[llvm] [AMDGPU][Uniformity][TTI] Make Uniformity Analysis Operand-Aware via Custom Divergence Checks (PR #137639)

[via llvm-commits]

https://github.com/ruiling commented:

Thanks for making the change so quickly. It all looks good to me. But there is one mistake we need to fix. Thanks!

https://github.com/llvm/llvm-project/pull/137639