[llvm] PowerPC/VSX: Select FMINNUM_IEEE and FMAXNUM_IEEE (PR #112195)

[Matt Arsenault via llvm-commits]

================
@@ -2735,39 +2744,23 @@ def : Pat<(f32 (fneg (fabs f32:$S))),
 // f32 Min.
 def : Pat<(f32 (fminnum_ieee f32:$A, f32:$B)),
           (f32 FpMinMax.F32Min)>;
-def : Pat<(f32 (fminnum_ieee (fcanonicalize f32:$A), f32:$B)),
-          (f32 FpMinMax.F32Min)>;
-def : Pat<(f32 (fminnum_ieee f32:$A, (fcanonicalize f32:$B))),
-          (f32 FpMinMax.F32Min)>;
----------------
arsenm wrote:

Don't understand these patterns folding out a canonicalize on each individual operand. If FpMinMax.F32Min has the IEEE snan behavior, both need to be guaranteed quiet. Can these be removed as a precommit? 

https://github.com/llvm/llvm-project/pull/112195