diff options
Diffstat (limited to 'test/CodeGen/AMDGPU/llvm.rint.f16.ll')
| -rw-r--r-- | test/CodeGen/AMDGPU/llvm.rint.f16.ll | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/test/CodeGen/AMDGPU/llvm.rint.f16.ll b/test/CodeGen/AMDGPU/llvm.rint.f16.ll index 59e81a7acc0b..30cb969a76e5 100644 --- a/test/CodeGen/AMDGPU/llvm.rint.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.rint.f16.ll @@ -42,13 +42,13 @@ entry: ; VI: v_or_b32_e32 v[[R_V2_F16:[0-9]+]], v[[R_F16_1]], v[[R_F16_0]] ; GFX9: v_rndne_f16_e32 v[[R_F16_0:[0-9]+]], v[[A_V2_F16]] -; GFX9: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]] -; GFX9: v_rndne_f16_e32 v[[R_F16_1:[0-9]+]], v[[A_F16_1]] +; GFX9: v_rndne_f16_sdwa v[[R_F16_1:[0-9]+]], v[[A_V2_F16]] dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 ; GFX9: v_and_b32_e32 v[[R_F16_LO:[0-9]+]], 0xffff, v[[R_F16_0]] ; GFX9: v_lshl_or_b32 v[[R_V2_F16:[0-9]+]], v[[R_F16_1]], 16, v[[R_F16_LO]] ; GCN: buffer_store_dword v[[R_V2_F16]] ; GCN: s_endpgm + define amdgpu_kernel void @rint_v2f16( <2 x half> addrspace(1)* %r, <2 x half> addrspace(1)* %a) { |