1 ;RUN: llc < %s -march=r600 -mcpu=SI | FileCheck %s 2 3 ; CHECK: S_ENDPGM 4 5 define void @main() { 6 main_body: 7 call void @llvm.AMDGPU.shader.type(i32 1) 8 %0 = load <4 x i32> addrspace(2)* addrspace(8)* inttoptr (i32 6 to <4 x i32> addrspace(2)* addrspace(8)*) 9 %1 = getelementptr <4 x i32> addrspace(2)* %0, i32 0 10 %2 = load <4 x i32> addrspace(2)* %1 11 %3 = call i32 @llvm.SI.vs.load.buffer.index() 12 %4 = call <4 x float> @llvm.SI.vs.load.input(<4 x i32> %2, i32 0, i32 %3) 13 %5 = extractelement <4 x float> %4, i32 0 14 %6 = extractelement <4 x float> %4, i32 1 15 %7 = extractelement <4 x float> %4, i32 2 16 %8 = extractelement <4 x float> %4, i32 3 17 %9 = load <4 x i32> addrspace(2)* addrspace(8)* inttoptr (i32 6 to <4 x i32> addrspace(2)* addrspace(8)*) 18 %10 = getelementptr <4 x i32> addrspace(2)* %9, i32 1 19 %11 = load <4 x i32> addrspace(2)* %10 20 %12 = call i32 @llvm.SI.vs.load.buffer.index() 21 %13 = call <4 x float> @llvm.SI.vs.load.input(<4 x i32> %11, i32 0, i32 %12) 22 %14 = extractelement <4 x float> %13, i32 0 23 %15 = extractelement <4 x float> %13, i32 1 24 %16 = extractelement <4 x float> %13, i32 2 25 %17 = extractelement <4 x float> %13, i32 3 26 call void @llvm.SI.export(i32 15, i32 0, i32 0, i32 32, i32 0, float %14, float %15, float %16, float %17) 27 call void @llvm.SI.export(i32 15, i32 0, i32 1, i32 12, i32 0, float %5, float %6, float %7, float %8) 28 ret void 29 } 30 31 declare void @llvm.AMDGPU.shader.type(i32) 32 33 declare i32 @llvm.SI.vs.load.buffer.index() readnone 34 35 declare <4 x float> @llvm.SI.vs.load.input(<4 x i32>, i32, i32) 36 37 declare void @llvm.SI.export(i32, i32, i32, i32, i32, float, float, float, float) 38
