = icmp sgt i32 {{[^,]+}}, 20 251 // CHECK: br i1 [[IF]], label %[[TRY:[^,]+]], label %[[IFELSE:[^,]+]] 252 // CHECK: [[TRY]] 253 // CHECK-64: [[BNSIZE:%.+]] = mul nuw i64 [[VLA0:%.+]], 4 254 // CHECK-32: [[BNSZSIZE:%.+]] = mul nuw i32 [[VLA0:%.+]], 4 255 // CHECK-32: [[BNSIZE:%.+]] = sext i32 [[BNSZSIZE]] to i64 256 // CHECK: [[CNELEMSIZE2:%.+]] = mul nuw i[[SZ]] 5, [[VLA1:%.+]] 257 // CHECK-64: [[CNSIZE:%.+]] = mul nuw i64 [[CNELEMSIZE2]], 8 258 // CHECK-32: [[CNSZSIZE:%.+]] = mul nuw i32 [[CNELEMSIZE2]], 8 259 // CHECK-32: [[CNSIZE:%.+]] = sext i32 [[CNSZSIZE]] to i64 260 261 // CHECK-DAG: [[RET:%.+]] = call i32 @__tgt_target_mapper(i64 -1, i8* @{{[^,]+}}, i32 9, i8** [[BPR:%[^,]+]], i8** [[PR:%[^,]+]], i64* [[SR:%[^,]+]], i64* getelementptr inbounds ([9 x i64], [9 x i64]* [[MAPT4]], i32 0, i32 0), i8** null) 262 // CHECK-DAG: [[BPR]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[BP:%[^,]+]], i32 0, i32 0 263 // CHECK-DAG: [[PR]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[P:%[^,]+]], i32 0, i32 0 264 // CHECK-DAG: [[SR]] = getelementptr inbounds [9 x i64], [9 x i64]* [[S:%[^,]+]], i32 0, i32 0 265 266 // CHECK-DAG: [[SADDR0:%.+]] = getelementptr inbounds [9 x i64], [9 x i64]* [[S]], i32 0, i32 [[IDX0:0]] 267 // CHECK-DAG: [[BPADDR0:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[BP]], i32 0, i32 [[IDX0]] 268 // CHECK-DAG: [[PADDR0:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[P]], i32 0, i32 [[IDX0]] 269 // CHECK-DAG: [[SADDR1:%.+]] = getelementptr inbounds [9 x i64], [9 x i64]* [[S]], i32 0, i32 [[IDX1:1]] 270 // CHECK-DAG: [[BPADDR1:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[BP]], i32 0, i32 [[IDX1]] 271 // CHECK-DAG: [[PADDR1:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[P]], i32 0, i32 [[IDX1]] 272 // CHECK-DAG: [[SADDR2:%.+]] = getelementptr inbounds [9 x i64], [9 x i64]* [[S]], i32 0, i32 [[IDX2:2]] 273 // CHECK-DAG: [[BPADDR2:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[BP]], i32 0, i32 [[IDX2]] 274 // CHECK-DAG: [[PADDR2:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[P]], i32 0, i32 [[IDX2]] 275 // CHECK-DAG: [[SADDR3:%.+]] = getelementptr inbounds [9 x i64], [9 x i64]* [[S]], i32 0, i32 [[IDX3:3]] 276 // CHECK-DAG: [[BPADDR3:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[BP]], i32 0, i32 [[IDX3]] 277 // CHECK-DAG: [[PADDR3:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[P]], i32 0, i32 [[IDX3]] 278 // CHECK-DAG: [[SADDR4:%.+]] = getelementptr inbounds [9 x i64], [9 x i64]* [[S]], i32 0, i32 [[IDX4:4]] 279 // CHECK-DAG: [[BPADDR4:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[BP]], i32 0, i32 [[IDX4]] 280 // CHECK-DAG: [[PADDR4:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[P]], i32 0, i32 [[IDX4]] 281 // CHECK-DAG: [[SADDR5:%.+]] = getelementptr inbounds [9 x i64], [9 x i64]* [[S]], i32 0, i32 [[IDX5:5]] 282 // CHECK-DAG: [[BPADDR5:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[BP]], i32 0, i32 [[IDX5]] 283 // CHECK-DAG: [[PADDR5:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[P]], i32 0, i32 [[IDX5]] 284 // CHECK-DAG: [[SADDR6:%.+]] = getelementptr inbounds [9 x i64], [9 x i64]* [[S]], i32 0, i32 [[IDX6:6]] 285 // CHECK-DAG: [[BPADDR6:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[BP]], i32 0, i32 [[IDX6]] 286 // CHECK-DAG: [[PADDR6:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[P]], i32 0, i32 [[IDX6]] 287 // CHECK-DAG: [[SADDR7:%.+]] = getelementptr inbounds [9 x i64], [9 x i64]* [[S]], i32 0, i32 [[IDX7:7]] 288 // CHECK-DAG: [[BPADDR7:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[BP]], i32 0, i32 [[IDX7]] 289 // CHECK-DAG: [[PADDR7:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[P]], i32 0, i32 [[IDX7]] 290 // CHECK-DAG: [[SADDR8:%.+]] = getelementptr inbounds [9 x i64], [9 x i64]* [[S]], i32 0, i32 [[IDX8:8]] 291 // CHECK-DAG: [[BPADDR8:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[BP]], i32 0, i32 [[IDX8]] 292 // CHECK-DAG: [[PADDR8:%.+]] = getelementptr inbounds [9 x i8*], [9 x i8*]* [[P]], i32 0, i32 [[IDX8]] 293 294 // The names below are not necessarily consistent with the names used for the 295 // addresses above as some are repeated. 