[llvm-branch-commits] [llvm] AMDGPU: Create pseudo to real mapping for flat/buffer atomic fmin/fmax (PR #95591)

[Matt Arsenault via llvm-branch-commits]

================
@@ -1608,14 +1598,14 @@ defm : FlatSignedAtomicIntrPat <"FLAT_ATOMIC_FMAX", "int_amdgcn_flat_atomic_fmax
 }
 
 let OtherPredicates = [isGFX10Only] in {
-defm : GlobalFLATAtomicPats <"GLOBAL_ATOMIC_FMIN_X2", "atomic_load_fmin_global", f64>;
-defm : GlobalFLATAtomicPats <"GLOBAL_ATOMIC_FMAX_X2", "atomic_load_fmax_global", f64>;
-defm : GlobalFLATAtomicIntrPats <"GLOBAL_ATOMIC_FMIN_X2", "int_amdgcn_global_atomic_fmin", f64>;
-defm : GlobalFLATAtomicIntrPats <"GLOBAL_ATOMIC_FMAX_X2", "int_amdgcn_global_atomic_fmax", f64>;
-defm : FlatSignedAtomicPat <"FLAT_ATOMIC_FMIN_X2", "atomic_load_fmin_flat", f64>;
-defm : FlatSignedAtomicPat <"FLAT_ATOMIC_FMAX_X2", "atomic_load_fmax_flat", f64>;
-defm : FlatSignedAtomicIntrPat <"FLAT_ATOMIC_FMIN_X2", "int_amdgcn_flat_atomic_fmin", f64>;
-defm : FlatSignedAtomicIntrPat <"FLAT_ATOMIC_FMAX_X2", "int_amdgcn_flat_atomic_fmax", f64>;
+defm : GlobalFLATAtomicPats <"GLOBAL_ATOMIC_MIN_F64", "atomic_load_fmin_global", f64>;
----------------
arsenm wrote:

Yes, there's a lot of predicate junk here that needs to be cleaned up. Most of these atomics are imprecisely guarded with isGFX10/11/12 checks that are available on a subset of gfx9 

https://github.com/llvm/llvm-project/pull/95591