diff options
Diffstat (limited to 'test/CodeGen/AMDGPU/llvm.cos.f16.ll')
-rw-r--r-- | test/CodeGen/AMDGPU/llvm.cos.f16.ll | 12 |
1 files changed, 6 insertions, 6 deletions
diff --git a/test/CodeGen/AMDGPU/llvm.cos.f16.ll b/test/CodeGen/AMDGPU/llvm.cos.f16.ll index d836ea36ef63..8931de63e74b 100644 --- a/test/CodeGen/AMDGPU/llvm.cos.f16.ll +++ b/test/CodeGen/AMDGPU/llvm.cos.f16.ll @@ -1,5 +1,5 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s -; RUN: llc -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=SI %s +; RUN: llc -amdgpu-scalarize-global-loads=false -march=amdgcn -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN -check-prefix=VI %s declare half @llvm.cos.f16(half %a) declare <2 x half> @llvm.cos.v2f16(<2 x half> %a) @@ -29,8 +29,8 @@ entry: ; SI: v_lshrrev_b32_e32 v[[A_F16_1:[0-9]+]], 16, v[[A_V2_F16]] ; SI: v_cvt_f32_f16_e32 v[[A_F32_1:[0-9]+]], v[[A_F16_1]] ; SI: v_cvt_f32_f16_e32 v[[A_F32_0:[0-9]+]], v[[A_V2_F16]] -; SI-DAG: v_mul_f32_e32 v[[M_F32_0:[0-9]+]], v[[HALF_PIE]], v[[A_F32_0]] -; SI-DAG: v_mul_f32_e32 v[[M_F32_1:[0-9]+]], v[[HALF_PIE]], v[[A_F32_1]] +; SI-DAG: v_mul_f32_e32 v[[M_F32_0:[0-9]+]], v[[A_F32_0]], v[[HALF_PIE]] +; SI-DAG: v_mul_f32_e32 v[[M_F32_1:[0-9]+]], v[[A_F32_1]], v[[HALF_PIE]] ; VI-DAG: v_cvt_f32_f16_e32 v[[A_F32_0:[0-9]+]], v[[A_V2_F16]] ; VI-DAG: v_cvt_f32_f16_sdwa v[[A_F32_1:[0-9]+]], v[[A_V2_F16]] dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_1 @@ -48,8 +48,8 @@ entry: ; GCN-NOT: and ; SI: v_lshlrev_b32_e32 v[[R_F16_HI:[0-9]+]], 16, v[[R_F16_1]] -; SI: v_or_b32_e32 v[[R_V2_F16:[0-9]+]], v[[R_F16_HI]], v[[R_F16_0]] -; VI: v_or_b32_e32 v[[R_V2_F16:[0-9]+]], v[[R_F16_1]], v[[R_F16_0]] +; SI: v_or_b32_e32 v[[R_V2_F16:[0-9]+]], v[[R_F16_0]], v[[R_F16_HI]] +; VI: v_or_b32_e32 v[[R_V2_F16:[0-9]+]], v[[R_F16_0]], v[[R_F16_1]] ; GCN: buffer_store_dword v[[R_V2_F16]] ; GCN: s_endpgm define amdgpu_kernel void @cos_v2f16( |