296 // CHECK-DAG: [[CBPADDR2:%.+]] = bitcast i8** [[BPADDR2]] to i[[SZ]]* 297 // CHECK-DAG: [[CPADDR2:%.+]] = bitcast i8** [[PADDR2]] to i[[SZ]]* 298 // CHECK-DAG: store i[[SZ]] [[VLA0]], i[[SZ]]* [[CBPADDR2]] 299 // CHECK-DAG: store i[[SZ]] [[VLA0]], i[[SZ]]* [[CPADDR2]] 300 // CHECK-DAG: store i64 {{4|8}}, i64* [[SADDR2]] 301 302 // CHECK-DAG: [[CBPADDR6:%.+]] = bitcast i8** [[BPADDR6]] to i[[SZ]]* 303 // CHECK-DAG: [[CPADDR6:%.+]] = bitcast i8** [[PADDR6]] to i[[SZ]]* 304 // CHECK-DAG: store i[[SZ]] [[VLA1]], i[[SZ]]* [[CBPADDR6]] 305 // CHECK-DAG: store i[[SZ]] [[VLA1]], i[[SZ]]* [[CPADDR6]] 306 // CHECK-DAG: store i64 {{4|8}}, i64* [[SADDR6]] 307 308 // CHECK-DAG: [[CBPADDR5:%.+]] = bitcast i8** [[BPADDR5]] to i[[SZ]]* 309 // CHECK-DAG: [[CPADDR5:%.+]] = bitcast i8** [[PADDR5]] to i[[SZ]]* 310 // CHECK-DAG: store i[[SZ]] 5, i[[SZ]]* [[CBPADDR5]] 311 // CHECK-DAG: store i[[SZ]] 5, i[[SZ]]* [[CPADDR5]] 312 // CHECK-DAG: store i64 {{4|8}}, i64* [[SADDR5]] 313 314 // CHECK-DAG: [[CBPADDR0:%.+]] = bitcast i8** [[BPADDR0]] to i[[SZ]]* 315 // CHECK-DAG: [[CPADDR0:%.+]] = bitcast i8** [[PADDR0]] to i[[SZ]]* 316 // CHECK-DAG: store i[[SZ]] [[A_CVAL]], i[[SZ]]* [[CBPADDR0]] 317 // CHECK-DAG: store i[[SZ]] [[A_CVAL]], i[[SZ]]* [[CPADDR0]] 318 // CHECK-DAG: store i64 4, i64* [[SADDR0]] 319 320 // CHECK-DAG: [[CBPADDR1:%.+]] = bitcast i8** [[BPADDR1]] to [10 x float]** 321 // CHECK-DAG: [[CPADDR1:%.+]] = bitcast i8** [[PADDR1]] to [10 x float]** 322 // CHECK-DAG: store [10 x float]* %{{.+}}, [10 x float]** [[CBPADDR1]] 323 // CHECK-DAG: store [10 x float]* %{{.+}}, [10 x float]** [[CPADDR1]] 324 // CHECK-DAG: store i64 40, i64* [[SADDR1]] 325 326 // CHECK-DAG: [[CBPADDR3:%.+]] = bitcast i8** [[BPADDR3]] to float** 327 // CHECK-DAG: [[CPADDR3:%.+]] = bitcast i8** [[PADDR3]] to float** 328 // CHECK-DAG: store float* %{{.+}}, float** [[CBPADDR3]] 329 // CHECK-DAG: store float* %{{.+}}, float** [[CPADDR3]] 330 // CHECK-DAG: store i64 [[BNSIZE]], i64* [[SADDR3]] 331 332 // CHECK-DAG: [[CBPADDR4:%.+]] = bitcast i8** [[BPADDR4]] to [5 x [10 x double]]** 333 // CHECK-DAG: [[CPADDR4:%.+]] = bitcast i8** [[PADDR4]] to [5 x [10 x double]]** 334 // CHECK-DAG: store [5 x [10 x double]]* %{{.+}}, [5 x [10 x double]]** [[CBPADDR4]] 335 // CHECK-DAG: store [5 x [10 x double]]* %{{.+}}, [5 x [10 x double]]** [[CPADDR4]] 336 // CHECK-DAG: store i64 400, i64* [[SADDR4]] 337 338 // CHECK-DAG: [[CBPADDR7:%.+]] = bitcast i8** [[BPADDR7]] to double** 339 // CHECK-DAG: [[CPADDR7:%.+]] = bitcast i8** [[PADDR7]] to double** 340 // CHECK-DAG: store double* %{{.+}}, double** [[CBPADDR7]] 341 // CHECK-DAG: store double* %{{.+}}, double** [[CPADDR7]] 342 // CHECK-DAG: store i64 [[CNSIZE]], i64* [[SADDR7]] 343 344 // CHECK-DAG: [[CBPADDR8:%.+]] = bitcast i8** [[BPADDR8]] to [[TT]]** 345 // CHECK-DAG: [[CPADDR8:%.+]] = bitcast i8** [[PADDR8]] to [[TT]]** 346 // CHECK-DAG: store [[TT]]* %{{.+}}, [[TT]]** [[CBPADDR8]] 347 // CHECK-DAG: store [[TT]]* %{{.+}}, [[TT]]** [[CPADDR8]] 348 // CHECK-DAG: store i64 {{12|16}}, i64* [[SADDR8]] 349 350 // CHECK: [[ERROR:%.+]] = icmp ne i32 [[RET]], 0 351 // CHECK-NEXT: br i1 [[ERROR]], label %[[FAIL:.+]], label %[[END:[^,]+]] 352 // CHECK: [[FAIL]] 353 // CHECK: call void [[HVT4:@.+]]({{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}) 354 // CHECK-NEXT: br label %[[END]] 355 // CHECK: [[END]] 356 // CHECK-NEXT: br label %[[IFEND:.+]] 357 // CHECK: [[IFELSE]] 358 // CHECK: call void [[HVT4]]({{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}) 359 // CHECK-NEXT: br label %[[IFEND]] 360 361 // CHECK: [[IFEND]] 362 #pragma omp target if(n>20) 363 { 364 a += 1; 365 b[2] += 1.0; 366 bn[3] += 1.0; 367 c[1][2] += 1.0; 368 cn[1][3] += 1.0; 369 d.X += 1; 370 d.Y += 1; 371 } 372 373 return a; 374 } 375 376 // Check that the offloading functions are emitted and that the arguments are 377 // correct and loaded correctly for the target regions in foo(). 378 379 // CHECK: define internal void [[HVT0]]() 380 381 // CHECK: define internal void [[HVT1]](i[[SZ]] %{{.+}}) 382 // Create stack storage and store argument in there. 383 // CHECK: [[AA_ADDR:%.