[llvm-branch-commits] [llvm] InstCombine: Handle fmul in SimplifyDemandedFPClass (PR #173872)

[Yingwei Zheng via llvm-branch-commits]

================
@@ -2280,6 +2290,127 @@ Value *InstCombinerImpl::SimplifyDemandedUseFPClass(Value *V,
     Known = KnownLHS | KnownRHS;
     break;
   }
+  case Instruction::FMul: {
+    KnownFPClass KnownLHS, KnownRHS;
+
+    Value *X = I->getOperand(0);
+    Value *Y = I->getOperand(1);
+
+    FPClassTest SrcDemandedMask =
+        DemandedMask & (fcNan | fcZero | fcSubnormal | fcNormal);
+
+    if (DemandedMask & fcInf) {
+      // mul x, inf = inf
+      // mul large_x, large_y = inf
+      SrcDemandedMask |= fcSubnormal | fcNormal | fcInf;
+    }
+
+    if (DemandedMask & fcNan) {
+      // mul +/-inf, 0 => nan
+      SrcDemandedMask |= fcZero | fcInf;
+
+      // TODO: Mode check
+      // mul +/-inf, sub => nan if daz
+      SrcDemandedMask |= fcSubnormal;
+    }
+
+    if (X == Y) {
+      if (SimplifyDemandedFPClass(I, 0, SrcDemandedMask, KnownLHS, Depth + 1))
+        return I;
----------------
dtcxzyw wrote:

It is a bit weird that the RHS doesn't get simplified as well.

https://github.com/llvm/llvm-project/pull/173872