1 ; RUN: llc -march=amdgcn -mcpu=tahiti -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s
2 ; RUN: llc -march=amdgcn -mcpu=tonga -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s
4 declare double @llvm.maxnum.f64(double, double) nounwind readnone
6 ; SI-LABEL: {{^}}test_fmax3_f64:
7 ; SI-DAG: buffer_load_dwordx2 [[REGA:v\[[0-9]+:[0-9]+\]]], off, s[{{[0-9]+:[0-9]+}}], 0{{$}}
8 ; SI-DAG: buffer_load_dwordx2 [[REGB:v\[[0-9]+:[0-9]+\]]], off, s[{{[0-9]+:[0-9]+}}], 0 offset:8
9 ; SI: v_max_f64 [[REGA]], [[REGA]], [[REGB]]
10 ; SI: buffer_load_dwordx2 [[REGC:v\[[0-9]+:[0-9]+\]]], off, s[{{[0-9]+:[0-9]+}}], 0 offset:16
11 ; SI: v_max_f64 [[RESULT:v\[[0-9]+:[0-9]+\]]], [[REGA]], [[REGC]]
12 ; SI: buffer_store_dwordx2 [[RESULT]],
14 define amdgpu_kernel void @test_fmax3_f64(double addrspace(1)* %out, double addrspace(1)* %aptr) nounwind {
15 %bptr = getelementptr double, double addrspace(1)* %aptr, i32 1
16 %cptr = getelementptr double, double addrspace(1)* %aptr, i32 2
17 %a = load volatile double, double addrspace(1)* %aptr, align 8
18 %b = load volatile double, double addrspace(1)* %bptr, align 8
19 %c = load volatile double, double addrspace(1)* %cptr, align 8
20 %f0 = call double @llvm.maxnum.f64(double %a, double %b) nounwind readnone
21 %f1 = call double @llvm.maxnum.f64(double %f0, double %c) nounwind readnone
22 store double %f1, double addrspace(1)* %out, align 8