+]] = alloca i[[SZ]], align 384 // CHECK: store i[[SZ]] %{{.+}}, i[[SZ]]* [[AA_ADDR]], align 385 // CHECK-64: [[AA_CADDR:%.+]] = bitcast i[[SZ]]* [[AA_ADDR]] to i32* 386 // CHECK-64: load i32, i32* [[AA_CADDR]], align 387 // CHECK-32: load i32, i32* [[AA_ADDR]], align 388 389 // CHECK: define internal void [[HVT2]](i[[SZ]] %{{.+}}) 390 // Create stack storage and store argument in there. 391 // CHECK: [[AA_ADDR:%.+]] = alloca i[[SZ]], align 392 // CHECK: store i[[SZ]] %{{.+}}, i[[SZ]]* [[AA_ADDR]], align 393 // CHECK: [[AA_CADDR:%.+]] = bitcast i[[SZ]]* [[AA_ADDR]] to i16* 394 // CHECK: load i16, i16* [[AA_CADDR]], align 395 396 // CHECK: define internal void [[HVT3]] 397 // Create stack storage and store argument in there. 398 // CHECK: [[A_ADDR:%.+]] = alloca i[[SZ]], align 399 // CHECK: [[AA_ADDR:%.+]] = alloca i[[SZ]], align 400 // CHECK-DAG: store i[[SZ]] %{{.+}}, i[[SZ]]* [[A_ADDR]], align 401 // CHECK-DAG: store i[[SZ]] %{{.+}}, i[[SZ]]* [[AA_ADDR]], align 402 // CHECK-64-DAG:[[A_CADDR:%.+]] = bitcast i[[SZ]]* [[A_ADDR]] to i32* 403 // CHECK-DAG: [[AA_CADDR:%.+]] = bitcast i[[SZ]]* [[AA_ADDR]] to i16* 404 // CHECK-64-DAG:load i32, i32* [[A_CADDR]], align 405 // CHECK-32-DAG:load i32, i32* [[A_ADDR]], align 406 // CHECK-DAG: load i16, i16* [[AA_CADDR]], align 407 408 // CHECK: define internal void [[HVT4]] 409 // Create local storage for each capture. 410 // CHECK: [[LOCAL_A:%.+]] = alloca i[[SZ]] 411 // CHECK: [[LOCAL_B:%.+]] = alloca [10 x float]* 412 // CHECK: [[LOCAL_VLA1:%.+]] = alloca i[[SZ]] 413 // CHECK: [[LOCAL_BN:%.+]] = alloca float* 414 // CHECK: [[LOCAL_C:%.+]] = alloca [5 x [10 x double]]* 415 // CHECK: [[LOCAL_VLA2:%.+]] = alloca i[[SZ]] 416 // CHECK: [[LOCAL_VLA3:%.+]] = alloca i[[SZ]] 417 // CHECK: [[LOCAL_CN:%.+]] = alloca double* 418 // CHECK: [[LOCAL_D:%.+]] = alloca [[TT]]* 419 // CHECK-DAG: store i[[SZ]] [[ARG_A:%.+]], i[[SZ]]* [[LOCAL_A]] 420 // CHECK-DAG: store [10 x float]* [[ARG_B:%.+]], [10 x float]** [[LOCAL_B]] 421 // CHECK-DAG: store i[[SZ]] [[ARG_VLA1:%.+]], i[[SZ]]* [[LOCAL_VLA1]] 422 // CHECK-DAG: store float* [[ARG_BN:%.+]], float** [[LOCAL_BN]] 423 // CHECK-DAG: store [5 x [10 x double]]* [[ARG_C:%.+]], [5 x [10 x double]]** [[LOCAL_C]] 424 // CHECK-DAG: store i[[SZ]] [[ARG_VLA2:%.+]], i[[SZ]]* [[LOCAL_VLA2]] 425 // CHECK-DAG: store i[[SZ]] [[ARG_VLA3:%.+]], i[[SZ]]* [[LOCAL_VLA3]] 426 // CHECK-DAG: store double* [[ARG_CN:%.+]], double** [[LOCAL_CN]] 427 // CHECK-DAG: store [[TT]]* [[ARG_D:%.+]], [[TT]]** [[LOCAL_D]] 428 429 // CHECK-64-DAG:[[REF_A:%.+]] = bitcast i[[SZ]]* [[LOCAL_A]] to i32* 430 // CHECK-DAG: [[REF_B:%.+]] = load [10 x float]*, [10 x float]** [[LOCAL_B]], 431 // CHECK-DAG: [[VAL_VLA1:%.+]] = load i[[SZ]], i[[SZ]]* [[LOCAL_VLA1]], 432 // CHECK-DAG: [[REF_BN:%.+]] = load float*, float** [[LOCAL_BN]], 433 // CHECK-DAG: [[REF_C:%.+]] = load [5 x [10 x double]]*, [5 x [10 x double]]** [[LOCAL_C]], 434 // CHECK-DAG: [[VAL_VLA2:%.+]] = load i[[SZ]], i[[SZ]]* [[LOCAL_VLA2]], 435 // CHECK-DAG: [[VAL_VLA3:%.+]] = load i[[SZ]], i[[SZ]]* [[LOCAL_VLA3]], 436 // CHECK-DAG: [[REF_CN:%.+]] = load double*, double** [[LOCAL_CN]], 437 // CHECK-DAG: [[REF_D:%.+]] = load [[TT]]*, [[TT]]** [[LOCAL_D]], 438 439 // Use captures. 440 // CHECK-64-DAG: load i32, i32* [[REF_A]] 441 // CHECK-32-DAG: load i32, i32* [[LOCAL_A]] 442 // CHECK-DAG: getelementptr inbounds [10 x float], [10 x float]* [[REF_B]], i[[SZ]] 0, i[[SZ]] 2 443 // CHECK-DAG: getelementptr inbounds float, float* [[REF_BN]], i[[SZ]] 3 444 // CHECK-DAG: getelementptr inbounds [5 x [10 x double]], [5 x [10 x double]]* [[REF_C]], i[[SZ]] 0, i[[SZ]] 1 445 // CHECK-DAG: getelementptr inbounds double, double* [[REF_CN]], i[[SZ]] %{{.+}} 446 // CHECK-DAG: getelementptr inbounds [[TT]], [[TT]]* [[REF_D]], i32 0, i32 0 447 448 template<typename tx> 449 tx ftemplate(int n) { 450 tx a = 0; 451 short aa = 0; 452 tx b[10]; 453 454 #pragma omp target if(n>40) 455 { 456 a += 1; 457 aa += 1; 458 b[2] += 1; 459 } 460 461 return a; 462 } 463 464 static 465 int fstatic(int n) { 466 int a = 0; 467 short aa = 0; 468 char aaa = 0; 469 int b[10]; 470 471 #pragma omp target if(n>50) 472 { 473 a += 1; 474 aa += 1; 475 aaa += 1; 476 b[2] += 1; 477 } 478 479 return a; 480 } 481 482 struct S1 { 483 double a; 484 485 int r1(int n){ 486 int b = n+1; 487 short int c[2][n]; 488 489 #pragma omp target if(n>60) 490 { 491 this->a = (double)b + 1.5; 492 c[1][1] = ++a; 493 } 494 495 return c[1][1] + (int)b; 496 } 497 }; 498 499 // CHECK: define {{.