[X86][SelectionDAG] Add SADDO_CARRY and SSUBO_CARRY to support multipart signed add/sub overflow legalization.

[craig.topper]

Summary

This passes existing X86 test but I'm not sure if it handles all type
legalization cases it needs to.

Alternative to D89200

[craig.topper]

Add support for expanding SADDO_CARRY and SSUBO_CARRY in type legalizer.
Default to operation action to Expand.
Rebase after committing tests for i128 on a 32-bit target to cover the second rounding of expanding.

[efriedma]

The title is still "WIP", but I don't see anything obviously missing.

LGTM

[craig.topper]

@ecnelises why does this review show blocked by you? I don't see any comments.

llvm/include/llvm/CodeGen/ISDOpcodes.h

 //===-- llvm/CodeGen/ISDOpcodes.h - CodeGen opcodes -------------*- C++ -*-===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 //
enum NodeType {
   /// is an incoming carry. These nodes produce two results: the normal
   /// result of the add or sub, and the output carry so they can be chained
   /// together. The use of this opcode is preferable to adde/sube if the
   /// target supports it, as the carry is a regular value rather than a
   /// glue, which allows further optimisation.
   ADDCARRY,
   SUBCARRY,
 
+  /// Carry-using overflow-aware nodes for multiple precision addition and
+  /// subtraction. These nodes take three operands: The first two are normal lhs
+  /// and /// rhs to the add or sub, and the third is a boolean indicating if
+  /// there is an incoming carry. They produce two results: the normal result of
+  /// the add or sub, and a boolean that indicates if an overflow occured (*not*
+  /// flag, because it may be a store to memory, etc.). If the type of the
+  /// boolean is not i1 then the high bits conform to getBooleanContents.
+  SADDO_CARRY,
+  SSUBO_CARRY,
+
   /// RESULT, BOOL = [SU]ADDO(LHS, RHS) - Overflow-aware nodes for addition.
   /// These nodes take two operands: the normal LHS and RHS to the add. They
   /// produce two results: the normal result of the add, and a boolean that
   /// indicates if an overflow occurred (*not* a flag, because it may be store
   /// to memory, etc.).  If the type of the boolean is not i1 then the high
   /// bits conform to getBooleanContents.
   /// These nodes are generated from llvm.[su]add.with.overflow intrinsics.
   SADDO,

llvm/lib/CodeGen/SelectionDAG/DAGCombiner.cpp

 //===- DAGCombiner.cpp - Implement a DAG node combiner --------------------===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 //
private:
     SDValue visitSUBSAT(SDNode *N);
     SDValue visitADDC(SDNode *N);
     SDValue visitADDO(SDNode *N);
     SDValue visitUADDOLike(SDValue N0, SDValue N1, SDNode *N);
     SDValue visitSUBC(SDNode *N);
     SDValue visitSUBO(SDNode *N);
     SDValue visitADDE(SDNode *N);
     SDValue visitADDCARRY(SDNode *N);
+    SDValue visitSADDO_CARRY(SDNode *N);
     SDValue visitADDCARRYLike(SDValue N0, SDValue N1, SDValue CarryIn, SDNode *N);
     SDValue visitSUBE(SDNode *N);
     SDValue visitSUBCARRY(SDNode *N);
+    SDValue visitSSUBO_CARRY(SDNode *N);
     SDValue visitMUL(SDNode *N);
     SDValue visitMULFIX(SDNode *N);
     SDValue useDivRem(SDNode *N);
     SDValue visitSDIV(SDNode *N);
     SDValue visitSDIVLike(SDValue N0, SDValue N1, SDNode *N);
     SDValue visitUDIV(SDNode *N);
     SDValue visitUDIVLike(SDValue N0, SDValue N1, SDNode *N);
     SDValue visitREM(SDNode *N);
SDValue DAGCombiner::visit(SDNode *N) {
   case ISD::ADDC:               return visitADDC(N);
   case ISD::SADDO:
   case ISD::UADDO:              return visitADDO(N);
   case ISD::SUBC:               return visitSUBC(N);
   case ISD::SSUBO:
   case ISD::USUBO:              return visitSUBO(N);
   case ISD::ADDE:               return visitADDE(N);
   case ISD::ADDCARRY:           return visitADDCARRY(N);
+  case ISD::SADDO_CARRY:        return visitSADDO_CARRY(N);
   case ISD::SUBE:               return visitSUBE(N);
   case ISD::SUBCARRY:           return visitSUBCARRY(N);
+  case ISD::SSUBO_CARRY:        return visitSSUBO_CARRY(N);
   case ISD::SMULFIX:
   case ISD::SMULFIXSAT:
   case ISD::UMULFIX:
   case ISD::UMULFIXSAT:         return visitMULFIX(N);
   case ISD::MUL:                return visitMUL(N);
   case ISD::SDIV:               return visitSDIV(N);
   case ISD::UDIV:               return visitUDIV(N);
   case ISD::SREM:
if (SDValue Combined = visitADDCARRYLike(N0, N1, CarryIn, N))
     return Combined;
 
   if (SDValue Combined = visitADDCARRYLike(N1, N0, CarryIn, N))
     return Combined;
 
   return SDValue();
 }
 
+SDValue DAGCombiner::visitSADDO_CARRY(SDNode *N) {
+  SDValue N0 = N->getOperand(0);
+  SDValue N1 = N->getOperand(1);
+  SDValue CarryIn = N->getOperand(2);
+  SDLoc DL(N);
+
+  // canonicalize constant to RHS
+  ConstantSDNode *N0C = dyn_cast<ConstantSDNode>(N0);
+  ConstantSDNode *N1C = dyn_cast<ConstantSDNode>(N1);
+  if (N0C && !N1C)
+    return DAG.getNode(ISD::SADDO_CARRY, DL, N->getVTList(), N1, N0, CarryIn);
+
+  // fold (saddo_carry x, y, false) -> (saddo x, y)
+  if (isNullConstant(CarryIn)) {
+    if (!LegalOperations ||
+        TLI.isOperationLegalOrCustom(ISD::SADDO, N->getValueType(0)))
+      return DAG.getNode(ISD::SADDO, DL, N->getVTList(), N0, N1);
+  }
+
+  return SDValue();
+}
+
 /**
  * If we are facing some sort of diamond carry propapagtion pattern try to
  * break it up to generate something like:
  *   (addcarry X, 0, (addcarry A, B, Z):Carry)
  *
  * The end result is usually an increase in operation required, but because the
  * carry is now linearized, other tranforms can kick in and optimize the DAG.
  *
if (isNullConstant(CarryIn)) {
     if (!LegalOperations ||
         TLI.isOperationLegalOrCustom(ISD::USUBO, N->getValueType(0)))
       return DAG.getNode(ISD::USUBO, SDLoc(N), N->getVTList(), N0, N1);
   }
 
   return SDValue();
 }
 
+SDValue DAGCombiner::visitSSUBO_CARRY(SDNode *N) {
+  SDValue N0 = N->getOperand(0);
+  SDValue N1 = N->getOperand(1);
+  SDValue CarryIn = N->getOperand(2);
+
+  // fold (ssubo_carry x, y, false) -> (ssubo x, y)
+  if (isNullConstant(CarryIn)) {
+    if (!LegalOperations ||
+        TLI.isOperationLegalOrCustom(ISD::SSUBO, N->getValueType(0)))
+      return DAG.getNode(ISD::SSUBO, SDLoc(N), N->getVTList(), N0, N1);
+  }
+
+  return SDValue();
+}
+
 // Notice that "mulfix" can be any of SMULFIX, SMULFIXSAT, UMULFIX and
 // UMULFIXSAT here.
 SDValue DAGCombiner::visitMULFIX(SDNode *N) {
   SDValue N0 = N->getOperand(0);
   SDValue N1 = N->getOperand(1);
   SDValue Scale = N->getOperand(2);
   EVT VT = N0.getValueType();
 

llvm/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp

 //===----- LegalizeIntegerTypes.cpp - Legalization of integer types -------===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 //
#endif
   case ISD::SMULO:
   case ISD::UMULO:       Res = PromoteIntRes_XMULO(N, ResNo); break;
 
   case ISD::ADDE:
   case ISD::SUBE:
   case ISD::ADDCARRY:
   case ISD::SUBCARRY:    Res = PromoteIntRes_ADDSUBCARRY(N, ResNo); break;
 
