Index: clang/lib/Headers/openmp_wrappers/new =================================================================== --- clang/lib/Headers/openmp_wrappers/new +++ clang/lib/Headers/openmp_wrappers/new @@ -13,7 +13,7 @@ // which do not use nothrow_t are provided without the header. #include_next -#if defined(__NVPTX__) && defined(_OPENMP) +#if (defined(__NVPTX__) || defined(__AMDGPU__)) && defined(_OPENMP) #include Index: clang/test/Headers/openmp_new_nothrow.cpp =================================================================== --- clang/test/Headers/openmp_new_nothrow.cpp +++ clang/test/Headers/openmp_new_nothrow.cpp @@ -1,6 +1,9 @@ // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 2 -// RUN: %clang_cc1 -std=c++03 -internal-isystem %S/../../lib/Headers/openmp_wrappers -include __clang_openmp_device_functions.h -internal-isystem %S/Inputs/include -verify -fopenmp -x c++ -triple nvptx64-nvidia-cuda -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm -fopenmp-is-device -o - %s | FileCheck -check-prefixes=CHECK,CXX03 %s -// RUN: %clang_cc1 -std=c++11 -internal-isystem %S/../../lib/Headers/openmp_wrappers -include __clang_openmp_device_functions.h -internal-isystem %S/Inputs/include -verify -fopenmp -x c++ -triple nvptx64-nvidia-cuda -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm -fopenmp-is-device -o - %s | FileCheck -check-prefixes=CHECK,CXX11 %s +// RUN: %clang_cc1 -std=c++03 -internal-isystem %S/../../lib/Headers/openmp_wrappers -include __clang_openmp_device_functions.h -internal-isystem %S/Inputs/include -verify -fopenmp -x c++ -triple nvptx64-nvidia-cuda -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm -fopenmp-is-device -o - %s | FileCheck -check-prefixes=NVPTX,NVPTX-CXX03 %s +// RUN: %clang_cc1 -std=c++11 -internal-isystem %S/../../lib/Headers/openmp_wrappers -include __clang_openmp_device_functions.h -internal-isystem %S/Inputs/include -verify -fopenmp -x c++ -triple nvptx64-nvidia-cuda -fopenmp-targets=nvptx64-nvidia-cuda -emit-llvm -fopenmp-is-device -o - %s | FileCheck -check-prefixes=NVPTX,NVPTX-NVPTX-CXX11 %s + +// RUN: %clang_cc1 -std=c++03 -internal-isystem %S/../../lib/Headers/openmp_wrappers -include __clang_openmp_device_functions.h -internal-isystem %S/Inputs/include -verify -fopenmp -x c++ -triple amdgcn-amd-amdhsa -fopenmp-targets=amdgcn-amd-amdhsa -emit-llvm -fopenmp-is-device -o - %s | FileCheck -check-prefixes=AMDGPU,AMDGPU-CXX03 %s +// RUN: %clang_cc1 -std=c++11 -internal-isystem %S/../../lib/Headers/openmp_wrappers -include __clang_openmp_device_functions.h -internal-isystem %S/Inputs/include -verify -fopenmp -x c++ -triple amdgcn-amd-amdhsa -fopenmp-targets=amdgcn-amd-amdhsa -emit-llvm -fopenmp-is-device -o - %s | FileCheck -check-prefixes=AMDGPU,AMDGPU-CXX11 %s // expected-no-diagnostics @@ -10,71 +13,159 @@ #pragma omp begin declare target extern const std::nothrow_t nothrow; -// CHECK-LABEL: define hidden noundef ptr @_Z17new_stuff_nothrowv -// CHECK-SAME: () #[[ATTR0:[0-9]+]] { -// CHECK-NEXT: entry: -// CHECK-NEXT: [[CALL:%.*]] = call noalias noundef ptr @_ZnwmRKSt9nothrow_t(i64 noundef 4, ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR4:[0-9]+]] -// CHECK-NEXT: ret ptr [[CALL]] +// NVPTX-LABEL: define hidden noundef ptr @_Z17new_stuff_nothrowv +// NVPTX-SAME: () #[[ATTR0:[0-9]+]] { +// NVPTX-NEXT: entry: +// NVPTX-NEXT: [[CALL:%.*]] = call noalias noundef ptr @_ZnwmRKSt9nothrow_t(i64 noundef 4, ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR4:[0-9]+]] +// NVPTX-NEXT: ret ptr [[CALL]] +// +// AMDGPU-LABEL: define hidden noundef ptr @_Z17new_stuff_nothrowv +// AMDGPU-SAME: () #[[ATTR0:[0-9]+]] { +// AMDGPU-NEXT: entry: +// AMDGPU-NEXT: [[RETVAL:%.*]] = alloca ptr, align 8, addrspace(5) +// AMDGPU-NEXT: [[RETVAL_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[RETVAL]] to ptr +// AMDGPU-NEXT: [[CALL:%.