[llvm-branch-commits] [llvm] InstCombine: Handle fdiv in SimplifyDemandedFPClass (PR #175946)

[Yingwei Zheng via llvm-branch-commits]

================
@@ -2331,6 +2331,136 @@ Value *InstCombinerImpl::SimplifyDemandedUseFPClass(Instruction *I,
     FPClassTest ValidResults = DemandedMask & Known.KnownFPClasses;
     return getFPClassConstant(VTy, ValidResults, /*IsCanonicalizing=*/true);
   }
+  case Instruction::FDiv: {
+    Value *X = I->getOperand(0);
+    Value *Y = I->getOperand(1);
+    if (X == Y && isGuaranteedNotToBeUndef(X, SQ.AC, CxtI, SQ.DT, Depth + 1)) {
+      // If the source is 0, inf or nan, the result is a nan
+
+      Value *IsZeroOrNan = Builder.CreateFCmpFMF(
+          FCmpInst::FCMP_UEQ, I->getOperand(0), ConstantFP::getZero(VTy), FMF);
+
+      Value *Fabs =
+          Builder.CreateUnaryIntrinsic(Intrinsic::fabs, I->getOperand(0), FMF);
+      Value *IsInfOrNan = Builder.CreateFCmpFMF(
+          FCmpInst::FCMP_UEQ, Fabs, ConstantFP::getInfinity(VTy), FMF);
+
+      Value *IsInfOrZeroOrNan = Builder.CreateOr(IsInfOrNan, IsZeroOrNan);
----------------
dtcxzyw wrote:

Use is.fpclass instead. When is_zero/is_inf gets simplified, InstCombine will convert it into the canonical form `fcmp fabs(x), inf/0`.

https://github.com/llvm/llvm-project/pull/175946