[llvm] [LV, VP]VP intrinsics support for the Loop Vectorizer + adding new tail-folding mode using EVL. (PR #76172)

[Florian Hahn via llvm-commits]

================
@@ -9445,14 +9552,31 @@ void VPWidenMemoryInstructionRecipe::execute(VPTransformState &State) {
     }
   }
 
+  auto MaskValue = [&](unsigned Part) -> Value * {
+    if (isMaskRequired)
+      return BlockInMaskParts[Part];
+    return nullptr;
+  };
+
   // Handle Stores:
   if (SI) {
     State.setDebugLocFrom(SI->getDebugLoc());
 
     for (unsigned Part = 0; Part < State.UF; ++Part) {
       Instruction *NewSI = nullptr;
       Value *StoredVal = State.get(StoredValue, Part);
-      if (CreateGatherScatter) {
+      if (State.EVL) {
----------------
fhahn wrote:

As mentioned earlier, I think we are in agreement that the recipe should be split up (also independent of EVL support). It's something I am planning to get around to do sometime soon, but there are a number of other changes I would like  to wrap up first.

So I don't think we need to block the current patch on the refactoring, but leave a TODO to split up the recipe in general.

https://github.com/llvm/llvm-project/pull/76172