AMDGPU/GlobalISel: Select G_SUB

[arsenm]

Attach right version

[arsenm]

ping

[tstellar]

LGTM.

[arsenm]

r365484

lib/Target/AMDGPU/AMDGPUInstructionSelector.h

private:
   MachineOperand getSubOperand64(MachineOperand &MO,
                                  const TargetRegisterClass &SubRC,
                                  unsigned SubIdx) const;
   bool selectCOPY(MachineInstr &I) const;
   bool selectPHI(MachineInstr &I) const;
   bool selectG_TRUNC(MachineInstr &I) const;
   bool selectG_SZA_EXT(MachineInstr &I) const;
   bool selectG_CONSTANT(MachineInstr &I) const;
-  bool selectG_ADD(MachineInstr &I) const;
+  bool selectG_ADD_SUB(MachineInstr &I) const;
   bool selectG_EXTRACT(MachineInstr &I) const;
   bool selectG_MERGE_VALUES(MachineInstr &I) const;
   bool selectG_UNMERGE_VALUES(MachineInstr &I) const;
   bool selectG_GEP(MachineInstr &I) const;
   bool selectG_IMPLICIT_DEF(MachineInstr &I) const;
   bool selectG_INSERT(MachineInstr &I) const;
   bool selectG_INTRINSIC(MachineInstr &I, CodeGenCoverage &CoverageInfo) const;
   bool selectG_INTRINSIC_W_SIDE_EFFECTS(MachineInstr &I,

lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp

case AMDGPU::sub1:
     return MachineOperand::CreateImm(Imm.getHiBits(32).getSExtValue());
   }
 }
 
 static int64_t getConstant(const MachineInstr *MI) {
   return MI->getOperand(1).getCImm()->getSExtValue();
 }
 
-bool AMDGPUInstructionSelector::selectG_ADD(MachineInstr &I) const {
+bool AMDGPUInstructionSelector::selectG_ADD_SUB(MachineInstr &I) const {
   MachineBasicBlock *BB = I.getParent();
   MachineFunction *MF = BB->getParent();
   MachineRegisterInfo &MRI = MF->getRegInfo();
   Register DstReg = I.getOperand(0).getReg();
   const DebugLoc &DL = I.getDebugLoc();
   unsigned Size = RBI.getSizeInBits(DstReg, MRI, TRI);
   const RegisterBank *DstRB = RBI.getRegBank(DstReg, MRI, TRI);
   const bool IsSALU = DstRB->getID() == AMDGPU::SGPRRegBankID;
+  const bool Sub = I.getOpcode() == TargetOpcode::G_SUB;
 
   if (Size == 32) {
     if (IsSALU) {
+      const unsigned Opc = Sub ? AMDGPU::S_SUB_U32 : AMDGPU::S_ADD_U32;
       MachineInstr *Add =
-        BuildMI(*BB, &I, DL, TII.get(AMDGPU::S_ADD_U32), DstReg)
+        BuildMI(*BB, &I, DL, TII.get(Opc), DstReg)
         .add(I.getOperand(1))
         .add(I.getOperand(2));
       I.eraseFromParent();
       return constrainSelectedInstRegOperands(*Add, TII, TRI, RBI);
     }
 
     if (STI.hasAddNoCarry()) {
-      I.setDesc(TII.get(AMDGPU::V_ADD_U32_e64));
+      const unsigned Opc = Sub ? AMDGPU::V_SUB_U32_e64 : AMDGPU::V_ADD_U32_e64;
+      I.setDesc(TII.get(Opc));
       I.addOperand(*MF, MachineOperand::CreateImm(0));
       I.addOperand(*MF, MachineOperand::CreateReg(AMDGPU::EXEC, false, true));
       return constrainSelectedInstRegOperands(I, TII, TRI, RBI);
     }
 
