[llvm] [SLP] Do not skip tiny trees with gathered loads to vectorize (PR #190040)

[Alexey Bataev via llvm-commits]

alexey-bataev wrote:

> Looks like this PR causes llvm/clang to enter an infinite loop when building clang-tblgen (with avx512+LTO+PGO), but I haven't been able to extract a minimal reproducer yet. I'm not sure if the code itself is trapped in an infinite loop or if it's caused by a miscompile.

Ok, going to revert

https://github.com/llvm/llvm-project/pull/190040