xref: /llvm-project/llvm/test/CodeGen/AMDGPU/fcopysign.f32.r600.ll (revision 9e9907f1cfa424366fba58d9520f9305b537cec9)
1; RUN: llc -mtriple=r600 -mcpu=cypress < %s | FileCheck -check-prefix=EG %s
2
3declare float @llvm.copysign.f32(float, float) nounwind readnone
4declare <2 x float> @llvm.copysign.v2f32(<2 x float>, <2 x float>) nounwind readnone
5declare <4 x float> @llvm.copysign.v4f32(<4 x float>, <4 x float>) nounwind readnone
6
7; EG: BFI_INT
8define amdgpu_kernel void @test_copysign_f32(ptr addrspace(1) %out, float %mag, float %sign) nounwind {
9  %result = call float @llvm.copysign.f32(float %mag, float %sign)
10  store float %result, ptr addrspace(1) %out, align 4
11  ret void
12}
13
14; EG: BFI_INT
15; EG: BFI_INT
16define amdgpu_kernel void @test_copysign_v2f32(ptr addrspace(1) %out, <2 x float> %mag, <2 x float> %sign) nounwind {
17  %result = call <2 x float> @llvm.copysign.v2f32(<2 x float> %mag, <2 x float> %sign)
18  store <2 x float> %result, ptr addrspace(1) %out, align 8
19  ret void
20}
21
22; EG: BFI_INT
23; EG: BFI_INT
24; EG: BFI_INT
25; EG: BFI_INT
26define amdgpu_kernel void @test_copysign_v4f32(ptr addrspace(1) %out, <4 x float> %mag, <4 x float> %sign) nounwind {
27  %result = call <4 x float> @llvm.copysign.v4f32(<4 x float> %mag, <4 x float> %sign)
28  store <4 x float> %result, ptr addrspace(1) %out, align 16
29  ret void
30}
31