; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 6
; RUN: opt -passes=debugify < %s | llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx900 | FileCheck %s

@lds = addrspace(3) global [512 x float] poison, align 4

define amdgpu_kernel void @simple_write2_one_val_f32(ptr addrspace(1) %C, ptr addrspace(1) %in) #0 {
  %x.i = tail call i32 @llvm.amdgcn.workitem.id.x() #1
  %in.gep = getelementptr float, ptr addrspace(1) %in, i32 %x.i
  %val = load float, ptr addrspace(1) %in.gep, align 4
  %arrayidx0 = getelementptr inbounds [512 x float], ptr addrspace(3) @lds, i32 0, i32 %x.i
  store float %val, ptr addrspace(3) %arrayidx0, align 4
  %add.x = add nsw i32 %x.i, 8
  %arrayidx1 = getelementptr inbounds [512 x float], ptr addrspace(3) @lds, i32 0, i32 %add.x
  store float %val, ptr addrspace(3) %arrayidx1, align 4
  ret void
}

define amdgpu_kernel void @simple_read2_f32(ptr addrspace(1) %out) #0 {
  %x.i = tail call i32 @llvm.amdgcn.workitem.id.x() #1
  %arrayidx0 = getelementptr inbounds [512 x float], ptr addrspace(3) @lds, i32 0, i32 %x.i
  %val0 = load float, ptr addrspace(3) %arrayidx0, align 4
  %add.x = add nsw i32 %x.i, 8
  %arrayidx1 = getelementptr inbounds [512 x float], ptr addrspace(3) @lds, i32 0, i32 %add.x
  %val1 = load float, ptr addrspace(3) %arrayidx1, align 4
  %sum = fadd float %val0, %val1
  %out.gep = getelementptr inbounds float, ptr addrspace(1) %out, i32 %x.i
  store float %sum, ptr addrspace(1) %out.gep, align 4
  ret void
}

attributes #0 = { nounwind }
;; NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
; CHECK: {{.*}}