diff options
Diffstat (limited to 'llvm/test/CodeGen/AMDGPU/split-smrd.ll')
| -rw-r--r-- | llvm/test/CodeGen/AMDGPU/split-smrd.ll | 22 |
1 files changed, 9 insertions, 13 deletions
diff --git a/llvm/test/CodeGen/AMDGPU/split-smrd.ll b/llvm/test/CodeGen/AMDGPU/split-smrd.ll index d07da103093..66631b03c89 100644 --- a/llvm/test/CodeGen/AMDGPU/split-smrd.ll +++ b/llvm/test/CodeGen/AMDGPU/split-smrd.ll @@ -1,11 +1,11 @@ -; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck %s +; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s ; FIXME: Move this to sgpr-copy.ll when this is fixed on VI. ; Make sure that when we split an smrd instruction in order to move it to ; the VALU, we are also moving its users to the VALU. -; CHECK-LABEL: {{^}}split_smrd_add_worklist: -; CHECK: image_sample v{{[0-9]+}}, v[{{[0-9]+:[0-9]+}}], s[{{[0-9]+:[0-9]+}}], s[{{[0-9]+:[0-9]+}}] dmask:0x1 +; GCN-LABEL: {{^}}split_smrd_add_worklist: +; GCN: image_sample v{{[0-9]+}}, v[{{[0-9]+:[0-9]+}}], s[{{[0-9]+:[0-9]+}}], s[{{[0-9]+:[0-9]+}}] dmask:0x1 define amdgpu_ps void @split_smrd_add_worklist([34 x <8 x i32>] addrspace(2)* byval %arg) #0 { bb: %tmp = call float @llvm.SI.load.const(<16 x i8> undef, i32 96) @@ -24,24 +24,20 @@ bb3: ; preds = %bb %tmp9 = call <4 x float> @llvm.SI.image.sample.v2i32(<2 x i32> <i32 1061158912, i32 1048576000>, <8 x i32> %tmp8, <4 x i32> undef, i32 15, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0) %tmp10 = extractelement <4 x float> %tmp9, i32 0 %tmp12 = call i32 @llvm.SI.packf16(float %tmp10, float undef) - %tmp13 = bitcast i32 %tmp12 to float - call void @llvm.SI.export(i32 15, i32 1, i32 1, i32 0, i32 1, float undef, float %tmp13, float undef, float undef) + %tmp13 = bitcast i32 %tmp12 to <2 x half> + call void @llvm.amdgcn.exp.compr.v2f16(i32 0, i32 15, <2 x half> %tmp13, <2 x half> undef, i1 true, i1 true) #0 ret void } -; Function Attrs: nounwind readnone -declare float @llvm.SI.load.const(<16 x i8>, i32) #1 - -declare void @llvm.SI.export(i32, i32, i32, i32, i32, float, float, float, float) +declare void @llvm.amdgcn.exp.compr.v2f16(i32, i32, <2 x half>, <2 x half>, i1, i1) #0 +declare float @llvm.SI.load.const(<16 x i8>, i32) #1 declare <4 x float> @llvm.SI.image.sample.v2i32(<2 x i32>, <8 x i32>, <4 x i32>, i32, i32, i32, i32, i32, i32, i32, i32) #1 - declare i32 @llvm.SI.packf16(float, float) #1 attributes #0 = { nounwind } attributes #1 = { nounwind readnone } !0 = !{!1, !1, i64 0, i32 1} -!1 = !{!"const", !3} -!2 = !{!1, !1, i64 0} -!3 = !{!"tbaa root"} +!1 = !{!"const", !2} +!2 = !{!"tbaa root"} |