+    const unsigned Opc = Sub ? AMDGPU::V_SUB_I32_e64 : AMDGPU::V_ADD_I32_e64;
+
     Register UnusedCarry = MRI.createVirtualRegister(TRI.getWaveMaskRegClass());
     MachineInstr *Add
-      = BuildMI(*BB, &I, DL, TII.get(AMDGPU::V_ADD_I32_e64), DstReg)
+      = BuildMI(*BB, &I, DL, TII.get(Opc), DstReg)
       .addDef(UnusedCarry, RegState::Dead)
       .add(I.getOperand(1))
       .add(I.getOperand(2))
       .addImm(0);
     I.eraseFromParent();
     return constrainSelectedInstRegOperands(*Add, TII, TRI, RBI);
   }
 
+  assert(!Sub && "illegal sub should not reach here");
+
   const TargetRegisterClass &RC
     = IsSALU ? AMDGPU::SReg_64_XEXECRegClass : AMDGPU::VReg_64RegClass;
   const TargetRegisterClass &HalfRC
     = IsSALU ? AMDGPU::SReg_32RegClass : AMDGPU::VGPR_32RegClass;
 
   MachineOperand Lo1(getSubOperand64(I.getOperand(1), HalfRC, AMDGPU::sub0));
   MachineOperand Lo2(getSubOperand64(I.getOperand(2), HalfRC, AMDGPU::sub0));
   MachineOperand Hi1(getSubOperand64(I.getOperand(1), HalfRC, AMDGPU::sub1));
if (DstRC && !RBI.constrainGenericRegister(Dst.getReg(), *DstRC, MRI))
       return false;
   }
 
   MI.eraseFromParent();
   return true;
 }
 
 bool AMDGPUInstructionSelector::selectG_GEP(MachineInstr &I) const {
-  return selectG_ADD(I);
+  return selectG_ADD_SUB(I);
 }
 
 bool AMDGPUInstructionSelector::selectG_IMPLICIT_DEF(MachineInstr &I) const {
   MachineBasicBlock *BB = I.getParent();
   MachineFunction *MF = BB->getParent();
   MachineRegisterInfo &MRI = MF->getRegInfo();
   const MachineOperand &MO = I.getOperand(0);
 
bool AMDGPUInstructionSelector::select(MachineInstr &I,
   if (!isPreISelGenericOpcode(I.getOpcode())) {
     if (I.isCopy())
       return selectCOPY(I);
     return true;
   }
 
   switch (I.getOpcode()) {
   case TargetOpcode::G_ADD:
-    if (selectG_ADD(I))
+  case TargetOpcode::G_SUB:
+    if (selectG_ADD_SUB(I))
       return true;
     LLVM_FALLTHROUGH;
   default:
     return selectImpl(I, CoverageInfo);
   case TargetOpcode::G_INTTOPTR:
   case TargetOpcode::G_BITCAST:
     return selectCOPY(I);
   case TargetOpcode::G_CONSTANT:

