[llvm] r288865 - AMDGPU: Add llvm.amdgcn.interp.mov intrinsic
Tom Stellard via llvm-commits
llvm-commits at lists.llvm.org
Tue Dec 6 15:52:14 PST 2016
Author: tstellar
Date: Tue Dec 6 17:52:13 2016
New Revision: 288865
URL: http://llvm.org/viewvc/llvm-project?rev=288865&view=rev
Log:
AMDGPU: Add llvm.amdgcn.interp.mov intrinsic
Reviewers: arsenm, nhaehnle
Subscribers: kzhuravl, wdng, yaxunl, tony-tye, llvm-commits
Differential Revision: https://reviews.llvm.org/D26725
Modified:
llvm/trunk/include/llvm/IR/IntrinsicsAMDGPU.td
llvm/trunk/lib/Target/AMDGPU/SIISelLowering.cpp
llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.interp.ll
Modified: llvm/trunk/include/llvm/IR/IntrinsicsAMDGPU.td
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/include/llvm/IR/IntrinsicsAMDGPU.td?rev=288865&r1=288864&r2=288865&view=diff
==============================================================================
--- llvm/trunk/include/llvm/IR/IntrinsicsAMDGPU.td (original)
+++ llvm/trunk/include/llvm/IR/IntrinsicsAMDGPU.td Tue Dec 6 17:52:13 2016
@@ -475,6 +475,14 @@ def int_amdgcn_s_getreg :
GCCBuiltin<"__builtin_amdgcn_s_getreg">,
Intrinsic<[llvm_i32_ty], [llvm_i32_ty], [IntrReadMem]>;
+// __builtin_amdgcn_interp_mov <param>, <attr_chan>, <attr>, <m0>
+// param values: 0 = P10, 1 = P20, 2 = P0
+def int_amdgcn_interp_mov :
+ GCCBuiltin<"__builtin_amdgcn_interp_mov">,
+ Intrinsic<[llvm_float_ty],
+ [llvm_i32_ty, llvm_i32_ty, llvm_i32_ty, llvm_i32_ty],
+ [IntrNoMem]>;
+
// __builtin_amdgcn_interp_p1 <i>, <attr_chan>, <attr>, <m0>
def int_amdgcn_interp_p1 :
GCCBuiltin<"__builtin_amdgcn_interp_p1">,
Modified: llvm/trunk/lib/Target/AMDGPU/SIISelLowering.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AMDGPU/SIISelLowering.cpp?rev=288865&r1=288864&r2=288865&view=diff
==============================================================================
--- llvm/trunk/lib/Target/AMDGPU/SIISelLowering.cpp (original)
+++ llvm/trunk/lib/Target/AMDGPU/SIISelLowering.cpp Tue Dec 6 17:52:13 2016
@@ -2497,6 +2497,12 @@ SDValue SITargetLowering::LowerINTRINSIC
return DAG.getNode(AMDGPUISD::INTERP_P2, DL, MVT::f32, P1, J,
Op.getOperand(1), Op.getOperand(2), Glue);
}
+ case Intrinsic::amdgcn_interp_mov: {
+ SDValue M0 = copyToM0(DAG, DAG.getEntryNode(), DL, Op.getOperand(4));
+ SDValue Glue = M0.getValue(1);
+ return DAG.getNode(AMDGPUISD::INTERP_MOV, DL, MVT::f32, Op.getOperand(1),
+ Op.getOperand(2), Op.getOperand(3), Glue);
+ }
case Intrinsic::amdgcn_interp_p1: {
SDValue M0 = copyToM0(DAG, DAG.getEntryNode(), DL, Op.getOperand(4));
SDValue Glue = M0.getValue(1);
Modified: llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.interp.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.interp.ll?rev=288865&r1=288864&r2=288865&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.interp.ll (original)
+++ llvm/trunk/test/CodeGen/AMDGPU/llvm.amdgcn.interp.ll Tue Dec 6 17:52:13 2016
@@ -14,7 +14,9 @@ main_body:
%p1_0 = call float @llvm.amdgcn.interp.p2(float %p0_0, float %j, i32 0, i32 0, i32 %3)
%p0_1 = call float @llvm.amdgcn.interp.p1(float %i, i32 1, i32 0, i32 %3)
%p1_1 = call float @llvm.amdgcn.interp.p2(float %p0_1, float %j, i32 1, i32 0, i32 %3)
- call void @llvm.SI.export(i32 15, i32 1, i32 1, i32 0, i32 1, float %p0_0, float %p0_0, float %p1_1, float %p1_1)
+ %const = call float @llvm.amdgcn.interp.mov(i32 2, i32 0, i32 0, i32 %3)
+ %w = fadd float %p1_1, %const
+ call void @llvm.SI.export(i32 15, i32 1, i32 1, i32 0, i32 1, float %p0_0, float %p0_0, float %p1_1, float %w)
ret void
}
@@ -24,6 +26,8 @@ declare float @llvm.amdgcn.interp.p1(flo
; Function Attrs: nounwind readnone
declare float @llvm.amdgcn.interp.p2(float, float, i32, i32, i32) #0
+declare float @llvm.amdgcn.interp.mov(i32, i32, i32, i32) #0
+
declare void @llvm.SI.export(i32, i32, i32, i32, i32, float, float, float, float)
attributes #0 = { nounwind readnone }
More information about the llvm-commits
mailing list