+  case ISD::SADDO_CARRY:
+  case ISD::SSUBO_CARRY: Res = PromoteIntRes_SADDSUBO_CARRY(N, ResNo); break;
+
   case ISD::SADDSAT:
   case ISD::UADDSAT:
   case ISD::SSUBSAT:
   case ISD::USUBSAT:
   case ISD::SSHLSAT:
   case ISD::USHLSAT:     Res = PromoteIntRes_ADDSUBSHLSAT(N); break;
 
   case ISD::SMULFIX:
SDValue Res = DAG.getNode(N->getOpcode(), SDLoc(N), DAG.getVTList(ValueVTs),
                             LHS, RHS, N->getOperand(2));
 
   // Update the users of the original carry/borrow value.
   ReplaceValueWith(SDValue(N, 1), Res.getValue(1));
 
   return SDValue(Res.getNode(), 0);
 }
 
+SDValue DAGTypeLegalizer::PromoteIntRes_SADDSUBO_CARRY(SDNode *N,
+                                                       unsigned ResNo) {
+  assert(ResNo == 1 && "Don't know how to promote other results yet.");
+  return PromoteIntRes_Overflow(N);
+}
+
 SDValue DAGTypeLegalizer::PromoteIntRes_ABS(SDNode *N) {
   SDValue Op0 = SExtPromotedInteger(N->getOperand(0));
   return DAG.getNode(ISD::ABS, SDLoc(N), Op0.getValueType(), Op0);
 }
 
 SDValue DAGTypeLegalizer::PromoteIntRes_XMULO(SDNode *N, unsigned ResNo) {
   // Promote the overflow bit trivially.
   if (ResNo == 1)
bool DAGTypeLegalizer::PromoteIntegerOperand(SDNode *N, unsigned OpNo) {
   case ISD::EXTRACT_SUBVECTOR: Res = PromoteIntOp_EXTRACT_SUBVECTOR(N); break;
 
   case ISD::SHL:
   case ISD::SRA:
   case ISD::SRL:
   case ISD::ROTL:
   case ISD::ROTR: Res = PromoteIntOp_Shift(N); break;
 
+  case ISD::SADDO_CARRY:
+  case ISD::SSUBO_CARRY:
   case ISD::ADDCARRY:
   case ISD::SUBCARRY: Res = PromoteIntOp_ADDSUBCARRY(N, OpNo); break;
 
   case ISD::FRAMEADDR:
   case ISD::RETURNADDR: Res = PromoteIntOp_FRAMERETURNADDR(N); break;
 
   case ISD::PREFETCH: Res = PromoteIntOp_PREFETCH(N, OpNo); break;
 

 }
 
 void DAGTypeLegalizer::ExpandIntRes_SADDSUBO(SDNode *Node,
                                              SDValue &Lo, SDValue &Hi) {
   SDValue LHS = Node->getOperand(0);
   SDValue RHS = Node->getOperand(1);
   SDLoc dl(Node);
 
+  SDValue Ovf;
+
+  unsigned CarryOp;
+  switch(Node->getOpcode()) {
+  default: llvm_unreachable("Node has unexpected Opcode");
+  case ISD::SADDO: CarryOp = ISD::SADDO_CARRY; break;
+  case ISD::SSUBO: CarryOp = ISD::SSUBO_CARRY; break;
+  }
+
+  bool HasCarryOp = TLI.isOperationLegalOrCustom(
+      CarryOp, TLI.getTypeToExpandTo(*DAG.getContext(), LHS.getValueType()));
+
+  if (HasCarryOp) {
+    // Expand the subcomponents.
+    SDValue LHSL, LHSH, RHSL, RHSH;
+    GetExpandedInteger(LHS, LHSL, LHSH);
+    GetExpandedInteger(RHS, RHSL, RHSH);
+    SDVTList VTList = DAG.getVTList(LHSL.getValueType(), Node->getValueType(1));
+
+    Lo = DAG.getNode(Node->getOpcode() == ISD::SADDO ?
+                     ISD::UADDO : ISD::USUBO, dl, VTList, { LHSL, RHSL });
+    Hi = DAG.getNode(CarryOp, dl, VTList, { LHSH, RHSH, Lo.getValue(1) });
+
+    Ovf = Hi.getValue(1);
+  } else {
     // Expand the result by simply replacing it with the equivalent
     // non-overflow-checking operation.
     SDValue Sum = DAG.getNode(Node->getOpcode() == ISD::SADDO ?
                               ISD::ADD : ISD::SUB, dl, LHS.getValueType(),
                               LHS, RHS);
     SplitInteger(Sum, Lo, Hi);
 
     // Compute the overflow.
     //
     //   LHSSign -> LHS >= 0
     //   RHSSign -> RHS >= 0
     //   SumSign -> Sum >= 0
     //
     //   Add:
     //   Overflow -> (LHSSign == RHSSign) && (LHSSign != SumSign)
     //   Sub:
     //   Overflow -> (LHSSign != RHSSign) && (LHSSign != SumSign)
     //
     EVT OType = Node->getValueType(1);
     SDValue Zero = DAG.getConstant(0, dl, LHS.getValueType());
 
     SDValue LHSSign = DAG.getSetCC(dl, OType, LHS, Zero, ISD::SETGE);
     SDValue RHSSign = DAG.getSetCC(dl, OType, RHS, Zero, ISD::SETGE);
     SDValue SignsMatch = DAG.getSetCC(dl, OType, LHSSign, RHSSign,
                                       Node->getOpcode() == ISD::SADDO ?
                                       ISD::SETEQ : ISD::SETNE);
 
     SDValue SumSign = DAG.getSetCC(dl, OType, Sum, Zero, ISD::SETGE);
     SDValue SumSignNE = DAG.getSetCC(dl, OType, LHSSign, SumSign, ISD::SETNE);
 
-  SDValue Cmp = DAG.getNode(ISD::AND, dl, OType, SignsMatch, SumSignNE);
+    Ovf = DAG.getNode(ISD::AND, dl, OType, SignsMatch, SumSignNE);
+  }
 
