[llvm] [ValueTracking] Add support for most `llvm.vector.reduce.*` ops. (PR #88320)

[Nikita Popov via llvm-commits]

================
@@ -5004,6 +5077,15 @@ void computeKnownFPClass(const Value *V, const APInt &DemandedElts,
 
       break;
     }
+      // reduce min/max will choose an element from one of the vector elements,
+      // so we can infer and class information that is common to all elements.
+    case Intrinsic::vector_reduce_fmax:
+    case Intrinsic::vector_reduce_fmin:
+    case Intrinsic::vector_reduce_fmaximum:
+    case Intrinsic::vector_reduce_fminimum:
+      computeKnownFPClass(II->getArgOperand(0), Known, InterestedClasses,
+                          Depth + 1, Q);
+      break;
----------------
nikic wrote:

Can you please split this into a separate PR for review by FP folks? I'm not sure things are quite as simple as that (e.g., do we need to account for denormal flush like the scalar min/max handling above)?

https://github.com/llvm/llvm-project/pull/88320