[AArch64][GlobalISel] Support the neg_addsub_shifted_imm32 pattern

[paquette]

Summary

Add an equivalent ComplexRendererFns function for SelectNegArithImmed. This allows us to select immediate adds of -1 by turning them into subtracts.

Update select-binop.mir to show that the pattern works.

[aemerson]

LGTM.

llvm/trunk/lib/Target/AArch64/AArch64InstrFormats.td

class neg_addsub_shifted_imm<ValueType Ty>
   let EncoderMethod = "getAddSubImmOpValue";
   let ParserMatchClass = AddSubImmOperand;
   let MIOperandInfo = (ops i32imm, i32imm);
 }
 
 def neg_addsub_shifted_imm32 : neg_addsub_shifted_imm<i32>;
 def neg_addsub_shifted_imm64 : neg_addsub_shifted_imm<i64>;
 
+def gi_neg_addsub_shifted_imm32 :
+    GIComplexOperandMatcher<s32, "selectNegArithImmed">,
+    GIComplexPatternEquiv<neg_addsub_shifted_imm32>;
+
+def gi_neg_addsub_shifted_imm64 :
+    GIComplexOperandMatcher<s64, "selectNegArithImmed">,
+    GIComplexPatternEquiv<neg_addsub_shifted_imm64>;
+
 // An extend operand:
 //  {5-3} - extend type
 //  {2-0} - imm3
 def arith_extend : Operand<i32> {
   let PrintMethod = "printArithExtend";
   let ParserMatchClass = ExtendOperand;
 }
 def arith_extend64 : Operand<i32> {

llvm/trunk/lib/Target/AArch64/AArch64InstructionSelector.cpp

private:
   // Equivalent to the i32shift_a and friends from AArch64InstrInfo.td.
   // We use these manually instead of using the importer since it doesn't
   // support SDNodeXForm.
   ComplexRendererFns selectShiftA_32(const MachineOperand &Root) const;
   ComplexRendererFns selectShiftB_32(const MachineOperand &Root) const;
   ComplexRendererFns selectShiftA_64(const MachineOperand &Root) const;
   ComplexRendererFns selectShiftB_64(const MachineOperand &Root) const;
 
+  ComplexRendererFns select12BitValueWithLeftShift(uint64_t Immed) const;
   ComplexRendererFns selectArithImmed(MachineOperand &Root) const;
+  ComplexRendererFns selectNegArithImmed(MachineOperand &Root) const;
 
   ComplexRendererFns selectAddrModeUnscaled(MachineOperand &Root,
                                             unsigned Size) const;
 
   ComplexRendererFns selectAddrModeUnscaled8(MachineOperand &Root) const {
     return selectAddrModeUnscaled(Root, 1);
   }
   ComplexRendererFns selectAddrModeUnscaled16(MachineOperand &Root) const {

 AArch64InstructionSelector::selectShiftB_64(const MachineOperand &Root) const {
   auto MaybeImmed = getImmedFromMO(Root);
   if (MaybeImmed == None || *MaybeImmed > 63)
     return None;
   uint64_t Enc = 63 - *MaybeImmed;
   return {{[=](MachineInstrBuilder &MIB) { MIB.addImm(Enc); }}};
 }
 
+/// Helper to select an immediate value that can be represented as a 12-bit
+/// value shifted left by either 0 or 12. If it is possible to do so, return
+/// the immediate and shift value. If not, return None.
+///
+/// Used by selectArithImmed and selectNegArithImmed.
+InstructionSelector::ComplexRendererFns
+AArch64InstructionSelector::select12BitValueWithLeftShift(
+    uint64_t Immed) const {
+  unsigned ShiftAmt;
+  if (Immed >> 12 == 0) {
+    ShiftAmt = 0;
+  } else if ((Immed & 0xfff) == 0 && Immed >> 24 == 0) {
+    ShiftAmt = 12;
+    Immed = Immed >> 12;
+  } else
+    return None;
+
+  unsigned ShVal = AArch64_AM::getShifterImm(AArch64_AM::LSL, ShiftAmt);
+  return {{
+      [=](MachineInstrBuilder &MIB) { MIB.addImm(Immed); },
+      [=](MachineInstrBuilder &MIB) { MIB.addImm(ShVal); },
+  }};
+}
+
 /// SelectArithImmed - Select an immediate value that can be represented as
 /// a 12-bit value shifted left by either 0 or 12.  If so, return true with
 /// Val set to the 12-bit value and Shift set to the shifter operand.
 InstructionSelector::ComplexRendererFns
 AArch64InstructionSelector::selectArithImmed(MachineOperand &Root) const {
   // This function is called from the addsub_shifted_imm ComplexPattern,
   // which lists [imm] as the list of opcode it's interested in, however
   // we still need to check whether the operand is actually an immediate
   // here because the ComplexPattern opcode list is only used in
   // root-level opcode matching.
   auto MaybeImmed = getImmedFromMO(Root);
   if (MaybeImmed == None)
     return None;
+  return select12BitValueWithLeftShift(*MaybeImmed);
+}
+
+/// SelectNegArithImmed - As above, but negates the value before trying to
+/// select it.
+InstructionSelector::ComplexRendererFns
+AArch64InstructionSelector::selectNegArithImmed(MachineOperand &Root) const {
+  // We need a register here, because we need to know if we have a 64 or 32
+  // bit immediate.
+  if (!Root.isReg())
+    return None;
+  auto MaybeImmed = getImmedFromMO(Root);
+  if (MaybeImmed == None)
+    return None;
   uint64_t Immed = *MaybeImmed;
-  unsigned ShiftAmt;
 
