[llvm] [InstCombine] Improve select simplification based on known bits (PR #97289)

[Yingwei Zheng via llvm-commits]

dtcxzyw wrote:

> > Can we revert #66740 after landing this patch? @cyyself found that #66740 caused a significant compile-time regression in SLPVectorizer when building RTL simulator for https://github.com/OpenXiangShan/XiangShan.
> 
> I don't really get how the optimization in this PR would help the motivating issues for #66740?

Sorry, I misread the fold in #66740 as `B == odd ? (B & 1) : 0 -> B == odd ? 1 : 0`.


https://github.com/llvm/llvm-project/pull/97289