diff options
Diffstat (limited to 'llvm/test/CodeGen/AMDGPU/fmed3-cast-combine.ll')
-rw-r--r-- | llvm/test/CodeGen/AMDGPU/fmed3-cast-combine.ll | 6 |
1 files changed, 2 insertions, 4 deletions
diff --git a/llvm/test/CodeGen/AMDGPU/fmed3-cast-combine.ll b/llvm/test/CodeGen/AMDGPU/fmed3-cast-combine.ll index d35871e..99b163d 100644 --- a/llvm/test/CodeGen/AMDGPU/fmed3-cast-combine.ll +++ b/llvm/test/CodeGen/AMDGPU/fmed3-cast-combine.ll @@ -790,8 +790,7 @@ define bfloat @fmed3_f32_fpext_f16_fptrunc_bf16(half %arg0, half %arg1, half %ar ; GFX8-NEXT: v_bfe_u32 v1, v0, 16, 1 ; GFX8-NEXT: v_add_u32_e32 v1, vcc, v1, v0 ; GFX8-NEXT: v_add_u32_e32 v1, vcc, 0x7fff, v1 -; GFX8-NEXT: v_and_b32_e32 v2, 0xff800000, v0 -; GFX8-NEXT: v_or_b32_e32 v2, 0x400000, v2 +; GFX8-NEXT: v_or_b32_e32 v2, 0x400000, v0 ; GFX8-NEXT: v_cmp_u_f32_e32 vcc, v0, v0 ; GFX8-NEXT: v_cndmask_b32_e32 v0, v1, v2, vcc ; GFX8-NEXT: v_lshrrev_b32_e32 v0, 16, v0 @@ -806,9 +805,8 @@ define bfloat @fmed3_f32_fpext_f16_fptrunc_bf16(half %arg0, half %arg1, half %ar ; GFX9-NEXT: s_movk_i32 s4, 0x7fff ; GFX9-NEXT: v_med3_f32 v0, v0, v1, v2 ; GFX9-NEXT: v_bfe_u32 v1, v0, 16, 1 -; GFX9-NEXT: v_and_b32_e32 v2, 0xff800000, v0 ; GFX9-NEXT: v_add3_u32 v1, v1, v0, s4 -; GFX9-NEXT: v_or_b32_e32 v2, 0x400000, v2 +; GFX9-NEXT: v_or_b32_e32 v2, 0x400000, v0 ; GFX9-NEXT: v_cmp_u_f32_e32 vcc, v0, v0 ; GFX9-NEXT: v_cndmask_b32_e32 v0, v1, v2, vcc ; GFX9-NEXT: v_lshrrev_b32_e32 v0, 16, v0 |