[GlobalISel][X86] Add s8/s16/s64 handling of G_UADDE opcodes

llvm · Jun 16, 2023 · cad1eef · cad1eef
1 parent 835cdcb
commit cad1eef
Show file tree

Hide file tree

Showing 3 changed files with 42 additions and 19 deletions.
diff --git a/llvm/lib/Target/X86/X86InstructionSelector.cpp b/llvm/lib/Target/X86/X86InstructionSelector.cpp
@@ -1081,9 +1081,33 @@ bool X86InstructionSelector::selectUadde(MachineInstr &I,
   Register CarryInReg = I.getOperand(4).getReg();
 
   const LLT DstTy = MRI.getType(DstReg);
+  assert(DstTy.isScalar() && "G_UADDE only supported for scalar types");
 
-  if (DstTy != LLT::scalar(32))
-    return false;
+  // TODO: Handle immediate argument variants?
+  unsigned OpADC, OpADD;
+  switch (DstTy.getSizeInBits()) {
+  case 8:
+    OpADC = X86::ADC8rr;
+    OpADD = X86::ADD8rr;
+    break;
+  case 16:
+    OpADC = X86::ADC16rr;
+    OpADD = X86::ADD16rr;
+    break;
+  case 32:
+    OpADC = X86::ADC32rr;
+    OpADD = X86::ADD32rr;
+    break;
+  case 64:
+    OpADC = X86::ADC64rr;
+    OpADD = X86::ADD64rr;
+    break;
+  default:
+    llvm_unreachable("Can't select G_UADDE, unsupported type.");
+  }
+
+  const RegisterBank &DstRB = *RBI.getRegBank(DstReg, MRI, TRI);
+  const TargetRegisterClass *DstRC = getRegClass(DstTy, DstRB);
 
   // find CarryIn def instruction.
   MachineInstr *Def = MRI.getVRegDef(CarryInReg);
@@ -1092,23 +1116,23 @@ bool X86InstructionSelector::selectUadde(MachineInstr &I,
     Def = MRI.getVRegDef(CarryInReg);
   }
 
-  unsigned Opcode;
+  unsigned Opcode = 0;
   if (Def->getOpcode() == TargetOpcode::G_UADDE) {
     // carry set by prev ADD.
 
     BuildMI(*I.getParent(), I, I.getDebugLoc(), TII.get(X86::COPY), X86::EFLAGS)
         .addReg(CarryInReg);
 
-    if (!RBI.constrainGenericRegister(CarryInReg, X86::GR32RegClass, MRI))
+    if (!RBI.constrainGenericRegister(CarryInReg, *DstRC, MRI))
       return false;
 
-    Opcode = X86::ADC32rr;
+    Opcode = OpADC;
   } else if (auto val = getIConstantVRegVal(CarryInReg, MRI)) {
     // carry is constant, support only 0.
     if (*val != 0)
       return false;
 
-    Opcode = X86::ADD32rr;
+    Opcode = OpADD;
   } else
     return false;
 
@@ -1121,7 +1145,7 @@ bool X86InstructionSelector::selectUadde(MachineInstr &I,
       .addReg(X86::EFLAGS);
 
   if (!constrainSelectedInstRegOperands(AddInst, TII, TRI, RBI) ||
-      !RBI.constrainGenericRegister(CarryOutReg, X86::GR32RegClass, MRI))
+      !RBI.constrainGenericRegister(CarryOutReg, *DstRC, MRI))
     return false;
 
   I.eraseFromParent();

diff --git a/llvm/lib/Target/X86/X86LegalizerInfo.cpp b/llvm/lib/Target/X86/X86LegalizerInfo.cpp
@@ -148,12 +148,15 @@ X86LegalizerInfo::X86LegalizerInfo(const X86Subtarget &STI,
       .clampScalar(0, s8, sMaxScalar)
       .scalarize(0);
 
