[llvm] [LV] Vectorize FMax w/o fast-math flags. (PR #146711)

[Ramkumar Ramachandra via llvm-commits]

================
@@ -8909,6 +8914,9 @@ VPlanPtr LoopVectorizationPlanner::tryToBuildVPlanWithVPRecipes(
 
   // Adjust the recipes for any inloop reductions.
   adjustRecipesForReductions(Plan, RecipeBuilder, Range.Start);
+  if (!VPlanTransforms::runPass(
+          VPlanTransforms::handleFMaxReductionsWithoutFastMath, *Plan))
+    return nullptr;
----------------
artagnon wrote:

Not sure why the feature is designed to be in VPlanTransforms?

https://github.com/llvm/llvm-project/pull/146711