*}}@{{.*}}bar{{.*}} 500 int bar(int n){ 501 int a = 0; 502 503 // CHECK: call {{.*}}i32 [[FOO]](i32 {{.*}}) 504 a += foo(n); 505 506 S1 S; 507 // CHECK: call {{.*}}i32 [[FS1:@.+]]([[S1]]* {{.*}}, i32 {{.*}}) 508 a += S.r1(n); 509 510 // CHECK: call {{.*}}i32 [[FSTATIC:@.+]](i32 {{.*}}) 511 a += fstatic(n); 512 513 // CHECK: call {{.*}}i32 [[FTEMPLATE:@.+]](i32 {{.*}}) 514 a += ftemplate<int>(n); 515 516 return a; 517 } 518 519 // 520 // CHECK: define {{.*}}[[FS1]] 521 // 522 // CHECK: i8* @llvm.stacksave() 523 // CHECK-64: [[B_ADDR:%.+]] = bitcast i[[SZ]]* [[B_CADDR:%.+]] to i32* 524 // CHECK-64: store i32 %{{.+}}, i32* [[B_ADDR]], 525 // CHECK-64: [[B_CVAL:%.+]] = load i[[SZ]], i[[SZ]]* [[B_CADDR]], 526 527 // CHECK-32: store i32 %{{.+}}, i32* %__vla_expr 528 // CHECK-32: store i32 %{{.+}}, i32* [[B_ADDR:%.+]], 529 // CHECK-32: [[B_CVAL:%.+]] = load i[[SZ]], i[[SZ]]* [[B_ADDR]], 530 531 // CHECK: [[IF:%.+]] = icmp sgt i32 {{[^,]+}}, 60 532 // CHECK: br i1 [[IF]], label %[[TRY:[^,]+]], label %[[IFELSE:[^,]+]] 533 // CHECK: [[TRY]] 534 // We capture 2 VLA sizes in this target region 535 // CHECK: [[CELEMSIZE2:%.+]] = mul nuw i[[SZ]] 2, [[VLA0:%.+]] 536 // CHECK-64: [[CSIZE:%.+]] = mul nuw i64 [[CELEMSIZE2]], 2 537 // CHECK-32: [[CSZSIZE:%.+]] = mul nuw i32 [[CELEMSIZE2]], 2 538 // CHECK-32: [[CSIZE:%.+]] = sext i32 [[CSZSIZE]] to i64 539 540 // CHECK-DAG: [[RET:%.+]] = call i32 @__tgt_target_mapper(i64 -1, i8* @{{[^,]+}}, i32 6, i8** [[BPR:%[^,]+]], i8** [[PR:%[^,]+]], i64* [[SR:%[^,]+]], i64* getelementptr inbounds ([6 x i64], [6 x i64]* [[MAPT7]], i32 0, i32 0), i8** null) 541 // CHECK-DAG: [[BPR]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[BP:%.+]], i32 0, i32 0 542 // CHECK-DAG: [[PR]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[P:%.+]], i32 0, i32 0 543 // CHECK-DAG: [[SR]] = getelementptr inbounds [6 x i64], [6 x i64]* [[S:%.+]], i32 0, i32 0 544 // CHECK-DAG: [[SADDR0:%.+]] = getelementptr inbounds [6 x i64], [6 x i64]* [[S]], i32 0, i32 [[IDX0:0]] 545 // CHECK-DAG: [[BPADDR0:%.+]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[BP]], i32 0, i32 [[IDX0]] 546 // CHECK-DAG: [[PADDR0:%.+]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[P]], i32 0, i32 [[IDX0]] 547 // CHECK-DAG: [[SADDR1:%.+]] = getelementptr inbounds [6 x i64], [6 x i64]* [[S]], i32 0, i32 [[IDX1:1]] 548 // CHECK-DAG: [[BPADDR1:%.+]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[BP]], i32 0, i32 [[IDX1]] 549 // CHECK-DAG: [[PADDR1:%.+]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[P]], i32 0, i32 [[IDX1]] 550 // CHECK-DAG: [[SADDR2:%.+]] = getelementptr inbounds [6 x i64], [6 x i64]* [[S]], i32 0, i32 [[IDX2:2]] 551 // CHECK-DAG: [[BPADDR2:%.+]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[BP]], i32 0, i32 [[IDX2]] 552 // CHECK-DAG: [[PADDR2:%.+]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[P]], i32 0, i32 [[IDX2]] 553 // CHECK-DAG: [[SADDR3:%.+]] = getelementptr inbounds [6 x i64], [6 x i64]* [[S]], i32 0, i32 [[IDX3:3]] 554 // CHECK-DAG: [[BPADDR3:%.+]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[BP]], i32 0, i32 [[IDX3]] 555 // CHECK-DAG: [[PADDR3:%.+]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[P]], i32 0, i32 [[IDX3]] 556 // CHECK-DAG: [[SADDR4:%.+]] = getelementptr inbounds [6 x i64], [6 x i64]* [[S]], i32 0, i32 [[IDX4:4]] 557 // CHECK-DAG: [[BPADDR4:%.+]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[BP]], i32 0, i32 [[IDX4]] 558 // CHECK-DAG: [[PADDR4:%.+]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[P]], i32 0, i32 [[IDX4]] 559 // CHECK-DAG: [[SADDR5:%.+]] = getelementptr inbounds [6 x i64], [6 x i64]* [[S]], i32 0, i32 [[IDX5:5]] 560 // CHECK-DAG: [[BPADDR5:%.+]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[BP]], i32 0, i32 [[IDX5]] 561 // CHECK-DAG: [[PADDR5:%.