-  // TODO: Add all legal scalar types.
   // TODO: Add G_UADDO/G_USUBO/G_USUBE handling
   getActionDefinitionsBuilder(G_UADDE)
-      .legalFor({{s32, s1}})
+      .legalIf([=](const LegalityQuery &Query) -> bool {
+        return typePairInSet(0, 1, {{s8, s1}, {s16, s1}, {s32, s1}})(Query) ||
+               (Is64Bit && typePairInSet(0, 1, {{s64, s1}})(Query));
+      })
       .widenScalarToNextPow2(0, /*Min=*/32)
-      .clampScalar(0, s32, s32)
+      .clampScalar(0, s8, sMaxScalar)
+      .clampScalar(1, s1, s1)
       .scalarize(0);
 
   // integer multiply

diff --git a/llvm/test/CodeGen/X86/GlobalISel/legalize-add.mir b/llvm/test/CodeGen/X86/GlobalISel/legalize-add.mir
@@ -232,15 +232,11 @@ body:             |
     ; X64-NEXT: [[UV:%[0-9]+]]:_(s64), [[UV1:%[0-9]+]]:_(s64) = G_UNMERGE_VALUES [[DEF]](s128)
     ; X64-NEXT: [[UV2:%[0-9]+]]:_(s64), [[UV3:%[0-9]+]]:_(s64) = G_UNMERGE_VALUES [[DEF1]](s128)
     ; X64-NEXT: [[UADDO:%[0-9]+]]:_(s64), [[UADDO1:%[0-9]+]]:_(s1) = G_UADDO [[UV]], [[UV2]]
-    ; X64-NEXT: [[UV4:%[0-9]+]]:_(s32), [[UV5:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[UV1]](s64)
-    ; X64-NEXT: [[UV6:%[0-9]+]]:_(s32), [[UV7:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[UV3]](s64)
-    ; X64-NEXT: [[UADDE:%[0-9]+]]:_(s32), [[UADDE1:%[0-9]+]]:_(s1) = G_UADDE [[UV4]], [[UV6]], [[UADDO1]]
-    ; X64-NEXT: [[UADDE2:%[0-9]+]]:_(s32), [[UADDE3:%[0-9]+]]:_(s1) = G_UADDE [[UV5]], [[UV7]], [[UADDE1]]
-    ; X64-NEXT: [[MV:%[0-9]+]]:_(s64) = G_MERGE_VALUES [[UADDE]](s32), [[UADDE2]](s32)
-    ; X64-NEXT: [[MV1:%[0-9]+]]:_(s128) = G_MERGE_VALUES [[UADDO]](s64), [[MV]](s64)
-    ; X64-NEXT: [[UV8:%[0-9]+]]:_(s64), [[UV9:%[0-9]+]]:_(s64) = G_UNMERGE_VALUES [[MV1]](s128)
-    ; X64-NEXT: $rax = COPY [[UV8]](s64)
-    ; X64-NEXT: $rdx = COPY [[UV9]](s64)
+    ; X64-NEXT: [[UADDE:%[0-9]+]]:_(s64), [[UADDE1:%[0-9]+]]:_(s1) = G_UADDE [[UV1]], [[UV3]], [[UADDO1]]
+    ; X64-NEXT: [[MV:%[0-9]+]]:_(s128) = G_MERGE_VALUES [[UADDO]](s64), [[UADDE]](s64)
+    ; X64-NEXT: [[UV4:%[0-9]+]]:_(s64), [[UV5:%[0-9]+]]:_(s64) = G_UNMERGE_VALUES [[MV]](s128)
+    ; X64-NEXT: $rax = COPY [[UV4]](s64)
+    ; X64-NEXT: $rdx = COPY [[UV5]](s64)
     ; X64-NEXT: RET 0
     ; X32-LABEL: name: test_add_i128
     ; X32: [[DEF:%[0-9]+]]:_(s128) = IMPLICIT_DEF