[llvm] [LV] Vectorize maxnum/minnum w/o fast-math flags. (PR #148239)

[via llvm-commits]

================
@@ -652,3 +652,146 @@ void VPlanTransforms::attachCheckBlock(VPlan &Plan, Value *Cond,
     Term->addMetadata(LLVMContext::MD_prof, BranchWeights);
   }
 }
+
+bool VPlanTransforms::handleMaxMinNumReductionsWithoutFastMath(VPlan &Plan) {
+  VPRegionBlock *LoopRegion = Plan.getVectorLoopRegion();
+  VPReductionPHIRecipe *RedPhiR = nullptr;
+  bool HasUnsupportedPhi = false;
----------------
ayalz wrote:

Define these closer to their use - after the lambda?

https://github.com/llvm/llvm-project/pull/148239