+]] = getelementptr inbounds [6 x i8*], [6 x i8*]* [[P]], i32 0, i32 [[IDX5]] 562 563 // The names below are not necessarily consistent with the names used for the 564 // addresses above as some are repeated. 565 // CHECK-DAG: [[CBPADDR5:%.+]] = bitcast i8** [[BPADDR5]] to i16** 566 // CHECK-DAG: [[CPADDR5:%.+]] = bitcast i8** [[PADDR5]] to i16** 567 // CHECK-DAG: store i16* %{{.+}}, i16** [[CBPADDR5]] 568 // CHECK-DAG: store i16* %{{.+}}, i16** [[CPADDR5]] 569 // CHECK-DAG: store i64 [[CSIZE]], i64* [[SADDR5]] 570 571 // CHECK-DAG: [[CBPADDR4:%.+]] = bitcast i8** [[BPADDR4]] to i[[SZ]]* 572 // CHECK-DAG: [[CPADDR4:%.+]] = bitcast i8** [[PADDR4]] to i[[SZ]]* 573 // CHECK-DAG: store i[[SZ]] [[VLA0]], i[[SZ]]* [[CBPADDR4]] 574 // CHECK-DAG: store i[[SZ]] [[VLA0]], i[[SZ]]* [[CPADDR4]] 575 // CHECK-DAG: store i64 {{4|8}}, i64* [[SADDR4]] 576 577 // CHECK-DAG: [[CBPADDR3:%.+]] = bitcast i8** [[BPADDR3]] to i[[SZ]]* 578 // CHECK-DAG: [[CPADDR3:%.+]] = bitcast i8** [[PADDR3]] to i[[SZ]]* 579 // CHECK-DAG: store i[[SZ]] 2, i[[SZ]]* [[CBPADDR3]] 580 // CHECK-DAG: store i[[SZ]] 2, i[[SZ]]* [[CPADDR3]] 581 // CHECK-DAG: store i64 {{4|8}}, i64* [[SADDR3]] 582 583 // CHECK-DAG: [[CBPADDR2:%.+]] = bitcast i8** [[BPADDR2]] to i[[SZ]]* 584 // CHECK-DAG: [[CPADDR2:%.+]] = bitcast i8** [[PADDR2]] to i[[SZ]]* 585 // CHECK-DAG: store i[[SZ]] [[B_CVAL]], i[[SZ]]* [[CBPADDR2]] 586 // CHECK-DAG: store i[[SZ]] [[B_CVAL]], i[[SZ]]* [[CPADDR2]] 587 // CHECK-DAG: store i64 4, i64* [[SADDR2]] 588 589 // CHECK-DAG: [[CBPADDR0:%.+]] = bitcast i8** [[BPADDR0]] to [[S1]]** 590 // CHECK-DAG: [[CPADDR0:%.+]] = bitcast i8** [[PADDR0]] to double** 591 // CHECK-DAG: store [[S1]]* [[THIS:%.+]], [[S1]]** [[CBPADDR0]] 592 // CHECK-DAG: store double* [[A:%.+]], double** [[CPADDR0]] 593 // CHECK-DAG: store i64 %{{.+}}, i64* [[SADDR0]] 594 595 // CHECK-DAG: [[CBPADDR1:%.+]] = bitcast i8** [[BPADDR1]] to [[S1]]** 596 // CHECK-DAG: [[CPADDR1:%.+]] = bitcast i8** [[PADDR1]] to double** 597 // CHECK-DAG: store [[S1]]* [[THIS]], [[S1]]** [[CBPADDR1]] 598 // CHECK-DAG: store double* [[A]], double** [[CPADDR1]] 599 // CHECK-DAG: store i64 8, i64* [[SADDR1]] 600 601 // CHECK: [[ERROR:%.+]] = icmp ne i32 [[RET]], 0 602 // CHECK-NEXT: br i1 [[ERROR]], label %[[FAIL:.+]], label %[[END:[^,]+]] 603 // CHECK: [[FAIL]] 604 // CHECK: call void [[HVT7:@.+]]({{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}) 605 // CHECK-NEXT: br label %[[END]] 606 // CHECK: [[END]] 607 // CHECK-NEXT: br label %[[IFEND:.+]] 608 // CHECK: [[IFELSE]] 609 // CHECK: call void [[HVT7]]({{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}) 610 // CHECK-NEXT: br label %[[IFEND]] 611 612 // CHECK: [[IFEND]] 613 614 // 615 // CHECK: define {{.*}}[[FSTATIC]] 616 // 617 // CHECK: [[IF:%.+]] = icmp sgt i32 {{[^,]+}}, 50 618 // CHECK: br i1 [[IF]], label %[[IFTHEN:[^,]+]], label %[[IFELSE:[^,]+]] 619 // CHECK: [[IFTHEN]] 620 // CHECK-DAG: [[RET:%.+]] = call i32 @__tgt_target_mapper(i64 -1, i8* @{{[^,]+}}, i32 4, i8** [[BPR:%[^,]+]], i8** [[PR:%[^,]+]], i64* getelementptr inbounds ([4 x i64], [4 x i64]* [[SIZET6]], i32 0, i32 0), i64* getelementptr inbounds ([4 x i64], [4 x i64]* [[MAPT6]], i32 0, i32 0), i8** null) 621 // CHECK-DAG: [[BPR]] = getelementptr inbounds [4 x i8*], [4 x i8*]* [[BP:%.+]], i32 0, i32 0 622 // CHECK-DAG: [[PR]] = getelementptr inbounds [4 x i8*], [4 x i8*]* [[P:%.+]], i32 0, i32 0 623 624 // CHECK-DAG: [[BPADDR0:%.+]] = getelementptr inbounds [4 x i8*], [4 x i8*]* [[BP]], i32 0, i32 0 625 // CHECK-DAG: [[PADDR0:%.+]] = getelementptr inbounds [4 x i8*], [4 x i8*]* [[P]], i32 0, i32 0 626 // CHECK-DAG: [[CBPADDR0:%.+]] = bitcast i8** [[BPADDR0]] to i[[SZ]]* 627 // CHECK-DAG: [[CPADDR0:%.+]] = bitcast i8** [[PADDR0]] to i[[SZ]]* 628 // CHECK-DAG: store i[[SZ]] [[VAL0:%[^,]+]], i[[SZ]]* [[CBPADDR0]] 629 // CHECK-DAG: store i[[SZ]] [[VAL0]], i[[SZ]]* [[CPADDR0]] 630 631 // CHECK-DAG: [[BPADDR1:%.+]] = getelementptr inbounds [4 x i8*], [4 x i8*]* [[BP]], i32 0, i32 1 632 // CHECK-DAG: [[PADDR1:%.+]] = getelementptr inbounds [4 x i8*], [4 x i8*]* [[P]], i32 0, i32 1 633 // CHECK-DAG: [[CBPADDR1:%.