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

[Florian Hahn via llvm-commits]

================
@@ -0,0 +1,98 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
----------------
fhahn wrote:

should this pass `-force-vector-width`? at the moment we vectorize with VF = vscale x 1 and I think this is causing a UBSan failure the running the following tests

```
 LLVM :: Transforms/LoopVectorize/vectorize-vp-intrinsics-iv32.ll
  LLVM :: Transforms/LoopVectorize/vectorize-vp-intrinsics.ll
  LLVM :: Transforms/LoopVectorize/vplan-vp-intrinsics.ll
```

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