   // Use the calculated overflow everywhere.
-  ReplaceValueWith(SDValue(Node, 1), Cmp);
+  ReplaceValueWith(SDValue(Node, 1), Ovf);
 }
 
 void DAGTypeLegalizer::ExpandIntRes_SDIV(SDNode *N,
                                          SDValue &Lo, SDValue &Hi) {
   EVT VT = N->getValueType(0);
   SDLoc dl(N);
   SDValue Ops[2] = { N->getOperand(0), N->getOperand(1) };
 

llvm/lib/CodeGen/SelectionDAG/LegalizeTypes.h

 //===-- LegalizeTypes.h - DAG Type Legalizer class definition ---*- C++ -*-===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 //
private:
   SDValue PromoteIntRes_ZExtIntBinOp(SDNode *N);
   SDValue PromoteIntRes_SExtIntBinOp(SDNode *N);
   SDValue PromoteIntRes_SIGN_EXTEND_INREG(SDNode *N);
   SDValue PromoteIntRes_SRA(SDNode *N);
   SDValue PromoteIntRes_SRL(SDNode *N);
   SDValue PromoteIntRes_TRUNCATE(SDNode *N);
   SDValue PromoteIntRes_UADDSUBO(SDNode *N, unsigned ResNo);
   SDValue PromoteIntRes_ADDSUBCARRY(SDNode *N, unsigned ResNo);
+  SDValue PromoteIntRes_SADDSUBO_CARRY(SDNode *N, unsigned ResNo);
   SDValue PromoteIntRes_UNDEF(SDNode *N);
   SDValue PromoteIntRes_VAARG(SDNode *N);
   SDValue PromoteIntRes_VSCALE(SDNode *N);
   SDValue PromoteIntRes_XMULO(SDNode *N, unsigned ResNo);
   SDValue PromoteIntRes_ADDSUBSHLSAT(SDNode *N);
   SDValue PromoteIntRes_MULFIX(SDNode *N);
   SDValue PromoteIntRes_DIVFIX(SDNode *N);
   SDValue PromoteIntRes_FLT_ROUNDS(SDNode *N);

llvm/lib/CodeGen/SelectionDAG/SelectionDAGDumper.cpp

 //===- SelectionDAGDumper.cpp - Implement SelectionDAG::dump() ------------===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 //
#endif
   case ISD::EXTRACT_SUBVECTOR:          return "extract_subvector";
   case ISD::SCALAR_TO_VECTOR:           return "scalar_to_vector";
   case ISD::VECTOR_SHUFFLE:             return "vector_shuffle";
   case ISD::SPLAT_VECTOR:               return "splat_vector";
   case ISD::CARRY_FALSE:                return "carry_false";
   case ISD::ADDC:                       return "addc";
   case ISD::ADDE:                       return "adde";
   case ISD::ADDCARRY:                   return "addcarry";
+  case ISD::SADDO_CARRY:                return "saddo_carry";
   case ISD::SADDO:                      return "saddo";
   case ISD::UADDO:                      return "uaddo";
   case ISD::SSUBO:                      return "ssubo";
   case ISD::USUBO:                      return "usubo";
   case ISD::SMULO:                      return "smulo";
   case ISD::UMULO:                      return "umulo";
   case ISD::SUBC:                       return "subc";
   case ISD::SUBE:                       return "sube";
   case ISD::SUBCARRY:                   return "subcarry";
+  case ISD::SSUBO_CARRY:                return "ssubo_carry";
   case ISD::SHL_PARTS:                  return "shl_parts";
   case ISD::SRA_PARTS:                  return "sra_parts";
   case ISD::SRL_PARTS:                  return "srl_parts";
 
   case ISD::SADDSAT:                    return "saddsat";
   case ISD::UADDSAT:                    return "uaddsat";
   case ISD::SSUBSAT:                    return "ssubsat";
   case ISD::USUBSAT:                    return "usubsat";

llvm/lib/Target/X86/X86ISelLowering.cpp

 //===-- X86ISelLowering.cpp - X86 DAG Lowering Implementation -------------===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 //
for (auto VT : { MVT::i8, MVT::i16, MVT::i32, MVT::i64 }) {
     setOperationAction(ISD::USUBO, VT, Custom);
     setOperationAction(ISD::SMULO, VT, Custom);
     setOperationAction(ISD::UMULO, VT, Custom);
 
     // Support carry in as value rather than glue.
     setOperationAction(ISD::ADDCARRY, VT, Custom);
     setOperationAction(ISD::SUBCARRY, VT, Custom);
     setOperationAction(ISD::SETCCCARRY, VT, Custom);
+    setOperationAction(ISD::SADDO_CARRY, VT, Custom);
+    setOperationAction(ISD::SSUBO_CARRY, VT, Custom);
   }
 
   if (!Subtarget.is64Bit()) {
     // These libcalls are not available in 32-bit.
     setLibcallName(RTLIB::SHL_I128, nullptr);
     setLibcallName(RTLIB::SRL_I128, nullptr);
     setLibcallName(RTLIB::SRA_I128, nullptr);
     setLibcallName(RTLIB::MUL_I128, nullptr);
SDValue Swap = DAG.getAtomic(ISD::ATOMIC_SWAP, dl,
                                Node->getOperand(1), Node->getOperand(2),
                                Node->getMemOperand());
   return Swap.getValue(1);
 }
 
 static SDValue LowerADDSUBCARRY(SDValue Op, SelectionDAG &DAG) {
   SDNode *N = Op.getNode();
   MVT VT = N->getSimpleValueType(0);
+  unsigned Opc = Op.getOpcode();
 
   // Let legalize expand this if it isn't a legal type yet.
   if (!DAG.getTargetLoweringInfo().isTypeLegal(VT))
     return SDValue();
 
   SDVTList VTs = DAG.getVTList(VT, MVT::i32);
   SDLoc DL(N);
 
   // Set the carry flag.
   SDValue Carry = Op.getOperand(2);
   EVT CarryVT = Carry.getValueType();
   Carry = DAG.getNode(X86ISD::ADD, DL, DAG.getVTList(CarryVT, MVT::i32),
                       Carry, DAG.getAllOnesConstant(DL, CarryVT));
 
-  unsigned Opc = Op.getOpcode() == ISD::ADDCARRY ? X86ISD::ADC : X86ISD::SBB;
-  SDValue Sum = DAG.getNode(Opc, DL, VTs, Op.getOperand(0),
-                            Op.getOperand(1), Carry.getValue(1));
+  bool IsAdd = Opc == ISD::ADDCARRY || Opc == ISD::SADDO_CARRY;
+  SDValue Sum = DAG.getNode(IsAdd ? X86ISD::ADC : X86ISD::SBB, DL, VTs,
+                            Op.getOperand(0), Op.getOperand(1),
+                            Carry.getValue(1));
 
-  SDValue SetCC = getSETCC(X86::COND_B, Sum.getValue(1), DL, DAG);
+  bool IsSigned = Opc == ISD::SADDO_CARRY || Opc == ISD::SSUBO_CARRY;
+  SDValue SetCC = getSETCC(IsSigned ? X86::COND_O : X86::COND_B,
+                           Sum.getValue(1), DL, DAG);
   if (N->getValueType(1) == MVT::i1)
     SetCC = DAG.getNode(ISD::TRUNCATE, DL, MVT::i1, SetCC);
 
   return DAG.getNode(ISD::MERGE_VALUES, DL, N->getVTList(), Sum, SetCC);
 }
 
 static SDValue LowerFSINCOS(SDValue Op, const X86Subtarget &Subtarget,
                             SelectionDAG &DAG) {
SDValue X86TargetLowering::LowerOperation(SDValue Op, SelectionDAG &DAG) const {
   case ISD::SADDO:
   case ISD::UADDO:
   case ISD::SSUBO:
   case ISD::USUBO:
   case ISD::SMULO:
   case ISD::UMULO:              return LowerXALUO(Op, DAG);
   case ISD::READCYCLECOUNTER:   return LowerREADCYCLECOUNTER(Op, Subtarget,DAG);
   case ISD::BITCAST:            return LowerBITCAST(Op, Subtarget, DAG);
+  case ISD::SADDO_CARRY:
+  case ISD::SSUBO_CARRY:
   case ISD::ADDCARRY:
   case ISD::SUBCARRY:           return LowerADDSUBCARRY(Op, DAG);
   case ISD::ADD:
   case ISD::SUB:                return lowerAddSub(Op, DAG, Subtarget);
   case ISD::UADDSAT:
   case ISD::SADDSAT:
   case ISD::USUBSAT:
   case ISD::SSUBSAT:            return LowerADDSAT_SUBSAT(Op, DAG, Subtarget);