+]] = bitcast i8** [[BPADDR1]] to i[[SZ]]* 634 // CHECK-DAG: [[CPADDR1:%.+]] = bitcast i8** [[PADDR1]] to i[[SZ]]* 635 // CHECK-DAG: store i[[SZ]] [[VAL1:%[^,]+]], i[[SZ]]* [[CBPADDR1]] 636 // CHECK-DAG: store i[[SZ]] [[VAL1]], i[[SZ]]* [[CPADDR1]] 637 638 // CHECK-DAG: [[BPADDR2:%.+]] = getelementptr inbounds [4 x i8*], [4 x i8*]* [[BP]], i32 0, i32 2 639 // CHECK-DAG: [[PADDR2:%.+]] = getelementptr inbounds [4 x i8*], [4 x i8*]* [[P]], i32 0, i32 2 640 // CHECK-DAG: [[CBPADDR2:%.+]] = bitcast i8** [[BPADDR2]] to i[[SZ]]* 641 // CHECK-DAG: [[CPADDR2:%.+]] = bitcast i8** [[PADDR2]] to i[[SZ]]* 642 // CHECK-DAG: store i[[SZ]] [[VAL2:%[^,]+]], i[[SZ]]* [[CBPADDR2]] 643 // CHECK-DAG: store i[[SZ]] [[VAL2]], i[[SZ]]* [[CPADDR2]] 644 645 // CHECK-DAG: [[BPADDR3:%.+]] = getelementptr inbounds [4 x i8*], [4 x i8*]* [[BP]], i32 0, i32 3 646 // CHECK-DAG: [[PADDR3:%.+]] = getelementptr inbounds [4 x i8*], [4 x i8*]* [[P]], i32 0, i32 3 647 // CHECK-DAG: [[CBPADDR3:%.+]] = bitcast i8** [[BPADDR3]] to [10 x i32]** 648 // CHECK-DAG: [[CPADDR3:%.+]] = bitcast i8** [[PADDR3]] to [10 x i32]** 649 // CHECK-DAG: store [10 x i32]* [[VAL3:%[^,]+]], [10 x i32]** [[CBPADDR3]] 650 // CHECK-DAG: store [10 x i32]* [[VAL3]], [10 x i32]** [[CPADDR3]] 651 652 // CHECK: [[ERROR:%.+]] = icmp ne i32 [[RET]], 0 653 // CHECK-NEXT: br i1 [[ERROR]], label %[[FAIL:.+]], label %[[END:[^,]+]] 654 // CHECK: [[FAIL]] 655 // CHECK: call void [[HVT6:@.+]]({{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}) 656 // CHECK-NEXT: br label %[[END]] 657 // CHECK: [[END]] 658 // CHECK-NEXT: br label %[[IFEND:.+]] 659 // CHECK: [[IFELSE]] 660 // CHECK: call void [[HVT6]]({{[^,]+}}, {{[^,]+}}, {{[^,]+}}, {{[^,]+}}) 661 // CHECK-NEXT: br label %[[IFEND]] 662 663 // CHECK: [[IFEND]] 664 665 // 666 // CHECK: define {{.*}}[[FTEMPLATE]] 667 // 668 // CHECK: [[IF:%.+]] = icmp sgt i32 {{[^,]+}}, 40 669 // CHECK: br i1 [[IF]], label %[[IFTHEN:[^,]+]], label %[[IFELSE:[^,]+]] 670 // CHECK: [[IFTHEN]] 671 // CHECK-DAG: [[RET:%.+]] = call i32 @__tgt_target_mapper(i64 -1, i8* @{{[^,]+}}, i32 3, i8** [[BPR:%[^,]+]], i8** [[PR:%[^,]+]], i64* getelementptr inbounds ([3 x i64], [3 x i64]* [[SIZET5]], i32 0, i32 0), i64* getelementptr inbounds ([3 x i64], [3 x i64]* [[MAPT5]], i32 0, i32 0), i8** null) 672 // CHECK-DAG: [[BPR]] = getelementptr inbounds [3 x i8*], [3 x i8*]* [[BP:%.+]], i32 0, i32 0 673 // CHECK-DAG: [[PR]] = getelementptr inbounds [3 x i8*], [3 x i8*]* [[P:%.+]], i32 0, i32 0 674 675 // CHECK-DAG: [[BPADDR0:%.+]] = getelementptr inbounds [3 x i8*], [3 x i8*]* [[BP]], i32 0, i32 0 676 // CHECK-DAG: [[PADDR0:%.+]] = getelementptr inbounds [3 x i8*], [3 x i8*]* [[P]], i32 0, i32 0 677 // CHECK-DAG: [[CBPADDR0:%.+]] = bitcast i8** [[BPADDR0]] to i[[SZ]]* 678 // CHECK-DAG: [[CPADDR0:%.+]] = bitcast i8** [[PADDR0]] to i[[SZ]]* 679 // CHECK-DAG: store i[[SZ]] [[VAL0:%[^,]+]], i[[SZ]]* [[CBPADDR0]] 680 // CHECK-DAG: store i[[SZ]] [[VAL0]], i[[SZ]]* [[CPADDR0]] 681 682 // CHECK-DAG: [[BPADDR1:%.+]] = getelementptr inbounds [3 x i8*], [3 x i8*]* [[BP]], i32 0, i32 1 683 // CHECK-DAG: [[PADDR1:%.+]] = getelementptr inbounds [3 x i8*], [3 x i8*]* [[P]], i32 0, i32 1 684 // CHECK-DAG: [[CBPADDR1:%.+]] = bitcast i8** [[BPADDR1]] to i[[SZ]]* 685 // CHECK-DAG: [[CPADDR1:%.+]] = bitcast i8** [[PADDR1]] to i[[SZ]]* 686 // CHECK-DAG: store i[[SZ]] [[VAL1:%[^,]+]], i[[SZ]]* [[CBPADDR1]] 687 // CHECK-DAG: store i[[SZ]] [[VAL1]], i[[SZ]]* [[CPADDR1]] 688 689 // CHECK-DAG: [[BPADDR2:%.+]] = getelementptr inbounds [3 x i8*], [3 x i8*]* [[BP]], i32 0, i32 2 690 // CHECK-DAG: [[PADDR2:%.+]] = getelementptr inbounds [3 x i8*], [3 x i8*]* [[P]], i32 0, i32 2 691 // CHECK-DAG: [[CBPADDR2:%.+]] = bitcast i8** [[BPADDR2]] to [10 x i32]** 692 // CHECK-DAG: [[CPADDR2:%.+]] = bitcast i8** [[PADDR2]] to [10 x i32]** 693 // CHECK-DAG: store [10 x i32]* [[VAL2:%[^,]+]], [10 x i32]** [[CBPADDR2]] 694 // CHECK-DAG: store [10 x i32]* [[VAL2]], [10 x i32]** [[CPADDR2]] 695 696 // CHECK: [[ERROR:%.+]] = icmp ne i32 [[RET]], 0 697 // CHECK-NEXT: br i1 [[ERROR]], label %[[FAIL:.+]], label %[[END:[^,]+]] 698 // CHECK: [[FAIL]] 699 // CHECK: call void [[HVT5:@.