Home | History | Annotate | Download | only in AMDGPU
      1 ;RUN: llc < %s -march=amdgcn -mcpu=verde -verify-machineinstrs | FileCheck %s
      2 ;RUN: llc < %s -march=amdgcn -mcpu=tonga -verify-machineinstrs | FileCheck %s
      3 
      4 ; This shader has the potential to generated illegal VGPR to SGPR copies if
      5 ; the wrong register class is used for the REG_SEQUENCE instructions.
      6 
      7 ; CHECK: {{^}}main:
      8 ; CHECK: image_sample_b v{{\[[0-9]:[0-9]\]}}, 15, 0, 0, 0, 0, 0, 0, 0, v{{\[[0-9]:[0-9]\]}}
      9 
     10 define void @main(<16 x i8> addrspace(2)* inreg, <16 x i8> addrspace(2)* inreg, <32 x i8> addrspace(2)* inreg, i32 inreg, <2 x i32>, <2 x i32>, <2 x i32>, <3 x i32>, <2 x i32>, <2 x i32>, <2 x i32>, float, float, float, float, float, float, float, float, float) #0 {
     11 main_body:
     12   %20 = getelementptr <16 x i8>, <16 x i8> addrspace(2)* %0, i32 0
     13   %21 = load <16 x i8>, <16 x i8> addrspace(2)* %20, !tbaa !1
     14   %22 = call float @llvm.SI.load.const(<16 x i8> %21, i32 16)
     15   %23 = getelementptr <32 x i8>, <32 x i8> addrspace(2)* %2, i32 0
     16   %24 = load <32 x i8>, <32 x i8> addrspace(2)* %23, !tbaa !1
     17   %25 = getelementptr <16 x i8>, <16 x i8> addrspace(2)* %1, i32 0
     18   %26 = load <16 x i8>, <16 x i8> addrspace(2)* %25, !tbaa !1
     19   %27 = call float @llvm.SI.fs.interp(i32 0, i32 0, i32 %3, <2 x i32> %5)
     20   %28 = call float @llvm.SI.fs.interp(i32 1, i32 0, i32 %3, <2 x i32> %5)
     21   %29 = bitcast float %22 to i32
     22   %30 = bitcast float %27 to i32
     23   %31 = bitcast float %28 to i32
     24   %32 = insertelement <4 x i32> undef, i32 %29, i32 0
     25   %33 = insertelement <4 x i32> %32, i32 %30, i32 1
     26   %34 = insertelement <4 x i32> %33, i32 %31, i32 2
     27   %35 = insertelement <4 x i32> %34, i32 undef, i32 3
     28   %36 = call <4 x float> @llvm.SI.sampleb.v4i32(<4 x i32> %35, <32 x i8> %24, <16 x i8> %26, i32 2)
     29   %37 = extractelement <4 x float> %36, i32 0
     30   %38 = extractelement <4 x float> %36, i32 1
     31   %39 = extractelement <4 x float> %36, i32 2
     32   %40 = extractelement <4 x float> %36, i32 3
     33   call void @llvm.SI.export(i32 15, i32 1, i32 1, i32 0, i32 0, float %37, float %38, float %39, float %40)
     34   ret void
     35 }
     36 
     37 ; Function Attrs: nounwind readnone
     38 declare float @llvm.SI.load.const(<16 x i8>, i32) #1
     39 
     40 ; Function Attrs: nounwind readnone
     41 declare float @llvm.SI.fs.interp(i32, i32, i32, <2 x i32>) #1
     42 
     43 ; Function Attrs: nounwind readnone
     44 declare <4 x float> @llvm.SI.sampleb.v4i32(<4 x i32>, <32 x i8>, <16 x i8>, i32) #1
     45 
     46 declare void @llvm.SI.export(i32, i32, i32, i32, i32, float, float, float, float)
     47 
     48 attributes #0 = { "ShaderType"="0" }
     49 attributes #1 = { nounwind readnone }
     50 
     51 !0 = !{!"const", null}
     52 !1 = !{!0, !0, i64 0, i32 1}
     53