llvm/test/CodeGen/X86/known-bits.ll

; X64-NEXT:    retq
   %8 = add i128 %5, %7
   %9 = shl i128 %8, 54
   ret i128 %9
 }
 
 define {i32, i1} @knownbits_uaddo_saddo(i64 %a0, i64 %a1) nounwind {
 ; X32-LABEL: knownbits_uaddo_saddo:
 ; X32:       # %bb.0:
-; X32-NEXT:    pushl %ebx
 ; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
-; X32-NEXT:    movl %ecx, %edx
-; X32-NEXT:    addl %eax, %edx
-; X32-NEXT:    setb %bl
-; X32-NEXT:    testl %eax, %eax
-; X32-NEXT:    setns %al
-; X32-NEXT:    testl %ecx, %ecx
-; X32-NEXT:    setns %cl
-; X32-NEXT:    cmpb %al, %cl
-; X32-NEXT:    sete %al
-; X32-NEXT:    testl %edx, %edx
-; X32-NEXT:    setns %dl
-; X32-NEXT:    cmpb %dl, %cl
-; X32-NEXT:    setne %dl
-; X32-NEXT:    andb %al, %dl
-; X32-NEXT:    orb %bl, %dl
+; X32-NEXT:    addl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    setb %al
+; X32-NEXT:    seto %dl
+; X32-NEXT:    orb %al, %dl
 ; X32-NEXT:    xorl %eax, %eax
-; X32-NEXT:    popl %ebx
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: knownbits_uaddo_saddo:
 ; X64:       # %bb.0:
 ; X64-NEXT:    shlq $32, %rdi
 ; X64-NEXT:    shlq $32, %rsi
 ; X64-NEXT:    addq %rdi, %rsi
 ; X64-NEXT:    setb %al
; X64-NEXT:    retq
   %ret1 = insertvalue {i32, i1} %ret0, i1 %4, 1
   ret {i32, i1} %ret1
 }
 
 define {i32, i1} @knownbits_usubo_ssubo(i64 %a0, i64 %a1) nounwind {
 ; X32-LABEL: knownbits_usubo_ssubo:
 ; X32:       # %bb.0:
 ; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
-; X32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
-; X32-NEXT:    cmpl %eax, %ecx
-; X32-NEXT:    setb %dh
-; X32-NEXT:    setns %dl
-; X32-NEXT:    testl %ecx, %ecx
-; X32-NEXT:    setns %cl
-; X32-NEXT:    cmpb %dl, %cl
-; X32-NEXT:    setne %ch
-; X32-NEXT:    testl %eax, %eax
-; X32-NEXT:    setns %al
-; X32-NEXT:    cmpb %al, %cl
-; X32-NEXT:    setne %dl
-; X32-NEXT:    andb %ch, %dl
-; X32-NEXT:    orb %dh, %dl
+; X32-NEXT:    cmpl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    setb %al
+; X32-NEXT:    seto %dl
+; X32-NEXT:    orb %al, %dl
 ; X32-NEXT:    xorl %eax, %eax
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: knownbits_usubo_ssubo:
 ; X64:       # %bb.0:
 ; X64-NEXT:    shlq $32, %rdi
 ; X64-NEXT:    shlq $32, %rsi
 ; X64-NEXT:    cmpq %rsi, %rdi

llvm/test/CodeGen/X86/sadd_sat.ll

 ; X64-NEXT:    retq
   %tmp = call i32 @llvm.sadd.sat.i32(i32 %x, i32 %y);
   ret i32 %tmp;
 }
 
 define i64 @func2(i64 %x, i64 %y) nounwind {
 ; X86-LABEL: func2:
 ; X86:       # %bb.0:
-; X86-NEXT:    pushl %ebp
 ; X86-NEXT:    pushl %ebx
-; X86-NEXT:    pushl %edi
 ; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
 ; X86-NEXT:    movl {{[0-9]+}}(%esp), %esi
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %edi
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %ebx
-; X86-NEXT:    addl {{[0-9]+}}(%esp), %edi
-; X86-NEXT:    movl %ebx, %ebp
-; X86-NEXT:    adcl %esi, %ebp
-; X86-NEXT:    movl %ebp, %eax
+; X86-NEXT:    addl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    adcl {{[0-9]+}}(%esp), %esi
+; X86-NEXT:    seto %bl
+; X86-NEXT:    movl %esi, %eax
 ; X86-NEXT:    sarl $31, %eax
-; X86-NEXT:    xorl %ecx, %ecx
-; X86-NEXT:    testl %ebp, %ebp
-; X86-NEXT:    setns %cl
-; X86-NEXT:    movl %ecx, %edx
-; X86-NEXT:    addl $2147483647, %edx # imm = 0x7FFFFFFF
-; X86-NEXT:    testl %ebx, %ebx
-; X86-NEXT:    setns %bl
-; X86-NEXT:    cmpb %cl, %bl
-; X86-NEXT:    setne %cl
+; X86-NEXT:    testb %bl, %bl
+; X86-NEXT:    cmovel %ecx, %eax
+; X86-NEXT:    xorl %edx, %edx
 ; X86-NEXT:    testl %esi, %esi
-; X86-NEXT:    setns %ch
-; X86-NEXT:    cmpb %ch, %bl
-; X86-NEXT:    sete %ch
-; X86-NEXT:    testb %cl, %ch
-; X86-NEXT:    cmovel %ebp, %edx
-; X86-NEXT:    cmovel %edi, %eax
+; X86-NEXT:    setns %dl
+; X86-NEXT:    addl $2147483647, %edx # imm = 0x7FFFFFFF
+; X86-NEXT:    testb %bl, %bl
+; X86-NEXT:    cmovel %esi, %edx
 ; X86-NEXT:    popl %esi
-; X86-NEXT:    popl %edi
 ; X86-NEXT:    popl %ebx
-; X86-NEXT:    popl %ebp
 ; X86-NEXT:    retl
 ;
 ; X64-LABEL: func2:
 ; X64:       # %bb.0:
 ; X64-NEXT:    xorl %ecx, %ecx
 ; X64-NEXT:    movq %rdi, %rax
 ; X64-NEXT:    addq %rsi, %rax
 ; X64-NEXT:    setns %cl

llvm/test/CodeGen/X86/sadd_sat_plus.ll

; X64-NEXT:    retq
   %a = mul i32 %y, %z
   %tmp = call i32 @llvm.sadd.sat.i32(i32 %x, i32 %a)
   ret i32 %tmp
 }
 