*]] = call noalias noundef ptr @_ZnwmRKSt9nothrow_t(i64 noundef 4, ptr noundef nonnull align 1 dereferenceable(1) addrspacecast (ptr addrspace(1) @nothrow to ptr)) #[[ATTR4:[0-9]+]] +// AMDGPU-NEXT: ret ptr [[CALL]] // int* new_stuff_nothrow() { return new (nothrow) int; } -// CXX03-LABEL: define hidden noundef ptr @_Z23new_array_stuff_nothrowv -// CXX03-SAME: () #[[ATTR0]] { -// CXX03-NEXT: entry: -// CXX03-NEXT: [[CALL:%.*]] = call noalias noundef ptr @_ZnamRKSt9nothrow_t(i64 noundef 136, ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR4]] -// CXX03-NEXT: ret ptr [[CALL]] -// -// CXX11-LABEL: define hidden noundef ptr @_Z23new_array_stuff_nothrowv -// CXX11-SAME: () #[[ATTR0]] { -// CXX11-NEXT: entry: -// CXX11-NEXT: [[CALL:%.*]] = call noalias noundef ptr @_ZnamRKSt9nothrow_t(i64 noundef 136, ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR7:[0-9]+]] -// CXX11-NEXT: ret ptr [[CALL]] +// +// NVPTX-CXX11-LABEL: define hidden noundef ptr @_Z23new_array_stuff_nothrowv +// NVPTX-CXX11-SAME: () #[[ATTR0]] { +// NVPTX-CXX11-NEXT: entry: +// NVPTX-CXX11-NEXT: [[CALL:%.*]] = call noalias noundef ptr @_ZnamRKSt9nothrow_t(i64 noundef 136, ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR7:[0-9]+]] +// NVPTX-CXX11-NEXT: ret ptr [[CALL]] +// NVPTX-CXX03-LABEL: define hidden noundef ptr @_Z23new_array_stuff_nothrowv +// NVPTX-CXX03-SAME: () #[[ATTR0]] { +// NVPTX-CXX03-NEXT: entry: +// NVPTX-CXX03-NEXT: [[CALL:%.*]] = call noalias noundef ptr @_ZnamRKSt9nothrow_t(i64 noundef 136, ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR4]] +// NVPTX-CXX03-NEXT: ret ptr [[CALL]] +// +// NVPTX-NVPTX-CXX11-LABEL: define hidden noundef ptr @_Z23new_array_stuff_nothrowv +// NVPTX-NVPTX-CXX11-SAME: () #[[ATTR0]] { +// NVPTX-NVPTX-CXX11-NEXT: entry: +// NVPTX-NVPTX-CXX11-NEXT: [[CALL:%.*]] = call noalias noundef ptr @_ZnamRKSt9nothrow_t(i64 noundef 136, ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR7:[0-9]+]] +// NVPTX-NVPTX-CXX11-NEXT: ret ptr [[CALL]] +// +// AMDGPU-CXX03-LABEL: define hidden noundef ptr @_Z23new_array_stuff_nothrowv +// AMDGPU-CXX03-SAME: () #[[ATTR0]] { +// AMDGPU-CXX03-NEXT: entry: +// AMDGPU-CXX03-NEXT: [[RETVAL:%.*]] = alloca ptr, align 8, addrspace(5) +// AMDGPU-CXX03-NEXT: [[RETVAL_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[RETVAL]] to ptr +// AMDGPU-CXX03-NEXT: [[CALL:%.*]] = call noalias noundef ptr @_ZnamRKSt9nothrow_t(i64 noundef 136, ptr noundef nonnull align 1 dereferenceable(1) addrspacecast (ptr addrspace(1) @nothrow to ptr)) #[[ATTR4]] +// AMDGPU-CXX03-NEXT: ret ptr [[CALL]] +// +// AMDGPU-CXX11-LABEL: define hidden noundef ptr @_Z23new_array_stuff_nothrowv +// AMDGPU-CXX11-SAME: () #[[ATTR0]] { +// AMDGPU-CXX11-NEXT: entry: +// AMDGPU-CXX11-NEXT: [[RETVAL:%.*]] = alloca ptr, align 8, addrspace(5) +// AMDGPU-CXX11-NEXT: [[RETVAL_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[RETVAL]] to ptr +// AMDGPU-CXX11-NEXT: [[CALL:%.*]] = call noalias noundef ptr @_ZnamRKSt9nothrow_t(i64 noundef 136, ptr noundef nonnull align 1 dereferenceable(1) addrspacecast (ptr addrspace(1) @nothrow to ptr)) #[[ATTR7:[0-9]+]] +// AMDGPU-CXX11-NEXT: ret ptr [[CALL]] // int* new_array_stuff_nothrow() { return new (nothrow) int[34]; } -// CXX03-LABEL: define hidden void @_Z20delete_stuff_nothrowPi -// CXX03-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { -// CXX03-NEXT: entry: -// CXX03-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8 -// CXX03-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR]], align 8 -// CXX03-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR]], align 8 -// CXX03-NEXT: call void @_ZdlPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR6:[0-9]+]] -// CXX03-NEXT: ret void -// -// CXX11-LABEL: define hidden void @_Z20delete_stuff_nothrowPi -// CXX11-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { -// CXX11-NEXT: entry: -// CXX11-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8 -// CXX11-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR]], align 8 -// CXX11-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR]], align 8 -// CXX11-NEXT: call void @_ZdlPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR8:[0-9]+]] -// CXX11-NEXT: ret void +// +// NVPTX-CXX11-LABEL: define hidden void @_Z20delete_stuff_nothrowPi +// NVPTX-CXX11-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { +// NVPTX-CXX11-NEXT: entry: +// NVPTX-CXX11-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8 +// NVPTX-CXX11-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR]], align 8 +// NVPTX-CXX11-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR]], align 8 +// NVPTX-CXX11-NEXT: call void @_ZdlPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR8:[0-9]+]] +// NVPTX-CXX11-NEXT: ret void +// NVPTX-CXX03-LABEL: define hidden void @_Z20delete_stuff_nothrowPi +// NVPTX-CXX03-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { +// NVPTX-CXX03-NEXT: entry: +// NVPTX-CXX03-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8 +// NVPTX-CXX03-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR]], align 8 +// NVPTX-CXX03-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR]], align 8 +// NVPTX-CXX03-NEXT: call void @_ZdlPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR6:[0-9]+]] +// NVPTX-CXX03-NEXT: ret void +// +// NVPTX-NVPTX-CXX11-LABEL: define hidden void @_Z20delete_stuff_nothrowPi +// NVPTX-NVPTX-CXX11-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { +// NVPTX-NVPTX-CXX11-NEXT: entry: +// NVPTX-NVPTX-CXX11-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8 +// NVPTX-NVPTX-CXX11-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR]], align 8 +// NVPTX-NVPTX-CXX11-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR]], align 8 +// NVPTX-NVPTX-CXX11-NEXT: call void @_ZdlPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR8:[0-9]+]] +// NVPTX-NVPTX-CXX11-NEXT: ret void +// +// AMDGPU-CXX03-LABEL: define hidden void @_Z20delete_stuff_nothrowPi +// AMDGPU-CXX03-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { +// AMDGPU-CXX03-NEXT: entry: +// AMDGPU-CXX03-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8, addrspace(5) +// AMDGPU-CXX03-NEXT: [[PTR_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[PTR_ADDR]] to ptr +// AMDGPU-CXX03-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR_ASCAST]], align 8 +// AMDGPU-CXX03-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR_ASCAST]], align 8 +// AMDGPU-CXX03-NEXT: call void @_ZdlPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) addrspacecast (ptr addrspace(1) @nothrow to ptr)) #[[ATTR6:[0-9]+]] +// AMDGPU-CXX03-NEXT: ret void +// +// AMDGPU-CXX11-LABEL: define hidden void @_Z20delete_stuff_nothrowPi +// AMDGPU-CXX11-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { +// AMDGPU-CXX11-NEXT: entry: +// AMDGPU-CXX11-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8, addrspace(5) +// AMDGPU-CXX11-NEXT: [[PTR_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[PTR_ADDR]] to ptr +// AMDGPU-CXX11-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR_ASCAST]], align 8 +// AMDGPU-CXX11-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR_ASCAST]], align 8 +// AMDGPU-CXX11-NEXT: call void @_ZdlPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) addrspacecast (ptr addrspace(1) @nothrow to ptr)) #[[ATTR8:[0-9]+]] +// AMDGPU-CXX11-NEXT: ret void // void delete_stuff_nothrow(int* ptr) { operator delete(ptr, nothrow); } -// CXX03-LABEL: define hidden void @_Z26delete_array_stuff_nothrowPi -// CXX03-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { -// CXX03-NEXT: entry: -// CXX03-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8 -// CXX03-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR]], align 8 -// CXX03-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR]], align 