diff options
| author | Alexey Bataev <a.bataev@hotmail.com> | 2018-11-09 16:18:04 +0000 |
|---|---|---|
| committer | Alexey Bataev <a.bataev@hotmail.com> | 2018-11-09 16:18:04 +0000 |
| commit | 223835dbb1f8b240a8262100dbef25b02e6558c7 (patch) | |
| tree | 414efddea92fac35a901bf8a165cf94060e4fff3 /test/OpenMP/nvptx_target_teams_distribute_parallel_for_simd_codegen.cpp | |
| parent | 9096f2a5c68ea935cb3d1b08ab36a93c4d729cbd (diff) | |
| download | clang-223835dbb1f8b240a8262100dbef25b02e6558c7.tar.gz | |
[OPENMP][NVPTX]Allow to use shared memory for the
target|teams|distribute variables.
If the total size of the variables, declared in target|teams|distribute
regions, is less than the maximal size of shared memory available, the
buffer is allocated in the shared memory.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@346507 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'test/OpenMP/nvptx_target_teams_distribute_parallel_for_simd_codegen.cpp')
| -rw-r--r-- | test/OpenMP/nvptx_target_teams_distribute_parallel_for_simd_codegen.cpp | 9 |
1 files changed, 4 insertions, 5 deletions
diff --git a/test/OpenMP/nvptx_target_teams_distribute_parallel_for_simd_codegen.cpp b/test/OpenMP/nvptx_target_teams_distribute_parallel_for_simd_codegen.cpp index f7a3aaecc3..3b56bfd9a7 100644 --- a/test/OpenMP/nvptx_target_teams_distribute_parallel_for_simd_codegen.cpp +++ b/test/OpenMP/nvptx_target_teams_distribute_parallel_for_simd_codegen.cpp @@ -63,23 +63,22 @@ int bar(int n){ } // CHECK-DAG: [[MEM_TY:%.+]] = type { [4 x i8] } -// CHECK-DAG: [[GLOBAL_RD:@.+]] = weak global [{{[0-9]+}} x [{{[0-9]+}} x [[MEM_TY]]]] zeroinitializer -// CHECK-DAG: [[GLOBAL_RD_PTR:@.+]] = weak unnamed_addr constant i8* getelementptr inbounds ([{{[0-9]+}} x [{{[0-9]+}} x [[MEM_TY]]]], [{{[0-9]+}} x [{{[0-9]+}} x [[MEM_TY]]]]* [[GLOBAL_RD]], i{{[0-9]+}} 0, i{{[0-9]+}} 0, i{{[0-9]+}} 0, i{{[0-9]+}} 0, i{{[0-9]+}} 0) +// CHECK-DAG: [[SHARED_GLOBAL_RD:@.+]] = weak addrspace(3) global [[MEM_TY]] zeroinitializer // CHECK-DAG: [[KERNEL_PTR:@.+]] = internal addrspace(3) global i8* null // CHECK-DAG: [[KERNEL_SIZE:@.+]] = internal unnamed_addr constant i{{64|32}} 4 +// CHECK-DAG: [[KERNEL_SHARED:@.+]] = internal unnamed_addr constant i16 1 // CHECK-LABEL: define {{.*}}void {{@__omp_offloading_.+}}_l30( // CHECK-DAG: [[THREAD_LIMIT:%.+]] = call i32 @llvm.nvvm.read.ptx.sreg.ntid.x() // CHECK: call void @__kmpc_spmd_kernel_init(i32 [[THREAD_LIMIT]], i16 0, i16 1) -// CHECK: [[GLOBAL_RD:%.+]] = load i8*, i8** [[GLOBAL_RD_PTR]], -// CHECK: call void @__kmpc_get_team_static_memory(i8* [[GLOBAL_RD]], i{{64|32}} 4, i16 0, i8** addrspacecast (i8* addrspace(3)* [[KERNEL_PTR]] to i8**)) +// CHECK: call void @__kmpc_get_team_static_memory(i8* addrspacecast (i8 addrspace(3)* getelementptr inbounds ([[MEM_TY]], [[MEM_TY]] addrspace(3)* [[SHARED_GLOBAL_RD]], i32 0, i32 0, i32 0) to i8*), i{{64|32}} 4, i16 1, i8** addrspacecast (i8* addrspace(3)* [[KERNEL_PTR]] to i8**)) // CHECK: [[TEAM_ALLOC:%.+]] = load i8*, i8* addrspace(3)* [[KERNEL_PTR]], // CHECK: [[BC:%.+]] = bitcast i8* [[TEAM_ALLOC]] to [[REC:%.+]]* // CHECK: getelementptr inbounds [[REC]], [[REC]]* [[BC]], i{{[0-9]+}} 0, i{{[0-9]+}} 0 // CHECK: call void @__kmpc_for_static_init_4({{.+}}, {{.+}}, {{.+}} 91, // CHECK: {{call|invoke}} void [[OUTL1:@.+]]( // CHECK: call void @__kmpc_for_static_fini( -// CHECK: call void @__kmpc_restore_team_static_memory(i16 0) +// CHECK: call void @__kmpc_restore_team_static_memory(i16 1) // CHECK: call void @__kmpc_spmd_kernel_deinit() // CHECK: ret void |