[llvm] [LoopVectorize] Improve Vectorization of Low Trip Count Loops (PR #195823)

[Florian Hahn via llvm-commits]

https://github.com/fhahn commented:

> I've added this. No other changes are needed as the implementation won't perform the transformation when CM_EpilogueNotAllowedOptSize is set.

Is this true? Looking at `tc5_v4i32_reject_minsize`, it gets vectorized, followed by scalar loop AFAICT

https://github.com/llvm/llvm-project/pull/195823