[llvm-branch-commits] [llvm] [AMDGPU] Extending wave reduction intrinsics for `i64` types - 2 (PR #151309)

[via llvm-branch-commits]

================
@@ -5207,6 +5212,74 @@ static MachineBasicBlock *lowerWaveReduce(MachineInstr &MI,
             .addReg(NewAccumulator->getOperand(0).getReg());
         break;
       }
+      case AMDGPU::S_ADD_U64_PSEUDO:
+      case AMDGPU::S_SUB_U64_PSEUDO: {
+        Register DestSub0 = MRI.createVirtualRegister(&AMDGPU::SReg_32RegClass);
+        Register DestSub1 = MRI.createVirtualRegister(&AMDGPU::SReg_32RegClass);
+        Register Op1H_Op0L_Reg =
+            MRI.createVirtualRegister(&AMDGPU::SReg_32RegClass);
+        Register Op1L_Op0H_Reg =
+            MRI.createVirtualRegister(&AMDGPU::SReg_32RegClass);
+        Register CarryReg = MRI.createVirtualRegister(&AMDGPU::SReg_32RegClass);
+        Register AddReg = MRI.createVirtualRegister(&AMDGPU::SReg_32RegClass);
+        Register NegatedValLo =
+            MRI.createVirtualRegister(&AMDGPU::SReg_32RegClass);
+        Register NegatedValHi =
+            MRI.createVirtualRegister(&AMDGPU::SReg_32RegClass);
+
+        const TargetRegisterClass *Src1RC = MRI.getRegClass(SrcReg);
+        const TargetRegisterClass *Src1SubRC =
+            TRI->getSubRegisterClass(Src1RC, AMDGPU::sub0);
+
+        MachineOperand Op1L = TII->buildExtractSubRegOrImm(
+            MI, MRI, MI.getOperand(1), Src1RC, AMDGPU::sub0, Src1SubRC);
+        MachineOperand Op1H = TII->buildExtractSubRegOrImm(
+            MI, MRI, MI.getOperand(1), Src1RC, AMDGPU::sub1, Src1SubRC);
+
+        if (Opc == AMDGPU::S_SUB_U64_PSEUDO) {
+          BuildMI(BB, MI, DL, TII->get(AMDGPU::S_SUB_I32), NegatedValLo)
+              .addImm(0)
+              .addReg(NewAccumulator->getOperand(0).getReg());
+          BuildMI(BB, MI, DL, TII->get(AMDGPU::S_ASHR_I32), NegatedValHi)
+              .addReg(NegatedValLo)
+              .addImm(31)
+              .setOperandDead(3); // Dead scc
+          BuildMI(BB, MI, DL, TII->get(AMDGPU::S_MUL_I32), Op1L_Op0H_Reg)
+              .add(Op1L)
+              .addReg(NegatedValHi);
+        }
+        Register LowOpcode = Opc == AMDGPU::S_SUB_U64_PSEUDO
+                                 ? NegatedValLo
+                                 : NewAccumulator->getOperand(0).getReg();
+        BuildMI(BB, MI, DL, TII->get(AMDGPU::S_MUL_I32), DestSub0)
----------------
easyonaadit wrote:

I tried this, but the 64-bit pseudos weren't getting expanded. I don't think it's possible to replace one pseudo with another in the expand-pseudo pass.  

https://github.com/llvm/llvm-project/pull/151309