1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-22 04:22:57 +02:00
llvm-mirror/test/CodeGen/AMDGPU/llvm.amdgcn.rsq.legacy.ll
Matt Arsenault 42dd6642c6 AMDGPU: Fold more custom nodes to undef
This will help sneak undefs past GVN into the DAG for
some tests.

Also add missing intrinsic for rsq_legacy, even though the node
was already selected to the instruction. Also start passing
the debug location to intrinsic errors.

llvm-svn: 273181
2016-06-20 18:33:56 +00:00

40 lines
1.4 KiB
LLVM

; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=SI -check-prefix=FUNC %s
declare float @llvm.amdgcn.rsq.legacy(float) #0
; FUNC-LABEL: {{^}}rsq_legacy_f32:
; SI: v_rsq_legacy_f32_e32 {{v[0-9]+}}, {{s[0-9]+}}
define void @rsq_legacy_f32(float addrspace(1)* %out, float %src) #1 {
%rsq = call float @llvm.amdgcn.rsq.legacy(float %src) #0
store float %rsq, float addrspace(1)* %out, align 4
ret void
}
; TODO: Really these should be constant folded
; FUNC-LABEL: {{^}}rsq_legacy_f32_constant_4.0
; SI: v_rsq_legacy_f32_e32 {{v[0-9]+}}, 4.0
define void @rsq_legacy_f32_constant_4.0(float addrspace(1)* %out) #1 {
%rsq = call float @llvm.amdgcn.rsq.legacy(float 4.0) #0
store float %rsq, float addrspace(1)* %out, align 4
ret void
}
; FUNC-LABEL: {{^}}rsq_legacy_f32_constant_100.0
; SI: v_rsq_legacy_f32_e32 {{v[0-9]+}}, 0x42c80000
define void @rsq_legacy_f32_constant_100.0(float addrspace(1)* %out) #1 {
%rsq = call float @llvm.amdgcn.rsq.legacy(float 100.0) #0
store float %rsq, float addrspace(1)* %out, align 4
ret void
}
; FUNC-LABEL: {{^}}rsq_legacy_undef_f32:
; SI-NOT: v_rsq_legacy_f32
define void @rsq_legacy_undef_f32(float addrspace(1)* %out) #1 {
%rsq = call float @llvm.amdgcn.rsq.legacy(float undef)
store float %rsq, float addrspace(1)* %out, align 4
ret void
}
attributes #0 = { nounwind readnone }
attributes #1 = { nounwind }