[llvm] [AMDGPU] Adjust amdgpu fmax/fmin legalization (PR #202917)

[Matt Arsenault via llvm-commits]

================
@@ -2075,7 +2076,6 @@ def FeatureISAVersion12 : FeatureSet<
    FeatureFP8ConversionInsts,
    FeatureWMMA128bInsts,
    FeatureSWMMACGfx1200Insts,
-   FeatureIEEEMinimumMaximumInsts,
----------------
arsenm wrote:

Why are you removing this? Changing the targets is a separate change in any case 

https://github.com/llvm/llvm-project/pull/202917