+]]({{[^,]+}}, {{[^,]+}}, {{[^,]+}}) 700 // CHECK-NEXT: br label %[[END]] 701 // CHECK: [[END]] 702 // CHECK-NEXT: br label %[[IFEND:.+]] 703 // CHECK: [[IFELSE]] 704 // CHECK: call void [[HVT5]]({{[^,]+}}, {{[^,]+}}, {{[^,]+}}) 705 // CHECK-NEXT: br label %[[IFEND]] 706 707 // CHECK: [[IFEND]] 708 709 // OMP45: define internal void @__omp_offloading_{{.+}}_{{.+}}bar{{.+}}_l838(i[[SZ]] %{{.+}}) 710 711 // OMP45: define {{.*}}@{{.*}}zee{{.*}} 712 713 // OMP45: [[LOCAL_THIS:%.+]] = alloca [[S2]]* 714 // OMP45: [[BP:%.+]] = alloca [1 x i8*] 715 // OMP45: [[P:%.+]] = alloca [1 x i8*] 716 // OMP45: [[LOCAL_THIS1:%.+]] = load [[S2]]*, [[S2]]** [[LOCAL_THIS]] 717 718 // OMP45: call void @__kmpc_critical( 719 // OMP45: [[ARR_IDX:%.+]] = getelementptr inbounds [[S2]], [[S2]]* [[LOCAL_THIS1]], i[[SZ]] 0 720 // OMP45: [[ARR_IDX2:%.+]] = getelementptr inbounds [[S2]], [[S2]]* [[LOCAL_THIS1]], i[[SZ]] 0 721 722 // OMP45-DAG: [[BPADDR0:%.+]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[BP]], i32 0, i32 0 723 // OMP45-DAG: [[PADDR0:%.+]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[P]], i32 0, i32 0 724 // OMP45-DAG: [[CBPADDR0:%.+]] = bitcast i8** [[BPADDR0]] to [[S2]]** 725 // OMP45-DAG: [[CPADDR0:%.+]] = bitcast i8** [[PADDR0]] to [[S2]]** 726 // OMP45-DAG: store [[S2]]* [[ARR_IDX]], [[S2]]** [[CBPADDR0]] 727 // OMP45-DAG: store [[S2]]* [[ARR_IDX2]], [[S2]]** [[CPADDR0]] 728 729 // OMP45: [[BPR:%.+]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[BP]], i32 0, i32 0 730 // OMP45: [[PR:%.+]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[P]], i32 0, i32 0 731 // OMP45: [[RET:%.+]] = call i32 @__tgt_target_mapper(i64 -1, i8* @{{[^,]+}}, i32 1, i8** [[BPR:%[^,]+]], i8** [[PR:%[^,]+]], i64* getelementptr inbounds ([1 x i64], [1 x i64]* [[SIZET9]], i32 0, i32 0), i64* getelementptr inbounds ([1 x i64], [1 x i64]* [[MAPT10]], i32 0, i32 0), i8** null) 732 // OMP45-NEXT: [[ERROR:%.+]] = icmp ne i32 [[RET]], 0 733 // OMP45-NEXT: br i1 [[ERROR]], label %[[FAIL:[^,]+]], label %[[END:[^,]+]] 734 // OMP45: [[FAIL]] 735 // OMP45: call void [[HVT0:@.+]]([[S2]]* [[LOCAL_THIS1]]) 736 // OMP45-NEXT: br label %[[END]] 737 // OMP45: [[END]] 738 // OMP45: call void @__kmpc_end_critical( 739 740 // Check that the offloading functions are emitted and that the arguments are 741 // correct and loaded correctly for the target regions of the callees of bar(). 742 743 // CHECK: define internal void [[HVT7]] 744 // Create local storage for each capture. 745 // CHECK: [[LOCAL_THIS:%.+]] = alloca [[S1]]* 746 // CHECK: [[LOCAL_B:%.+]] = alloca i[[SZ]] 747 // CHECK: [[LOCAL_VLA1:%.+]] = alloca i[[SZ]] 748 // CHECK: [[LOCAL_VLA2:%.+]] = alloca i[[SZ]] 749 // CHECK: [[LOCAL_C:%.+]] = alloca i16* 750 // CHECK-DAG: store [[S1]]* [[ARG_THIS:%.+]], [[S1]]** [[LOCAL_THIS]] 751 // CHECK-DAG: store i[[SZ]] [[ARG_B:%.+]], i[[SZ]]* [[LOCAL_B]] 752 // CHECK-DAG: store i[[SZ]] [[ARG_VLA1:%.+]], i[[SZ]]* [[LOCAL_VLA1]] 753 // CHECK-DAG: store i[[SZ]] [[ARG_VLA2:%.+]], i[[SZ]]* [[LOCAL_VLA2]] 754 // CHECK-DAG: store i16* [[ARG_C:%.+]], i16** [[LOCAL_C]] 755 // Store captures in the context. 756 // CHECK-DAG: [[REF_THIS:%.+]] = load [[S1]]*, [[S1]]** [[LOCAL_THIS]], 757 // CHECK-64-DAG:[[REF_B:%.+]] = bitcast i[[SZ]]* [[LOCAL_B]] to i32* 758 // CHECK-DAG: [[VAL_VLA1:%.+]] = load i[[SZ]], i[[SZ]]* [[LOCAL_VLA1]], 759 // CHECK-DAG: [[VAL_VLA2:%.+]] = load i[[SZ]], i[[SZ]]* [[LOCAL_VLA2]], 760 // CHECK-DAG: [[REF_C:%.+]] = load i16*, i16** [[LOCAL_C]], 761 // Use captures. 762 // CHECK-DAG: getelementptr inbounds [[S1]], [[S1]]* [[REF_THIS]], i32 0, i32 0 763 // CHECK-64-DAG:load i32, i32* [[REF_B]] 764 // CHECK-32-DAG:load i32, i32* [[LOCAL_B]] 765 // CHECK-DAG: getelementptr inbounds i16, i16* [[REF_C]], i[[SZ]] %{{.+}} 766 767 768 // CHECK: define internal void [[HVT6]] 769 // Create local storage for each capture. 770 // CHECK: [[LOCAL_A:%.+]] = alloca i[[SZ]] 771 // CHECK: [[LOCAL_AA:%.+]] = alloca i[[SZ]] 772 // CHECK: [[LOCAL_AAA:%.+]] = alloca i[[SZ]] 773 // CHECK: [[LOCAL_B:%.+]] = alloca [10 x i32]* 774 // CHECK-DAG: store i[[SZ]] [[ARG_A:%.+]], i[[SZ]]* [[LOCAL_A]] 775 // CHECK-DAG: store i[[SZ]] [[ARG_AA:%.