[llvm-branch-commits] [llvm] [AMDGPU] Add wave reduce intrinsics for double types - 2 (PR #170812)

[Matt Arsenault via llvm-branch-commits]

================
@@ -5755,28 +5761,71 @@ static MachineBasicBlock *lowerWaveReduce(MachineInstr &MI,
         break;
       }
       case AMDGPU::V_ADD_F32_e64:
+      case AMDGPU::V_ADD_F64_e64:
+      case AMDGPU::V_ADD_F64_e64_gfx12:
----------------
arsenm wrote:

This is the same as the V_MUL_F64 case. Selection doesn't emit the encoded instruction, gfx12 emits V_ADD_F64_pseudo_e64

https://github.com/llvm/llvm-project/pull/170812