-  if (Immed >> 12 == 0) {
-    ShiftAmt = 0;
-  } else if ((Immed & 0xfff) == 0 && Immed >> 24 == 0) {
-    ShiftAmt = 12;
-    Immed = Immed >> 12;
-  } else
+  // This negation is almost always valid, but "cmp wN, #0" and "cmn wN, #0"
+  // have the opposite effect on the C flag, so this pattern mustn't match under
+  // those circumstances.
+  if (Immed == 0)
     return None;
 
-  unsigned ShVal = AArch64_AM::getShifterImm(AArch64_AM::LSL, ShiftAmt);
-  return {{
-      [=](MachineInstrBuilder &MIB) { MIB.addImm(Immed); },
-      [=](MachineInstrBuilder &MIB) { MIB.addImm(ShVal); },
-  }};
+  // Check if we're dealing with a 32-bit type on the root or a 64-bit type on
+  // the root.
+  MachineRegisterInfo &MRI = Root.getParent()->getMF()->getRegInfo();
+  if (MRI.getType(Root.getReg()).getSizeInBits() == 32)
+    Immed = ~((uint32_t)Immed) + 1;
+  else
+    Immed = ~Immed + 1ULL;
+
+  if (Immed & 0xFFFFFFFFFF000000ULL)
+    return None;
+
+  Immed &= 0xFFFFFFULL;
+  return select12BitValueWithLeftShift(Immed);
 }
 
 /// Return true if it is worth folding MI into an extended register. That is,
 /// if it's safe to pull it into the addressing mode of a load or store as a
 /// shift.
 bool AArch64InstructionSelector::isWorthFoldingIntoExtendedReg(
     MachineInstr &MI, const MachineRegisterInfo &MRI) const {
   // Always fold if there is one use, or if we're optimizing for size.

llvm/trunk/test/CodeGen/AArch64/GlobalISel/select-binop.mir

 # NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
 # RUN: llc -mtriple=aarch64-- -run-pass=instruction-select -verify-machineinstrs %s -o - | FileCheck %s
 
 --- |
   target datalayout = "e-m:o-i64:64-i128:128-n32:64-S128"
 
   define void @add_s32_gpr() { ret void }
   define void @add_s64_gpr() { ret void }
 
   define void @add_imm_s32_gpr() { ret void }
   define void @add_imm_s64_gpr() { ret void }
 
+  define void @add_neg_s32_gpr() { ret void }
+  define void @add_neg_s64_gpr() { ret void }
+  define void @add_neg_invalid_immed_s32() { ret void }
+  define void @add_neg_invalid_immed_s64() { ret void }
+  define void @add_imm_0_s32() { ret void }
+  define void @add_imm_0_s64() { ret void }
+
   define void @add_imm_s32_gpr_bb() { ret void }
 
   define void @sub_s32_gpr() { ret void }
   define void @sub_s64_gpr() { ret void }
 
   define void @or_s32_gpr() { ret void }
   define void @or_s64_gpr() { ret void }
   define void @or_v2s32_fpr() { ret void }
bb.0:
     ; CHECK: $x0 = COPY [[ADDXri]]
     %0(s64) = COPY $x0
     %1(s64) = G_CONSTANT i64 1
     %2(s64) = G_ADD %0, %1
     $x0 = COPY %2(s64)
 ...
 