 define i64 @func64(i64 %x, i64 %y, i64 %z) nounwind {
 ; X86-LABEL: func64:
 ; X86:       # %bb.0:
-; X86-NEXT:    pushl %ebp
 ; X86-NEXT:    pushl %ebx
-; X86-NEXT:    pushl %edi
 ; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
 ; X86-NEXT:    movl {{[0-9]+}}(%esp), %esi
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %edi
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %ebx
-; X86-NEXT:    addl {{[0-9]+}}(%esp), %edi
-; X86-NEXT:    movl %ebx, %ebp
-; X86-NEXT:    adcl %esi, %ebp
-; X86-NEXT:    movl %ebp, %eax
+; X86-NEXT:    addl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    adcl {{[0-9]+}}(%esp), %esi
+; X86-NEXT:    seto %bl
+; X86-NEXT:    movl %esi, %eax
 ; X86-NEXT:    sarl $31, %eax
-; X86-NEXT:    xorl %ecx, %ecx
-; X86-NEXT:    testl %ebp, %ebp
-; X86-NEXT:    setns %cl
-; X86-NEXT:    movl %ecx, %edx
-; X86-NEXT:    addl $2147483647, %edx # imm = 0x7FFFFFFF
-; X86-NEXT:    testl %ebx, %ebx
-; X86-NEXT:    setns %bl
-; X86-NEXT:    cmpb %cl, %bl
-; X86-NEXT:    setne %cl
+; X86-NEXT:    testb %bl, %bl
+; X86-NEXT:    cmovel %ecx, %eax
+; X86-NEXT:    xorl %edx, %edx
 ; X86-NEXT:    testl %esi, %esi
-; X86-NEXT:    setns %ch
-; X86-NEXT:    cmpb %ch, %bl
-; X86-NEXT:    sete %ch
-; X86-NEXT:    testb %cl, %ch
-; X86-NEXT:    cmovel %ebp, %edx
-; X86-NEXT:    cmovel %edi, %eax
+; X86-NEXT:    setns %dl
+; X86-NEXT:    addl $2147483647, %edx # imm = 0x7FFFFFFF
+; X86-NEXT:    testb %bl, %bl
+; X86-NEXT:    cmovel %esi, %edx
 ; X86-NEXT:    popl %esi
-; X86-NEXT:    popl %edi
 ; X86-NEXT:    popl %ebx
-; X86-NEXT:    popl %ebp
 ; X86-NEXT:    retl
 ;
 ; X64-LABEL: func64:
 ; X64:       # %bb.0:
 ; X64-NEXT:    xorl %ecx, %ecx
 ; X64-NEXT:    movq %rdi, %rax
 ; X64-NEXT:    addq %rdx, %rax
 ; X64-NEXT:    setns %cl

llvm/test/CodeGen/X86/sadd_sat_vec.ll

 ; AVX512-NEXT:    retq
   %z = call <8 x i64> @llvm.sadd.sat.v8i64(<8 x i64> %x, <8 x i64> %y)
   ret <8 x i64> %z
 }
 
 define <2 x i128> @v2i128(<2 x i128> %x, <2 x i128> %y) nounwind {
 ; SSE-LABEL: v2i128:
 ; SSE:       # %bb.0:
-; SSE-NEXT:    pushq %r15
-; SSE-NEXT:    pushq %r14
-; SSE-NEXT:    pushq %r13
-; SSE-NEXT:    pushq %r12
 ; SSE-NEXT:    pushq %rbx
 ; SSE-NEXT:    movq %rdi, %rax
-; SSE-NEXT:    movq {{[0-9]+}}(%rsp), %r11
-; SSE-NEXT:    movq {{[0-9]+}}(%rsp), %r14
 ; SSE-NEXT:    addq {{[0-9]+}}(%rsp), %rcx
-; SSE-NEXT:    movq %r8, %r13
-; SSE-NEXT:    adcq %r14, %r13
-; SSE-NEXT:    movq %r13, %r10
-; SSE-NEXT:    sarq $63, %r10
-; SSE-NEXT:    xorl %edi, %edi
-; SSE-NEXT:    testq %r13, %r13
-; SSE-NEXT:    setns %dil
-; SSE-NEXT:    movabsq $9223372036854775807, %r12 # imm = 0x7FFFFFFFFFFFFFFF
-; SSE-NEXT:    leaq (%rdi,%r12), %r15
+; SSE-NEXT:    adcq {{[0-9]+}}(%rsp), %r8
+; SSE-NEXT:    seto %r10b
+; SSE-NEXT:    movq %r8, %rbx
+; SSE-NEXT:    sarq $63, %rbx
+; SSE-NEXT:    testb %r10b, %r10b
+; SSE-NEXT:    cmoveq %rcx, %rbx
+; SSE-NEXT:    xorl %ecx, %ecx
 ; SSE-NEXT:    testq %r8, %r8
-; SSE-NEXT:    setns %r8b
-; SSE-NEXT:    cmpb %dil, %r8b
-; SSE-NEXT:    setne %dil
-; SSE-NEXT:    testq %r14, %r14
-; SSE-NEXT:    setns %bl
-; SSE-NEXT:    cmpb %bl, %r8b
-; SSE-NEXT:    sete %bl
-; SSE-NEXT:    testb %dil, %bl
-; SSE-NEXT:    cmoveq %r13, %r15
-; SSE-NEXT:    cmoveq %rcx, %r10
+; SSE-NEXT:    setns %cl
+; SSE-NEXT:    movabsq $9223372036854775807, %r11 # imm = 0x7FFFFFFFFFFFFFFF
+; SSE-NEXT:    addq %r11, %rcx
+; SSE-NEXT:    testb %r10b, %r10b
+; SSE-NEXT:    cmoveq %r8, %rcx
 ; SSE-NEXT:    addq %r9, %rsi
+; SSE-NEXT:    adcq {{[0-9]+}}(%rsp), %rdx
+; SSE-NEXT:    seto %r8b
 ; SSE-NEXT:    movq %rdx, %rdi
-; SSE-NEXT:    adcq %r11, %rdi
-; SSE-NEXT:    setns %bl
-; SSE-NEXT:    movzbl %bl, %ebx
-; SSE-NEXT:    addq %rbx, %r12
-; SSE-NEXT:    movq %rdi, %rcx
-; SSE-NEXT:    sarq $63, %rcx
-; SSE-NEXT:    testq %r11, %r11
-; SSE-NEXT:    setns %r8b
+; SSE-NEXT:    sarq $63, %rdi
+; SSE-NEXT:    testb %r8b, %r8b
+; SSE-NEXT:    cmoveq %rsi, %rdi
+; SSE-NEXT:    xorl %esi, %esi
 ; SSE-NEXT:    testq %rdx, %rdx
-; SSE-NEXT:    setns %dl
-; SSE-NEXT:    cmpb %r8b, %dl
-; SSE-NEXT:    sete %r8b
-; SSE-NEXT:    cmpb %bl, %dl
-; SSE-NEXT:    setne %dl
-; SSE-NEXT:    testb %dl, %r8b
-; SSE-NEXT:    cmoveq %rsi, %rcx
-; SSE-NEXT:    cmoveq %rdi, %r12
-; SSE-NEXT:    movq %r15, 24(%rax)
-; SSE-NEXT:    movq %r10, 16(%rax)
-; SSE-NEXT:    movq %r12, 8(%rax)
-; SSE-NEXT:    movq %rcx, (%rax)
+; SSE-NEXT:    setns %sil
+; SSE-NEXT:    addq %r11, %rsi
+; SSE-NEXT:    testb %r8b, %r8b
+; SSE-NEXT:    cmoveq %rdx, %rsi
+; SSE-NEXT:    movq %rbx, 16(%rax)
+; SSE-NEXT:    movq %rdi, (%rax)
+; SSE-NEXT:    movq %rcx, 24(%rax)
+; SSE-NEXT:    movq %rsi, 8(%rax)
 ; SSE-NEXT:    popq %rbx
-; SSE-NEXT:    popq %r12
-; SSE-NEXT:    popq %r13
-; SSE-NEXT:    popq %r14
-; SSE-NEXT:    popq %r15
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: v2i128:
 ; AVX:       # %bb.0:
-; AVX-NEXT:    pushq %r15
-; AVX-NEXT:    pushq %r14
-; AVX-NEXT:    pushq %r13
-; AVX-NEXT:    pushq %r12
 ; AVX-NEXT:    pushq %rbx
 ; AVX-NEXT:    movq %rdi, %rax
-; AVX-NEXT:    movq {{[0-9]+}}(%rsp), %r11
-; AVX-NEXT:    movq {{[0-9]+}}(%rsp), %r14
 ; AVX-NEXT:    addq {{[0-9]+}}(%rsp), %rcx
-; AVX-NEXT:    movq %r8, %r13
-; AVX-NEXT:    adcq %r14, %r13
-; AVX-NEXT:    movq %r13, %r10
-; AVX-NEXT:    sarq $63, %r10
-; AVX-NEXT:    xorl %edi, %edi
-; AVX-NEXT:    testq %r13, %r13
-; AVX-NEXT:    setns %dil
-; AVX-NEXT:    movabsq $9223372036854775807, %r12 # imm = 0x7FFFFFFFFFFFFFFF
-; AVX-NEXT:    leaq (%rdi,%r12), %r15
+; AVX-NEXT:    adcq {{[0-9]+}}(%rsp), %r8
+; AVX-NEXT:    seto %r10b
+; AVX-NEXT:    movq %r8, %rbx
+; AVX-NEXT:    sarq $63, %rbx
+; AVX-NEXT:    testb %r10b, %r10b
+; AVX-NEXT:    cmoveq %rcx, %rbx
+; AVX-NEXT:    xorl %ecx, %ecx
 ; AVX-NEXT:    testq %r8, %r8
-; AVX-NEXT:    setns %r8b
-; AVX-NEXT:    cmpb %dil, %r8b
-; AVX-NEXT:    setne %dil
-; AVX-NEXT:    testq %r14, %r14
-; AVX-NEXT:    setns %bl
-; AVX-NEXT:    cmpb %bl, %r8b
-; AVX-NEXT:    sete %bl
-; AVX-NEXT:    testb %dil, %bl
-; AVX-NEXT:    cmoveq %r13, %r15
-; AVX-NEXT:    cmoveq %rcx, %r10
+; AVX-NEXT:    setns %cl
+; AVX-NEXT:    movabsq $9223372036854775807, %r11 # imm = 0x7FFFFFFFFFFFFFFF
+; AVX-NEXT:    addq %r11, %rcx
+; AVX-NEXT:    testb %r10b, %r10b
+; AVX-NEXT:    cmoveq %r8, %rcx
 ; AVX-NEXT:    addq %r9, %rsi
+; AVX-NEXT:    adcq {{[0-9]+}}(%rsp), %rdx
+; AVX-NEXT:    seto %r8b
 ; AVX-NEXT:    movq %rdx, %rdi
-; AVX-NEXT:    adcq %r11, %rdi
-; AVX-NEXT:    setns %bl
-; AVX-NEXT:    movzbl %bl, %ebx
-; AVX-NEXT:    addq %rbx, %r12
-; AVX-NEXT:    movq %rdi, %rcx
-; AVX-NEXT:    sarq $63, %rcx
-; AVX-NEXT:    testq %r11, %r11
-; AVX-NEXT:    setns %r8b
+; AVX-NEXT:    sarq $63, %rdi
+; AVX-NEXT:    testb %r8b, %r8b
+; AVX-NEXT:    cmoveq %rsi, %rdi
+; AVX-NEXT:    xorl %esi, %esi
 ; AVX-NEXT:    testq %rdx, %rdx
-; AVX-NEXT:    setns %dl
-; AVX-NEXT:    cmpb %r8b, %dl
-; AVX-NEXT:    sete %r8b
-; AVX-NEXT:    cmpb %bl, %dl
-; AVX-NEXT:    setne %dl
-; AVX-NEXT:    testb %dl, %r8b
-; AVX-NEXT:    cmoveq %rsi, %rcx
-; AVX-NEXT:    cmoveq %rdi, %r12
-; AVX-NEXT:    movq %r15, 24(%rax)
-; AVX-NEXT:    movq %r10, 16(%rax)
-; AVX-NEXT:    movq %r12, 8(%rax)
-; AVX-NEXT:    movq %rcx, (%rax)
+; AVX-NEXT:    setns %sil
+; AVX-NEXT:    addq %r11, %rsi
+; AVX-NEXT:    testb %r8b, %r8b
+; AVX-NEXT:    cmoveq %rdx, %rsi
+; AVX-NEXT:    movq %rbx, 16(%rax)
+; AVX-NEXT:    movq %rdi, (%rax)
+; AVX-NEXT:    movq %rcx, 24(%rax)
+; AVX-NEXT:    movq %rsi, 8(%rax)
 ; AVX-NEXT:    popq %rbx
-; AVX-NEXT:    popq %r12
-; AVX-NEXT:    popq %r13
-; AVX-NEXT:    popq %r14
-; AVX-NEXT:    popq %r15
 ; AVX-NEXT:    retq
   %z = call <2 x i128> @llvm.sadd.sat.v2i128(<2 x i128> %x, <2 x i128> %y)
   ret <2 x i128> %z
 }

