[llvm] [SLP] Initial vectorization of non-power-of-2 ops. (PR #77790)

[Florian Hahn via llvm-commits]

https://github.com/fhahn commented:

> Would be good to see more tests for other targets. May help to check that TTI does not crash and vectorizatyion works for these targets too.

Excellent point, I added a number of X86 tests (check lines updated in the latest version of the PR) and also added extra X86 for issues I found while testing on X86. AArch64 and X86 are the architectures I can test locally, are there any particular other targets you think would be good to have tests for ?

https://github.com/llvm/llvm-project/pull/77790