[llvm] r317878 - [X86] Add support for combining FMADDSUB(A, B, FNEG(C))->FMSUBADD(A, B, C)

Craig Topper via llvm-commits llvm-commits at lists.llvm.org
Fri Nov 10 00:22:38 PST 2017


Author: ctopper
Date: Fri Nov 10 00:22:37 2017
New Revision: 317878

URL: http://llvm.org/viewvc/llvm-project?rev=317878&view=rev
Log:
[X86] Add support for combining FMADDSUB(A, B, FNEG(C))->FMSUBADD(A, B, C)

Support the opposite direction as well. Also add a TODO for not being able to combine FMSUB/FNMADD/FNMSUB with FNEG.

Modified:
    llvm/trunk/lib/Target/X86/X86ISelLowering.cpp
    llvm/trunk/test/CodeGen/X86/fma-fneg-combine.ll

Modified: llvm/trunk/lib/Target/X86/X86ISelLowering.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86ISelLowering.cpp?rev=317878&r1=317877&r2=317878&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86ISelLowering.cpp (original)
+++ llvm/trunk/lib/Target/X86/X86ISelLowering.cpp Fri Nov 10 00:22:37 2017
@@ -35474,6 +35474,7 @@ static SDValue combineSext(SDNode *N, Se
 
 static SDValue combineFMA(SDNode *N, SelectionDAG &DAG,
                           const X86Subtarget &Subtarget) {
+  // TODO: Handle FMSUB/FNMADD/FNMSUB as the starting opcode.
   SDLoc dl(N);
   EVT VT = N->getValueType(0);
 
@@ -35574,6 +35575,32 @@ static SDValue combineFMA(SDNode *N, Sel
   return SDValue();
 }
 
+// Combine FMADDSUB(A, B, FNEG(C)) -> FMSUBADD(A, B, C)
+static SDValue combineFMADDSUB(SDNode *N, SelectionDAG &DAG,
+                               const X86Subtarget &Subtarget) {
+  SDLoc dl(N);
+  EVT VT = N->getValueType(0);
+
+  SDValue NegVal = isFNEG(N->getOperand(2).getNode());
+  if (!NegVal)
+    return SDValue();
+
+  unsigned NewOpcode;
+  switch (N->getOpcode()) {
+  default: llvm_unreachable("Unexpected opcode!");
+  case X86ISD::FMADDSUB:     NewOpcode = X86ISD::FMSUBADD;     break;
+  case X86ISD::FMADDSUB_RND: NewOpcode = X86ISD::FMSUBADD_RND; break;
+  case X86ISD::FMSUBADD:     NewOpcode = X86ISD::FMADDSUB;     break;
+  case X86ISD::FMSUBADD_RND: NewOpcode = X86ISD::FMADDSUB_RND; break;
+  }
+
+  if (N->getNumOperands() == 4)
+    return DAG.getNode(NewOpcode, dl, VT, N->getOperand(0), N->getOperand(1),
+                       NegVal, N->getOperand(3));
+  return DAG.getNode(NewOpcode, dl, VT, N->getOperand(0), N->getOperand(1),
+                     NegVal);
+}
+
 static SDValue combineZext(SDNode *N, SelectionDAG &DAG,
                            TargetLowering::DAGCombinerInfo &DCI,
                            const X86Subtarget &Subtarget) {
@@ -36868,6 +36895,10 @@ SDValue X86TargetLowering::PerformDAGCom
   case X86ISD::FMADDS1:
   case X86ISD::FMADDS3:
   case ISD::FMA:            return combineFMA(N, DAG, Subtarget);
+  case X86ISD::FMADDSUB_RND:
+  case X86ISD::FMSUBADD_RND:
+  case X86ISD::FMADDSUB:
+  case X86ISD::FMSUBADD:    return combineFMADDSUB(N, DAG, Subtarget);
   case ISD::MGATHER:
   case ISD::MSCATTER:       return combineGatherScatter(N, DAG);
   case X86ISD::TESTM:       return combineTestM(N, DAG, Subtarget);

