[llvm] [X86] Improve FREEZE node elimination for SETCC operations (PR #192362)

[Feng Zou via llvm-commits]

fzou1 wrote:

> If we're confident the source node types to setcc/carry can't create poison, and we drop the canCreateUndefOrPoisonForTargetNode override I think we should be OK - isGuaranteedNotToBeUndefOrPoisonForTargetNode can't get us stuck in the same way (it'll just immediately fold freeze(setcc()) -> setcc()).

I'm okay to remove the canCreateUndefOrPoisonForTargetNode override for SETCC/SETCC_CARRY. And I've added two more LIT tests. Please help take a look. Thanks.

https://github.com/llvm/llvm-project/pull/192362