llvm/test/CodeGen/X86/ssub_sat.ll

 ; X64-NEXT:    retq
   %tmp = call i32 @llvm.ssub.sat.i32(i32 %x, i32 %y)
   ret i32 %tmp
 }
 
 define i64 @func2(i64 %x, i64 %y) nounwind {
 ; X86-LABEL: func2:
 ; X86:       # %bb.0:
-; X86-NEXT:    pushl %ebp
 ; X86-NEXT:    pushl %ebx
-; X86-NEXT:    pushl %edi
 ; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
 ; X86-NEXT:    movl {{[0-9]+}}(%esp), %esi
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %edi
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %ebx
-; X86-NEXT:    subl {{[0-9]+}}(%esp), %edi
-; X86-NEXT:    movl %ebx, %ebp
-; X86-NEXT:    sbbl %esi, %ebp
-; X86-NEXT:    movl %ebp, %eax
+; X86-NEXT:    subl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    sbbl {{[0-9]+}}(%esp), %esi
+; X86-NEXT:    seto %bl
+; X86-NEXT:    movl %esi, %eax
 ; X86-NEXT:    sarl $31, %eax
-; X86-NEXT:    xorl %ecx, %ecx
-; X86-NEXT:    testl %ebp, %ebp
-; X86-NEXT:    setns %cl
-; X86-NEXT:    movl %ecx, %edx
-; X86-NEXT:    addl $2147483647, %edx # imm = 0x7FFFFFFF
-; X86-NEXT:    testl %ebx, %ebx
-; X86-NEXT:    setns %bl
-; X86-NEXT:    cmpb %cl, %bl
-; X86-NEXT:    setne %cl
+; X86-NEXT:    testb %bl, %bl
+; X86-NEXT:    cmovel %ecx, %eax
+; X86-NEXT:    xorl %edx, %edx
 ; X86-NEXT:    testl %esi, %esi
-; X86-NEXT:    setns %ch
-; X86-NEXT:    cmpb %ch, %bl
-; X86-NEXT:    setne %ch
-; X86-NEXT:    testb %cl, %ch
-; X86-NEXT:    cmovel %ebp, %edx
-; X86-NEXT:    cmovel %edi, %eax
+; X86-NEXT:    setns %dl
+; X86-NEXT:    addl $2147483647, %edx # imm = 0x7FFFFFFF
+; X86-NEXT:    testb %bl, %bl
+; X86-NEXT:    cmovel %esi, %edx
 ; X86-NEXT:    popl %esi
-; X86-NEXT:    popl %edi
 ; X86-NEXT:    popl %ebx
-; X86-NEXT:    popl %ebp
 ; X86-NEXT:    retl
 ;
 ; X64-LABEL: func2:
 ; X64:       # %bb.0:
 ; X64-NEXT:    xorl %ecx, %ecx
 ; X64-NEXT:    cmpq %rsi, %rdi
 ; X64-NEXT:    setns %cl
 ; X64-NEXT:    movabsq $9223372036854775807, %rax # imm = 0x7FFFFFFFFFFFFFFF

