1; RUN:  llc -amdgpu-scalarize-global-loads=false  -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s
2; RUN:  llc -amdgpu-scalarize-global-loads=false  -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s
3
4declare i32 @llvm.amdgcn.sffbh.i32(i32) #1
5
6; GCN-LABEL: {{^}}s_flbit:
7; GCN: s_load_dword [[VAL:s[0-9]+]],
8; GCN: s_flbit_i32 [[SRESULT:s[0-9]+]], [[VAL]]
9; GCN: v_mov_b32_e32 [[VRESULT:v[0-9]+]], [[SRESULT]]
10; GCN: buffer_store_dword [[VRESULT]],
11define amdgpu_kernel void @s_flbit(i32 addrspace(1)* noalias %out, i32 %val) #0 {
12  %r = call i32 @llvm.amdgcn.sffbh.i32(i32 %val)
13  store i32 %r, i32 addrspace(1)* %out, align 4
14  ret void
15}
16
17; GCN-LABEL: {{^}}v_flbit:
18; GCN: buffer_load_dword [[VAL:v[0-9]+]],
19; GCN: v_ffbh_i32_e32 [[RESULT:v[0-9]+]], [[VAL]]
20; GCN: buffer_store_dword [[RESULT]],
21define amdgpu_kernel void @v_flbit(i32 addrspace(1)* noalias %out, i32 addrspace(1)* noalias %valptr) #0 {
22  %val = load i32, i32 addrspace(1)* %valptr, align 4
23  %r = call i32 @llvm.amdgcn.sffbh.i32(i32 %val)
24  store i32 %r, i32 addrspace(1)* %out, align 4
25  ret void
26}
27
28attributes #0 = { nounwind }
29attributes #1 = { nounwind readnone }
30