test/CodeGen/AMDGPU/GlobalISel/inst-select-sub.mir

+# NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
+# RUN: llc -march=amdgcn -mcpu=tahiti -run-pass=instruction-select -verify-machineinstrs -o - %s  | FileCheck -check-prefix=GFX6 %s
+# RUN: llc -march=amdgcn -mcpu=fiji -run-pass=instruction-select -verify-machineinstrs -o - %s  | FileCheck -check-prefix=GFX6 %s
+# RUN: llc -march=amdgcn -mcpu=gfx900 -run-pass=instruction-select -verify-machineinstrs -o - %s  | FileCheck -check-prefix=GFX9 %s
+# RUN: llc -march=amdgcn -mcpu=gfx1010 -run-pass=instruction-select -verify-machineinstrs -o - %s  | FileCheck -check-prefix=GFX9 %s
+
+---
+name:            sub_s32
+legalized:       true
+regBankSelected: true
+tracksRegLiveness: true
+
+body: |
+  bb.0:
+    liveins: $sgpr0, $sgpr1, $vgpr0, $vgpr3_vgpr4
+
+
+    ; GFX6-LABEL: name: sub_s32
+    ; GFX6: liveins: $sgpr0, $sgpr1, $vgpr0, $vgpr3_vgpr4
+    ; GFX6: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr0
+    ; GFX6: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr1
+    ; GFX6: [[COPY2:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+    ; GFX6: [[COPY3:%[0-9]+]]:vreg_64 = COPY $vgpr3_vgpr4
+    ; GFX6: [[S_SUB_U32_:%[0-9]+]]:sreg_32_xm0 = S_SUB_U32 [[COPY]], [[COPY1]], implicit-def $scc
+    ; GFX6: %7:vgpr_32, dead %12:sreg_64_xexec = V_SUB_I32_e64 [[COPY2]], [[S_SUB_U32_]], 0, implicit $exec
+    ; GFX6: %8:vgpr_32, dead %11:sreg_64_xexec = V_SUB_I32_e64 [[S_SUB_U32_]], %7, 0, implicit $exec
+    ; GFX6: %9:vgpr_32, dead %10:sreg_64_xexec = V_SUB_I32_e64 %8, [[COPY2]], 0, implicit $exec
+    ; GFX6: FLAT_STORE_DWORD [[COPY3]], %9, 0, 0, 0, 0, implicit $exec, implicit $flat_scr
+    ; GFX9-LABEL: name: sub_s32
+    ; GFX9: liveins: $sgpr0, $sgpr1, $vgpr0, $vgpr3_vgpr4
+    ; GFX9: [[COPY:%[0-9]+]]:sreg_32 = COPY $sgpr0
+    ; GFX9: [[COPY1:%[0-9]+]]:sreg_32 = COPY $sgpr1
+    ; GFX9: [[COPY2:%[0-9]+]]:vgpr_32 = COPY $vgpr0
+    ; GFX9: [[COPY3:%[0-9]+]]:vreg_64 = COPY $vgpr3_vgpr4
+    ; GFX9: [[S_SUB_U32_:%[0-9]+]]:sreg_32_xm0 = S_SUB_U32 [[COPY]], [[COPY1]], implicit-def $scc
+    ; GFX9: [[V_SUB_U32_e64_:%[0-9]+]]:vgpr_32 = V_SUB_U32_e64 [[COPY2]], [[S_SUB_U32_]], 0, implicit $exec
+    ; GFX9: [[V_SUB_U32_e64_1:%[0-9]+]]:vgpr_32 = V_SUB_U32_e64 [[S_SUB_U32_]], [[V_SUB_U32_e64_]], 0, implicit $exec
+    ; GFX9: [[V_SUB_U32_e64_2:%[0-9]+]]:vgpr_32 = V_SUB_U32_e64 [[V_SUB_U32_e64_1]], [[COPY2]], 0, implicit $exec
+    ; GFX9: FLAT_STORE_DWORD [[COPY3]], [[V_SUB_U32_e64_2]], 0, 0, 0, 0, implicit $exec, implicit $flat_scr
+    %0:sgpr(s32) = COPY $sgpr0
+    %1:sgpr(s32) = COPY $sgpr1
+    %2:vgpr(s32) = COPY $vgpr0
+    %3:vgpr(p1) = COPY $vgpr3_vgpr4
+    %4:sgpr(s32) = G_CONSTANT i32 1
+    %5:sgpr(s32) = G_CONSTANT i32 4096
+
+    ; sub ss
+    %6:sgpr(s32) = G_SUB %0, %1
+
+    ; sub vs
+    %7:vgpr(s32) = G_SUB %2, %6
+
+    ; sub sv
+    %8:vgpr(s32) = G_SUB %6, %7
+
+    ; sub vv
+    %9:vgpr(s32) = G_SUB %8, %2
+
+    G_STORE %9, %3 :: (store 4, addrspace 1)
+
+...