1 ; RUN: llc -march=amdgcn -mcpu=SI -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s 2 ; RUN: llc -march=amdgcn -mcpu=tonga -verify-machineinstrs < %s | FileCheck -check-prefix=SI %s 3 4 declare double @llvm.maxnum.f64(double, double) nounwind readnone 5 6 ; SI-LABEL: {{^}}test_fmax3_f64: 7 ; SI-DAG: buffer_load_dwordx2 [[REGA:v\[[0-9]+:[0-9]+\]]], s[{{[0-9]+:[0-9]+}}], 0{{$}} 8 ; SI-DAG: buffer_load_dwordx2 [[REGB:v\[[0-9]+:[0-9]+\]]], s[{{[0-9]+:[0-9]+}}], 0 offset:8 9 ; SI-DAG: buffer_load_dwordx2 [[REGC:v\[[0-9]+:[0-9]+\]]], s[{{[0-9]+:[0-9]+}}], 0 offset:16 10 ; SI: v_max_f64 [[REGA]], [[REGA]], [[REGB]] 11 ; SI: v_max_f64 [[RESULT:v\[[0-9]+:[0-9]+\]]], [[REGA]], [[REGC]] 12 ; SI: buffer_store_dwordx2 [[RESULT]], 13 ; SI: s_endpgm 14 define 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 double, double addrspace(1)* %aptr, align 8 18 %b = load double, double addrspace(1)* %bptr, align 8 19 %c = load 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 23 ret void 24 } 25