+]], i[[SZ]]* [[LOCAL_AA]] 776 // CHECK-DAG: store i[[SZ]] [[ARG_AAA:%.+]], i[[SZ]]* [[LOCAL_AAA]] 777 // CHECK-DAG: store [10 x i32]* [[ARG_B:%.+]], [10 x i32]** [[LOCAL_B]] 778 // Store captures in the context. 779 // CHECK-64-DAG: [[REF_A:%.+]] = bitcast i[[SZ]]* [[LOCAL_A]] to i32* 780 // CHECK-DAG: [[REF_AA:%.+]] = bitcast i[[SZ]]* [[LOCAL_AA]] to i16* 781 // CHECK-DAG: [[REF_AAA:%.+]] = bitcast i[[SZ]]* [[LOCAL_AAA]] to i8* 782 // CHECK-DAG: [[REF_B:%.+]] = load [10 x i32]*, [10 x i32]** [[LOCAL_B]], 783 // Use captures. 784 // CHECK-64-DAG: load i32, i32* [[REF_A]] 785 // CHECK-DAG: load i16, i16* [[REF_AA]] 786 // CHECK-DAG: load i8, i8* [[REF_AAA]] 787 // CHECK-32-DAG: load i32, i32* [[LOCAL_A]] 788 // CHECK-DAG: getelementptr inbounds [10 x i32], [10 x i32]* [[REF_B]], i[[SZ]] 0, i[[SZ]] 2 789 790 // CHECK: define internal void [[HVT5]] 791 // Create local storage for each capture. 792 // CHECK: [[LOCAL_A:%.+]] = alloca i[[SZ]] 793 // CHECK: [[LOCAL_AA:%.+]] = alloca i[[SZ]] 794 // CHECK: [[LOCAL_B:%.+]] = alloca [10 x i32]* 795 // CHECK-DAG: store i[[SZ]] [[ARG_A:%.+]], i[[SZ]]* [[LOCAL_A]] 796 // CHECK-DAG: store i[[SZ]] [[ARG_AA:%.+]], i[[SZ]]* [[LOCAL_AA]] 797 // CHECK-DAG: store [10 x i32]* [[ARG_B:%.+]], [10 x i32]** [[LOCAL_B]] 798 // Store captures in the context. 799 // CHECK-64-DAG:[[REF_A:%.+]] = bitcast i[[SZ]]* [[LOCAL_A]] to i32* 800 // CHECK-DAG: [[REF_AA:%.+]] = bitcast i[[SZ]]* [[LOCAL_AA]] to i16* 801 // CHECK-DAG: [[REF_B:%.+]] = load [10 x i32]*, [10 x i32]** [[LOCAL_B]], 802 // Use captures. 803 // CHECK-64-DAG: load i32, i32* [[REF_A]] 804 // CHECK-32-DAG: load i32, i32* [[LOCAL_A]] 805 // CHECK-DAG: load i16, i16* [[REF_AA]] 806 // CHECK-DAG: getelementptr inbounds [10 x i32], [10 x i32]* [[REF_B]], i[[SZ]] 0, i[[SZ]] 2 807 808 // OMP50: define internal void @__omp_offloading_{{.+}}_{{.+}}bar{{.+}}_l838(i[[SZ]] %{{.+}}) 809 810 // OMP50: define {{.*}}@{{.*}}zee{{.*}} 811 812 // OMP50: [[LOCAL_THIS:%.+]] = alloca [[S2]]* 813 // OMP50: [[BP:%.+]] = alloca [1 x i8*] 814 // OMP50: [[P:%.+]] = alloca [1 x i8*] 815 // OMP50: [[LOCAL_THIS1:%.+]] = load [[S2]]*, [[S2]]** [[LOCAL_THIS]] 816 // OMP50: [[ARR_IDX:%.+]] = getelementptr inbounds [[S2]], [[S2]]* [[LOCAL_THIS1]], i[[SZ]] 0 817 // OMP50: [[ARR_IDX2:%.+]] = getelementptr inbounds [[S2]], [[S2]]* [[LOCAL_THIS1]], i[[SZ]] 0 818 819 // OMP50-DAG: [[BPADDR0:%.+]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[BP]], i32 0, i32 0 820 // OMP50-DAG: [[PADDR0:%.+]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[P]], i32 0, i32 0 821 // OMP50-DAG: [[CBPADDR0:%.+]] = bitcast i8** [[BPADDR0]] to [[S2]]** 822 // OMP50-DAG: [[CPADDR0:%.+]] = bitcast i8** [[PADDR0]] to [[S2]]** 823 // OMP50-DAG: store [[S2]]* [[ARR_IDX]], [[S2]]** [[CBPADDR0]] 824 // OMP50-DAG: store [[S2]]* [[ARR_IDX2]], [[S2]]** [[CPADDR0]] 825 826 // OMP50: [[BPR:%.+]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[BP]], i32 0, i32 0 827 // OMP50: [[PR:%.+]] = getelementptr inbounds [1 x i8*], [1 x i8*]* [[P]], i32 0, i32 0 828 // OMP50: [[RET:%.+]] = call i32 @__tgt_target_mapper(i64 -1, i8* @{{[^,]+}}, i32 1, i8** [[BPR:%[^,]+]], i8** [[PR:%[^,]+]], i64* getelementptr inbounds ([1 x i64], [1 x i64]* [[SIZET9]], i32 0, i32 0), i64* getelementptr inbounds ([1 x i64], [1 x i64]* [[MAPT10]], i32 0, i32 0), i8** null) 829 // OMP50-NEXT: [[ERROR:%.+]] = icmp ne i32 [[RET]], 0 830 // OMP50-NEXT: br i1 [[ERROR]], label %[[FAIL:[^,]+]], label %[[END:[^,]+]] 831 // OMP50: [[FAIL]] 832 // OMP50: call void [[HVT0:@.+]]([[S2]]* [[LOCAL_THIS1]]) 833 // OMP50-NEXT: br label %[[END]] 834 // OMP50: [[END]] 835 836 void bar () { 837 #define pragma_target _Pragma("omp target") 838 pragma_target 839 { 840 global = 0; 841 #pragma omp parallel shared(global) 842 global = 1; 843 } 844 } 845 846 class S2 { 847 int a, b, c; 848 849 public: 850 void zee() { 851 #pragma omp critical 852 #pragma omp target map(this[0]) 853 a++; 854 } 855 }; 856 857 // CHECK: define internal void @.omp_offloading.requires_reg() 858 // CHECK: call void @__tgt_register_requires(i64 1) 859 // CHECK: ret void 860 861 int main () { 862 S2 bar; 863 bar.zee(); 864 } 865 866 #endif 867