8 -// CXX03-NEXT: call void @_ZdaPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR6]] -// CXX03-NEXT: ret void -// -// CXX11-LABEL: define hidden void @_Z26delete_array_stuff_nothrowPi -// CXX11-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { -// CXX11-NEXT: entry: -// CXX11-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8 -// CXX11-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR]], align 8 -// CXX11-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR]], align 8 -// CXX11-NEXT: call void @_ZdaPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR8]] -// CXX11-NEXT: ret void +// +// NVPTX-CXX11-LABEL: define hidden void @_Z26delete_array_stuff_nothrowPi +// NVPTX-CXX11-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { +// NVPTX-CXX11-NEXT: entry: +// NVPTX-CXX11-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8 +// NVPTX-CXX11-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR]], align 8 +// NVPTX-CXX11-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR]], align 8 +// NVPTX-CXX11-NEXT: call void @_ZdaPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR8]] +// NVPTX-CXX11-NEXT: ret void +// NVPTX-CXX03-LABEL: define hidden void @_Z26delete_array_stuff_nothrowPi +// NVPTX-CXX03-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { +// NVPTX-CXX03-NEXT: entry: +// NVPTX-CXX03-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8 +// NVPTX-CXX03-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR]], align 8 +// NVPTX-CXX03-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR]], align 8 +// NVPTX-CXX03-NEXT: call void @_ZdaPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR6]] +// NVPTX-CXX03-NEXT: ret void +// +// NVPTX-NVPTX-CXX11-LABEL: define hidden void @_Z26delete_array_stuff_nothrowPi +// NVPTX-NVPTX-CXX11-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { +// NVPTX-NVPTX-CXX11-NEXT: entry: +// NVPTX-NVPTX-CXX11-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8 +// NVPTX-NVPTX-CXX11-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR]], align 8 +// NVPTX-NVPTX-CXX11-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR]], align 8 +// NVPTX-NVPTX-CXX11-NEXT: call void @_ZdaPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) @nothrow) #[[ATTR8]] +// NVPTX-NVPTX-CXX11-NEXT: ret void +// +// AMDGPU-CXX03-LABEL: define hidden void @_Z26delete_array_stuff_nothrowPi +// AMDGPU-CXX03-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { +// AMDGPU-CXX03-NEXT: entry: +// AMDGPU-CXX03-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8, addrspace(5) +// AMDGPU-CXX03-NEXT: [[PTR_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[PTR_ADDR]] to ptr +// AMDGPU-CXX03-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR_ASCAST]], align 8 +// AMDGPU-CXX03-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR_ASCAST]], align 8 +// AMDGPU-CXX03-NEXT: call void @_ZdaPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) addrspacecast (ptr addrspace(1) @nothrow to ptr)) #[[ATTR6]] +// AMDGPU-CXX03-NEXT: ret void +// +// AMDGPU-CXX11-LABEL: define hidden void @_Z26delete_array_stuff_nothrowPi +// AMDGPU-CXX11-SAME: (ptr noundef [[PTR:%.*]]) #[[ATTR0]] { +// AMDGPU-CXX11-NEXT: entry: +// AMDGPU-CXX11-NEXT: [[PTR_ADDR:%.*]] = alloca ptr, align 8, addrspace(5) +// AMDGPU-CXX11-NEXT: [[PTR_ADDR_ASCAST:%.*]] = addrspacecast ptr addrspace(5) [[PTR_ADDR]] to ptr +// AMDGPU-CXX11-NEXT: store ptr [[PTR]], ptr [[PTR_ADDR_ASCAST]], align 8 +// AMDGPU-CXX11-NEXT: [[TMP0:%.*]] = load ptr, ptr [[PTR_ADDR_ASCAST]], align 8 +// AMDGPU-CXX11-NEXT: call void @_ZdaPvRKSt9nothrow_t(ptr noundef [[TMP0]], ptr noundef nonnull align 1 dereferenceable(1) addrspacecast (ptr addrspace(1) @nothrow to ptr)) #[[ATTR8]] +// AMDGPU-CXX11-NEXT: ret void // void delete_array_stuff_nothrow(int* ptr) { operator delete[](ptr, nothrow);