[llvm] expandFMINIMUMNUM_FMAXIMUMNUM: Quiet is not needed for NaN vs NaN (PR #139237)

[Matt Arsenault via llvm-commits]

================
@@ -1713,25 +1713,15 @@ define bfloat @v_min3_bf16_minimumnum_minimumnum__v_v_v_0(bfloat %a, bfloat %b,
 ; GFX8-NEXT:    v_lshlrev_b32_e32 v3, 16, v1
 ; GFX8-NEXT:    v_lshlrev_b32_e32 v4, 16, v0
 ; GFX8-NEXT:    v_cmp_lt_f32_e32 vcc, v4, v3
-; GFX8-NEXT:    v_cndmask_b32_e32 v3, v1, v0, vcc
-; GFX8-NEXT:    v_lshlrev_b32_e32 v3, 16, v3
-; GFX8-NEXT:    v_mul_f32_e32 v3, 1.0, v3
-; GFX8-NEXT:    v_bfe_u32 v4, v3, 16, 1
-; GFX8-NEXT:    v_add_u32_e32 v4, vcc, v4, v3
-; GFX8-NEXT:    s_movk_i32 s4, 0x7fff
-; GFX8-NEXT:    v_add_u32_e32 v4, vcc, s4, v4
-; GFX8-NEXT:    v_or_b32_e32 v5, 0x400000, v3
-; GFX8-NEXT:    v_cmp_u_f32_e32 vcc, v3, v3
-; GFX8-NEXT:    v_cndmask_b32_e32 v3, v4, v5, vcc
 ; GFX8-NEXT:    s_movk_i32 s4, 0x8000
-; GFX8-NEXT:    v_lshrrev_b32_e32 v4, 16, v3
+; GFX8-NEXT:    v_cndmask_b32_e32 v3, v1, v0, vcc
----------------
arsenm wrote:

I'm surprised how big the diff is here. I'm also surprised AMDGPU is going through this path for any type, bfloat should have promoted to float? 

https://github.com/llvm/llvm-project/pull/139237