llvm/test/CodeGen/X86/ssub_sat_plus.ll

; X64-NEXT:    retq
   %a = mul i32 %y, %z
   %tmp = call i32 @llvm.ssub.sat.i32(i32 %x, i32 %a)
   ret i32 %tmp
 }
 
 define i64 @func64(i64 %x, i64 %y, i64 %z) nounwind {
 ; X86-LABEL: func64:
 ; X86:       # %bb.0:
-; X86-NEXT:    pushl %ebp
 ; X86-NEXT:    pushl %ebx
-; X86-NEXT:    pushl %edi
 ; X86-NEXT:    pushl %esi
+; X86-NEXT:    movl {{[0-9]+}}(%esp), %ecx
 ; X86-NEXT:    movl {{[0-9]+}}(%esp), %esi
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %edi
-; X86-NEXT:    movl {{[0-9]+}}(%esp), %ebx
-; X86-NEXT:    subl {{[0-9]+}}(%esp), %edi
-; X86-NEXT:    movl %ebx, %ebp
-; X86-NEXT:    sbbl %esi, %ebp
-; X86-NEXT:    movl %ebp, %eax
+; X86-NEXT:    subl {{[0-9]+}}(%esp), %ecx
+; X86-NEXT:    sbbl {{[0-9]+}}(%esp), %esi
+; X86-NEXT:    seto %bl
+; X86-NEXT:    movl %esi, %eax
 ; X86-NEXT:    sarl $31, %eax
-; X86-NEXT:    xorl %ecx, %ecx
-; X86-NEXT:    testl %ebp, %ebp
-; X86-NEXT:    setns %cl
-; X86-NEXT:    movl %ecx, %edx
-; X86-NEXT:    addl $2147483647, %edx # imm = 0x7FFFFFFF
-; X86-NEXT:    testl %ebx, %ebx
-; X86-NEXT:    setns %bl
-; X86-NEXT:    cmpb %cl, %bl
-; X86-NEXT:    setne %cl
+; X86-NEXT:    testb %bl, %bl
+; X86-NEXT:    cmovel %ecx, %eax
+; X86-NEXT:    xorl %edx, %edx
 ; X86-NEXT:    testl %esi, %esi
-; X86-NEXT:    setns %ch
-; X86-NEXT:    cmpb %ch, %bl
-; X86-NEXT:    setne %ch
-; X86-NEXT:    testb %cl, %ch
-; X86-NEXT:    cmovel %ebp, %edx
-; X86-NEXT:    cmovel %edi, %eax
+; X86-NEXT:    setns %dl
+; X86-NEXT:    addl $2147483647, %edx # imm = 0x7FFFFFFF
+; X86-NEXT:    testb %bl, %bl
+; X86-NEXT:    cmovel %esi, %edx
 ; X86-NEXT:    popl %esi
-; X86-NEXT:    popl %edi
 ; X86-NEXT:    popl %ebx
-; X86-NEXT:    popl %ebp
 ; X86-NEXT:    retl
 ;
 ; X64-LABEL: func64:
 ; X64:       # %bb.0:
 ; X64-NEXT:    xorl %ecx, %ecx
 ; X64-NEXT:    cmpq %rdx, %rdi
 ; X64-NEXT:    setns %cl
 ; X64-NEXT:    movabsq $9223372036854775807, %rax # imm = 0x7FFFFFFFFFFFFFFF

llvm/test/CodeGen/X86/ssub_sat_vec.ll

 ; AVX512-NEXT:    retq
   %z = call <8 x i64> @llvm.ssub.sat.v8i64(<8 x i64> %x, <8 x i64> %y)
   ret <8 x i64> %z
 }
 
