[llvm] [AMDGPU] Prevent FMINIMUM and FMAXIMUM beeing fully scalarized (PR #91378)

[Matt Arsenault via llvm-commits]

https://github.com/arsenm approved this pull request.

lgtm but we're not getting the undef high element to scalar operation optimization 

https://github.com/llvm/llvm-project/pull/91378