[llvm-branch-commits] [llvm] ValueTracking: Handle amdgcn.rsq intrinsic in computeKnownFPClass (PR #171837)

[Matt Arsenault via llvm-branch-commits]

================
@@ -5553,6 +5553,37 @@ void computeKnownFPClass(const Value *V, const APInt &DemandedElts,
 
       // TODO: Copy inf handling from instructions
       break;
+    case Intrinsic::amdgcn_rsq: {
+      KnownFPClass KnownSrc;
+      // The only negative value that can be returned is -0 for -0 inputs.
+      Known.knownNot(fcNegInf | fcNegSubnormal | fcNegNormal);
+
+      computeKnownFPClass(II->getArgOperand(0), DemandedElts, InterestedClasses,
+                          KnownSrc, Q, Depth + 1);
+
+      if (KnownSrc.isKnownNever(fcSNan))
+        Known.knownNot(fcSNan);
+
+      // Negative -> nan
+      if (KnownSrc.isKnownNeverNaN() && KnownSrc.cannotBeOrderedLessThanZero())
+        Known.knownNot(fcNan);
+
+      Type *EltTy = II->getType()->getScalarType();
+
+      // f32 denormal always flushed.
+      if (EltTy->isFloatTy())
+        Known.knownNot(fcPosSubnormal);
----------------
arsenm wrote:

There was just no documentation before, the instruction never changed 

https://github.com/llvm/llvm-project/pull/171837