 ---
+name:            add_neg_s32_gpr
+legalized:       true
+regBankSelected: true
+
+registers:
+  - { id: 0, class: gpr }
+  - { id: 1, class: gpr }
+  - { id: 2, class: gpr }
+
+body:             |
+  bb.0:
+    liveins: $w1, $w2
+    ; We should be able to turn the ADD into a SUB.
+    ; CHECK-LABEL: name: add_neg_s32_gpr
+    ; CHECK: [[COPY:%[0-9]+]]:gpr32sp = COPY $w1
+    ; CHECK: [[SUBSWri:%[0-9]+]]:gpr32 = SUBSWri [[COPY]], 1, 0, implicit-def $nzcv
+    ; CHECK: $w2 = COPY [[SUBSWri]]
+    %0(s32) = COPY $w1
+    %1(s32) = G_CONSTANT i32 -1
+    %2(s32) = G_ADD %0, %1
+    $w2 = COPY %2(s32)
+...
+
+---
+name:            add_neg_s64_gpr
+legalized:       true
+regBankSelected: true
+
+registers:
+  - { id: 0, class: gpr }
+  - { id: 1, class: gpr }
+  - { id: 2, class: gpr }
+
+body:             |
+  bb.0:
+    liveins: $x0, $x1
+    ; We should be able to turn the ADD into a SUB.
+    ; CHECK-LABEL: name: add_neg_s64_gpr
+    ; CHECK: [[COPY:%[0-9]+]]:gpr64sp = COPY $x0
+    ; CHECK: [[SUBSXri:%[0-9]+]]:gpr64 = SUBSXri [[COPY]], 1, 0, implicit-def $nzcv
+    ; CHECK: $x0 = COPY [[SUBSXri]]
+    %0(s64) = COPY $x0
+    %1(s64) = G_CONSTANT i64 -1
+    %2(s64) = G_ADD %0, %1
+    $x0 = COPY %2(s64)
+...
+
+---
+name:            add_neg_invalid_immed_s32
+legalized:       true
+regBankSelected: true
+
+registers:
+  - { id: 0, class: gpr }
+  - { id: 1, class: gpr }
+  - { id: 2, class: gpr }
+
+body:             |
+  bb.0:
+    liveins: $x0, $x1
+    ; We can't select this if the value is out of range.
+    ; CHECK-LABEL: name: add_neg_invalid_immed_s32
+    ; CHECK: [[COPY:%[0-9]+]]:gpr64 = COPY $x0
+    ; CHECK: [[MOVi64imm:%[0-9]+]]:gpr64 = MOVi64imm -5000
+    ; CHECK: [[ADDXrr:%[0-9]+]]:gpr64 = ADDXrr [[COPY]], [[MOVi64imm]]
+    ; CHECK: $x0 = COPY [[ADDXrr]]
+    %0(s64) = COPY $x0
+    %1(s64) = G_CONSTANT i64 -5000
+    %2(s64) = G_ADD %0, %1
+    $x0 = COPY %2(s64)
+...
+
+---
+name:            add_neg_invalid_immed_s64
+legalized:       true
+regBankSelected: true
+
+registers:
+  - { id: 0, class: gpr }
+  - { id: 1, class: gpr }
+  - { id: 2, class: gpr }
+
+body:             |
+  bb.0:
+    liveins: $x0, $x1
+    ; We can't select this if the value is out of range.
+    ; CHECK-LABEL: name: add_neg_invalid_immed_s64
+    ; CHECK: [[COPY:%[0-9]+]]:gpr64 = COPY $x0
+    ; CHECK: [[MOVi64imm:%[0-9]+]]:gpr64 = MOVi64imm -5000
+    ; CHECK: [[ADDXrr:%[0-9]+]]:gpr64 = ADDXrr [[COPY]], [[MOVi64imm]]
+    ; CHECK: $x0 = COPY [[ADDXrr]]
+    %0(s64) = COPY $x0
+    %1(s64) = G_CONSTANT i64 -5000
+    %2(s64) = G_ADD %0, %1
+    $x0 = COPY %2(s64)
+...
+
+---
+name:            add_imm_0_s32
+legalized:       true
+regBankSelected: true
+
+registers:
+  - { id: 0, class: gpr }
+  - { id: 1, class: gpr }
+  - { id: 2, class: gpr }
+
+body:             |
+  bb.0:
+    liveins: $x0, $x1
+    ; We shouldn't get a SUB here, because "cmp wN, $0" and "cmp wN, #0" have
+    ; opposite effects on the C flag.
+    ; CHECK-LABEL: name: add_imm_0_s32
+    ; CHECK: [[COPY:%[0-9]+]]:gpr64sp = COPY $x0
+    ; CHECK: [[ADDXri:%[0-9]+]]:gpr64sp = ADDXri [[COPY]], 0, 0
+    ; CHECK: $x0 = COPY [[ADDXri]]
+    %0(s64) = COPY $x0
+    %1(s64) = G_CONSTANT i64 0
+    %2(s64) = G_ADD %0, %1
+    $x0 = COPY %2(s64)
+...
+
+---
+name:            add_imm_0_s64
+legalized:       true
+regBankSelected: true
+
+registers:
+  - { id: 0, class: gpr }
+  - { id: 1, class: gpr }
+  - { id: 2, class: gpr }
+
+body:             |
+  bb.0:
+    liveins: $x0, $x1
+    ; We shouldn't get a SUB here, because "cmp xN, $0" and "cmp xN, #0" have
+    ; opposite effects on the C flag.
+    ; CHECK-LABEL: name: add_imm_0_s64
+    ; CHECK: [[COPY:%[0-9]+]]:gpr64sp = COPY $x0
+    ; CHECK: [[ADDXri:%[0-9]+]]:gpr64sp = ADDXri [[COPY]], 0, 0
+    ; CHECK: $x0 = COPY [[ADDXri]]
+    %0(s64) = COPY $x0
+    %1(s64) = G_CONSTANT i64 0
+    %2(s64) = G_ADD %0, %1
+    $x0 = COPY %2(s64)
+...
+
+---
 name:            add_imm_s32_gpr_bb
 legalized:       true
 regBankSelected: true
 
 registers:
   - { id: 0, class: gpr }
   - { id: 1, class: gpr }
   - { id: 2, class: gpr }