1 ; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx906 -verify-machineinstrs -o - %s | FileCheck %s
3 target datalayout = "e-p:64:64-p1:64:64-p2:32:32-p3:32:32-p4:64:64-p5:32:32-p6:32:32-p7:160:256:256:32-p8:128:128-i64:64-v16:16-v24:32-v32:32-v48:64-v96:128-v192:256-v256:256-v512:512-v1024:1024-v2048:2048-n32:64-S32-A5-ni:7:8"
4 target triple = "amdgcn-amd-amdhsa"
6 ; CHECK-LABEL: {{^}}t0:
7 ; CHECK: s_load_dwordx2 s[[[PTR_LO:[0-9]+]]:[[PTR_HI:[0-9]+]]], s[4:5], 0x0
8 ; CHECK: v_mov_b32_e32 v{{[0-9]+}}, s[[PTR_HI]]
9 ; There should be no redundant copies from PTR_HI.
10 ; CHECK-NOT: v_mov_b32_e32 v{{[0-9]+}}, s[[PTR_HI]]
11 define protected amdgpu_kernel void @t0(ptr addrspace(1) %p, i32 %i0, i32 %j0, i32 %k0) {
13 %0 = tail call i32 @llvm.amdgcn.workitem.id.x()
17 %pi = getelementptr float, ptr addrspace(1) %p, i32 %i
18 %vi = load float, ptr addrspace(1) %pi
19 %pj = getelementptr float, ptr addrspace(1) %p, i32 %j
20 %vj = load float, ptr addrspace(1) %pj
21 %sum = fadd float %vi, %vj
22 %pk = getelementptr float, ptr addrspace(1) %p, i32 %k
23 store float %sum, ptr addrspace(1) %pk
27 declare i32 @llvm.amdgcn.workitem.id.x()