[llvm] [LV, VP]VP intrinsics support for the Loop Vectorizer (PR #76172)

[Alexey Bataev via llvm-commits]

================
@@ -4690,6 +4712,39 @@ LoopVectorizationCostModel::computeMaxVF(ElementCount UserVF, unsigned UserIC) {
   // FIXME: look for a smaller MaxVF that does divide TC rather than masking.
   if (Legal->prepareToFoldTailByMasking()) {
----------------
alexey-bataev wrote:

I simplified it as much as I could. If you have any particular comments, how it can be improved, please be more specific.

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