[clang] [llvm] [AMDGPU] Legalized FMINIMUM/FMAXIMUM operations to remove redundant canonicalization (PR #208176)

[Nikhil Kotikalapudi via cfe-commits]

nkotikal wrote:

> Is this still relevant? @Shoreshen is working on fixing some of this mess

Probably not, but what do you think? Every PR i've seen for this has a different method of implementing it (e.g. #221455 ), so if Shoreshen is planning to fix it, probably not worth pursuing.

https://github.com/llvm/llvm-project/pull/208176