[llvm] [LV] Vectorize fmin/fmax reductions (PR #198300)

[Nashe Mncube via llvm-commits]

nasherm wrote:

> I was just commenting on the apparent contradiction between what the PR title and what was in the test. I don't know enough about the semantics of the various kinds for fp min/max to have an opinion on whether we should be vectorizing or not.

I've updated the PR description to be more specific. I've also updated  `handleMultiUseReductions` so that we don't attempt to vectorize `FMaxNum/FMinNum` reductions

https://github.com/llvm/llvm-project/pull/198300