[llvm] [X86][LV] Prefer tail folding for i64 div/rem loops (PR #227721)

[Florian Hahn via llvm-commits]

https://github.com/fhahn commented:

This PR probably needs performance numbers to make sure this is always profitable

> I confirm that I have read the LLVM Developer Policy, LLVM Code-Review Policy, and LLVM AI Tool Use Policy, and have noted AI tool usage in the PR description.

The PR description should be written by yourself (see https://llvm.org/docs/AIToolPolicy.html#details)

Currently it looks like it was AI generated (I may be wrong though). Please take a look at other commit messages. Remove markdown, remove unneeded stuff like `Testing` section etc.

https://github.com/llvm/llvm-project/pull/227721