diff options
Diffstat (limited to 'llvm/test/CodeGen/AMDGPU/load-local-i8.ll')
| -rw-r--r-- | llvm/test/CodeGen/AMDGPU/load-local-i8.ll | 17 |
1 files changed, 17 insertions, 0 deletions
diff --git a/llvm/test/CodeGen/AMDGPU/load-local-i8.ll b/llvm/test/CodeGen/AMDGPU/load-local-i8.ll index 72f5408675f..e1931af042b 100644 --- a/llvm/test/CodeGen/AMDGPU/load-local-i8.ll +++ b/llvm/test/CodeGen/AMDGPU/load-local-i8.ll @@ -3,6 +3,9 @@ ; RUN: llc -march=amdgcn -mtriple=amdgcn---amdgiz -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -check-prefixes=GCN,GFX9,FUNC %s ; RUN: llc -march=r600 -mtriple=r600---amdgiz -mcpu=redwood -verify-machineinstrs < %s | FileCheck -check-prefix=EG -check-prefix=FUNC %s +; Testing for ds_read_b128 +; RUN: llc -march=amdgcn -mcpu=tonga -amdgpu-ds128 < %s | FileCheck -check-prefixes=CIVI,FUNC %s +; RUN: llc -march=amdgcn -mcpu=gfx900 -amdgpu-ds128 < %s | FileCheck -check-prefixes=CIVI,FUNC %s ; FUNC-LABEL: {{^}}local_load_i8: ; GCN-NOT: s_wqm_b64 @@ -1021,4 +1024,18 @@ define amdgpu_kernel void @local_sextload_v32i8_to_v32i16(<32 x i16> addrspace(3 ; ret void ; } +; Tests if ds_read_b128 gets generated for the 16 byte aligned load. +; FUNC-LABEL: {{^}}local_v16i8_to_128: +; SI-NOT: ds_read_b128 +; CIVI: ds_read_b128 +; EG: LDS_READ_RET +; EG: LDS_READ_RET +; EG: LDS_READ_RET +; EG: LDS_READ_RET +define amdgpu_kernel void @local_v16i8_to_128(<16 x i8> addrspace(3)* %out, <16 x i8> addrspace(3)* %in) { + %ld = load <16 x i8>, <16 x i8> addrspace(3)* %in, align 16 + store <16 x i8> %ld, <16 x i8> addrspace(3)* %out + ret void +} + attributes #0 = { nounwind } |