Modified: llvm/trunk/test/CodeGen/X86/fma-fneg-combine.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/fma-fneg-combine.ll?rev=317878&r1=317877&r2=317878&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/fma-fneg-combine.ll (original)
+++ llvm/trunk/test/CodeGen/X86/fma-fneg-combine.ll Fri Nov 10 00:22:37 2017
@@ -97,7 +97,7 @@ define <8 x float> @test8(<8 x float> %a
 ;
 ; KNL-LABEL: test8:
 ; KNL:       # BB#0: # %entry
-; KNL-NEXT:    vbroadcastss {{.*}}(%rip), %ymm3
+; KNL-NEXT:    vbroadcastss {{.*#+}} ymm3 = [-0,-0,-0,-0,-0,-0,-0,-0]
 ; KNL-NEXT:    vxorps %ymm3, %ymm2, %ymm2
 ; KNL-NEXT:    vfmsub213ps %ymm2, %ymm1, %ymm0
 ; KNL-NEXT:    retq
@@ -147,7 +147,7 @@ define <4 x float> @test11(<4 x float> %
 ;
 ; KNL-LABEL: test11:
 ; KNL:       # BB#0: # %entry
-; KNL-NEXT:    vbroadcastss {{.*}}(%rip), %xmm0
+; KNL-NEXT:    vbroadcastss {{.*#+}} xmm0 = [-0,-0,-0,-0]
 ; KNL-NEXT:    vxorps %xmm0, %xmm2, %xmm0
 ; KNL-NEXT:    kmovw %edi, %k1
 ; KNL-NEXT:    vfmadd231ss %xmm1, %xmm1, %xmm0 {%k1}
@@ -270,3 +270,38 @@ entry:
   ret <16 x float> %1
 }
 
+define <16 x float> @test16(<16 x float> %a, <16 x float> %b, <16 x float> %c, i16 %mask) {
+; SKX-LABEL: test16:
+; SKX:       # BB#0:
+; SKX-NEXT:    kmovd %edi, %k1
+; SKX-NEXT:    vfmsubadd132ps {rd-sae}, %zmm1, %zmm2, %zmm0 {%k1}
+; SKX-NEXT:    retq
+;
+; KNL-LABEL: test16:
+; KNL:       # BB#0:
+; KNL-NEXT:    kmovw %edi, %k1
+; KNL-NEXT:    vfmsubadd132ps {rd-sae}, %zmm1, %zmm2, %zmm0 {%k1}
+; KNL-NEXT:    retq
+  %sub.i = fsub <16 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
+  %res = call <16 x float> @llvm.x86.avx512.mask.vfmaddsub.ps.512(<16 x float> %a, <16 x float> %b, <16 x float> %sub.i, i16 %mask, i32 1)
+  ret <16 x float> %res
+}
+declare <16 x float> @llvm.x86.avx512.mask.vfmaddsub.ps.512(<16 x float>, <16 x float>, <16 x float>, i16, i32)
+
+define <8 x double> @test17(<8 x double> %a, <8 x double> %b, <8 x double> %c, i8 %mask) {
+; SKX-LABEL: test17:
+; SKX:       # BB#0:
+; SKX-NEXT:    kmovd %edi, %k1
+; SKX-NEXT:    vfmsubadd132pd %zmm1, %zmm2, %zmm0 {%k1}
+; SKX-NEXT:    retq
+;
+; KNL-LABEL: test17:
+; KNL:       # BB#0:
+; KNL-NEXT:    kmovw %edi, %k1
+; KNL-NEXT:    vfmsubadd132pd %zmm1, %zmm2, %zmm0 {%k1}
+; KNL-NEXT:    retq
+  %sub.i = fsub <8 x double> <double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00, double -0.000000e+00>, %c
+  %res = call <8 x double> @llvm.x86.avx512.mask.vfmaddsub.pd.512(<8 x double> %a, <8 x double> %b, <8 x double> %sub.i, i8 %mask, i32 4)
+  ret <8 x double> %res
+}
+declare <8 x double> @llvm.x86.avx512.mask.vfmaddsub.pd.512(<8 x double>, <8 x double>, <8 x double>, i8, i32)




More information about the llvm-commits mailing list