 define <2 x i128> @v2i128(<2 x i128> %x, <2 x i128> %y) nounwind {
 ; SSE-LABEL: v2i128:
 ; SSE:       # %bb.0:
-; SSE-NEXT:    pushq %r15
-; SSE-NEXT:    pushq %r14
-; SSE-NEXT:    pushq %r13
-; SSE-NEXT:    pushq %r12
 ; SSE-NEXT:    pushq %rbx
 ; SSE-NEXT:    movq %rdi, %rax
-; SSE-NEXT:    movq {{[0-9]+}}(%rsp), %r11
-; SSE-NEXT:    movq {{[0-9]+}}(%rsp), %r14
 ; SSE-NEXT:    subq {{[0-9]+}}(%rsp), %rcx
-; SSE-NEXT:    movq %r8, %r13
-; SSE-NEXT:    sbbq %r14, %r13
-; SSE-NEXT:    movq %r13, %r10
-; SSE-NEXT:    sarq $63, %r10
-; SSE-NEXT:    xorl %edi, %edi
-; SSE-NEXT:    testq %r13, %r13
-; SSE-NEXT:    setns %dil
-; SSE-NEXT:    movabsq $9223372036854775807, %r12 # imm = 0x7FFFFFFFFFFFFFFF
-; SSE-NEXT:    leaq (%rdi,%r12), %r15
+; SSE-NEXT:    sbbq {{[0-9]+}}(%rsp), %r8
+; SSE-NEXT:    seto %r10b
+; SSE-NEXT:    movq %r8, %rbx
+; SSE-NEXT:    sarq $63, %rbx
+; SSE-NEXT:    testb %r10b, %r10b
+; SSE-NEXT:    cmoveq %rcx, %rbx
+; SSE-NEXT:    xorl %ecx, %ecx
 ; SSE-NEXT:    testq %r8, %r8
-; SSE-NEXT:    setns %r8b
-; SSE-NEXT:    cmpb %dil, %r8b
-; SSE-NEXT:    setne %dil
-; SSE-NEXT:    testq %r14, %r14
-; SSE-NEXT:    setns %bl
-; SSE-NEXT:    cmpb %bl, %r8b
-; SSE-NEXT:    setne %bl
-; SSE-NEXT:    testb %dil, %bl
-; SSE-NEXT:    cmoveq %r13, %r15
-; SSE-NEXT:    cmoveq %rcx, %r10
+; SSE-NEXT:    setns %cl
+; SSE-NEXT:    movabsq $9223372036854775807, %r11 # imm = 0x7FFFFFFFFFFFFFFF
+; SSE-NEXT:    addq %r11, %rcx
+; SSE-NEXT:    testb %r10b, %r10b
+; SSE-NEXT:    cmoveq %r8, %rcx
 ; SSE-NEXT:    subq %r9, %rsi
+; SSE-NEXT:    sbbq {{[0-9]+}}(%rsp), %rdx
+; SSE-NEXT:    seto %r8b
 ; SSE-NEXT:    movq %rdx, %rdi
-; SSE-NEXT:    sbbq %r11, %rdi
-; SSE-NEXT:    setns %bl
-; SSE-NEXT:    movzbl %bl, %ebx
-; SSE-NEXT:    addq %rbx, %r12
-; SSE-NEXT:    movq %rdi, %rcx
-; SSE-NEXT:    sarq $63, %rcx
-; SSE-NEXT:    testq %r11, %r11
-; SSE-NEXT:    setns %r8b
+; SSE-NEXT:    sarq $63, %rdi
+; SSE-NEXT:    testb %r8b, %r8b
+; SSE-NEXT:    cmoveq %rsi, %rdi
+; SSE-NEXT:    xorl %esi, %esi
 ; SSE-NEXT:    testq %rdx, %rdx
-; SSE-NEXT:    setns %dl
-; SSE-NEXT:    cmpb %r8b, %dl
-; SSE-NEXT:    setne %r8b
-; SSE-NEXT:    cmpb %bl, %dl
-; SSE-NEXT:    setne %dl
-; SSE-NEXT:    testb %dl, %r8b
-; SSE-NEXT:    cmoveq %rsi, %rcx
-; SSE-NEXT:    cmoveq %rdi, %r12
-; SSE-NEXT:    movq %r15, 24(%rax)
-; SSE-NEXT:    movq %r10, 16(%rax)
-; SSE-NEXT:    movq %r12, 8(%rax)
-; SSE-NEXT:    movq %rcx, (%rax)
+; SSE-NEXT:    setns %sil
+; SSE-NEXT:    addq %r11, %rsi
+; SSE-NEXT:    testb %r8b, %r8b
+; SSE-NEXT:    cmoveq %rdx, %rsi
+; SSE-NEXT:    movq %rbx, 16(%rax)
+; SSE-NEXT:    movq %rdi, (%rax)
+; SSE-NEXT:    movq %rcx, 24(%rax)
+; SSE-NEXT:    movq %rsi, 8(%rax)
 ; SSE-NEXT:    popq %rbx
-; SSE-NEXT:    popq %r12
-; SSE-NEXT:    popq %r13
-; SSE-NEXT:    popq %r14
-; SSE-NEXT:    popq %r15
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: v2i128:
 ; AVX:       # %bb.0:
-; AVX-NEXT:    pushq %r15
-; AVX-NEXT:    pushq %r14
-; AVX-NEXT:    pushq %r13
-; AVX-NEXT:    pushq %r12
 ; AVX-NEXT:    pushq %rbx
 ; AVX-NEXT:    movq %rdi, %rax
-; AVX-NEXT:    movq {{[0-9]+}}(%rsp), %r11
-; AVX-NEXT:    movq {{[0-9]+}}(%rsp), %r14
 ; AVX-NEXT:    subq {{[0-9]+}}(%rsp), %rcx
-; AVX-NEXT:    movq %r8, %r13
-; AVX-NEXT:    sbbq %r14, %r13
-; AVX-NEXT:    movq %r13, %r10
-; AVX-NEXT:    sarq $63, %r10
-; AVX-NEXT:    xorl %edi, %edi
-; AVX-NEXT:    testq %r13, %r13
-; AVX-NEXT:    setns %dil
-; AVX-NEXT:    movabsq $9223372036854775807, %r12 # imm = 0x7FFFFFFFFFFFFFFF
-; AVX-NEXT:    leaq (%rdi,%r12), %r15
+; AVX-NEXT:    sbbq {{[0-9]+}}(%rsp), %r8
+; AVX-NEXT:    seto %r10b
+; AVX-NEXT:    movq %r8, %rbx
+; AVX-NEXT:    sarq $63, %rbx
+; AVX-NEXT:    testb %r10b, %r10b
+; AVX-NEXT:    cmoveq %rcx, %rbx
+; AVX-NEXT:    xorl %ecx, %ecx
 ; AVX-NEXT:    testq %r8, %r8
-; AVX-NEXT:    setns %r8b
-; AVX-NEXT:    cmpb %dil, %r8b
-; AVX-NEXT:    setne %dil
-; AVX-NEXT:    testq %r14, %r14
-; AVX-NEXT:    setns %bl
-; AVX-NEXT:    cmpb %bl, %r8b
-; AVX-NEXT:    setne %bl
-; AVX-NEXT:    testb %dil, %bl
-; AVX-NEXT:    cmoveq %r13, %r15
-; AVX-NEXT:    cmoveq %rcx, %r10
+; AVX-NEXT:    setns %cl
+; AVX-NEXT:    movabsq $9223372036854775807, %r11 # imm = 0x7FFFFFFFFFFFFFFF
+; AVX-NEXT:    addq %r11, %rcx
+; AVX-NEXT:    testb %r10b, %r10b
+; AVX-NEXT:    cmoveq %r8, %rcx
 ; AVX-NEXT:    subq %r9, %rsi
+; AVX-NEXT:    sbbq {{[0-9]+}}(%rsp), %rdx
+; AVX-NEXT:    seto %r8b
 ; AVX-NEXT:    movq %rdx, %rdi
-; AVX-NEXT:    sbbq %r11, %rdi
-; AVX-NEXT:    setns %bl
-; AVX-NEXT:    movzbl %bl, %ebx
-; AVX-NEXT:    addq %rbx, %r12
-; AVX-NEXT:    movq %rdi, %rcx
-; AVX-NEXT:    sarq $63, %rcx
-; AVX-NEXT:    testq %r11, %r11
-; AVX-NEXT:    setns %r8b
+; AVX-NEXT:    sarq $63, %rdi
+; AVX-NEXT:    testb %r8b, %r8b
+; AVX-NEXT:    cmoveq %rsi, %rdi
+; AVX-NEXT:    xorl %esi, %esi
 ; AVX-NEXT:    testq %rdx, %rdx
-; AVX-NEXT:    setns %dl
-; AVX-NEXT:    cmpb %r8b, %dl
-; AVX-NEXT:    setne %r8b
-; AVX-NEXT:    cmpb %bl, %dl
-; AVX-NEXT:    setne %dl
-; AVX-NEXT:    testb %dl, %r8b
-; AVX-NEXT:    cmoveq %rsi, %rcx
-; AVX-NEXT:    cmoveq %rdi, %r12
-; AVX-NEXT:    movq %r15, 24(%rax)
-; AVX-NEXT:    movq %r10, 16(%rax)
-; AVX-NEXT:    movq %r12, 8(%rax)
-; AVX-NEXT:    movq %rcx, (%rax)
+; AVX-NEXT:    setns %sil
+; AVX-NEXT:    addq %r11, %rsi
+; AVX-NEXT:    testb %r8b, %r8b
+; AVX-NEXT:    cmoveq %rdx, %rsi
+; AVX-NEXT:    movq %rbx, 16(%rax)
+; AVX-NEXT:    movq %rdi, (%rax)
+; AVX-NEXT:    movq %rcx, 24(%rax)
+; AVX-NEXT:    movq %rsi, 8(%rax)
 ; AVX-NEXT:    popq %rbx
-; AVX-NEXT:    popq %r12
-; AVX-NEXT:    popq %r13
-; AVX-NEXT:    popq %r14
-; AVX-NEXT:    popq %r15
 ; AVX-NEXT:    retq
   %z = call <2 x i128> @llvm.ssub.sat.v2i128(<2 x i128> %x, <2 x i128> %y)
   ret <2 x i128> %z
 }

llvm/test/CodeGen/X86/vec_saddo.ll

; AVX512-NEXT:    retq
   %res = sext <4 x i1> %obit to <4 x i32>
   store <4 x i1> %val, <4 x i1>* %p2
   ret <4 x i32> %res
 }
 
 define <2 x i32> @saddo_v2i128(<2 x i128> %a0, <2 x i128> %a1, <2 x i128>* %p2) nounwind {
 ; SSE2-LABEL: saddo_v2i128:
 ; SSE2:       # %bb.0:
-; SSE2-NEXT:    pushq %rbp
-; SSE2-NEXT:    pushq %rbx
-; SSE2-NEXT:    movq {{[0-9]+}}(%rsp), %r11
 ; SSE2-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; SSE2-NEXT:    testq %r9, %r9
-; SSE2-NEXT:    setns %al
-; SSE2-NEXT:    testq %rsi, %rsi
-; SSE2-NEXT:    setns %bl
-; SSE2-NEXT:    cmpb %al, %bl
-; SSE2-NEXT:    sete %bpl
 ; SSE2-NEXT:    addq %r8, %rdi
 ; SSE2-NEXT:    adcq %r9, %rsi
-; SSE2-NEXT:    setns %al
-; SSE2-NEXT:    cmpb %al, %bl
-; SSE2-NEXT:    setne %al
-; SSE2-NEXT:    andb %bpl, %al
+; SSE2-NEXT:    seto %r8b
 ; SSE2-NEXT:    addq {{[0-9]+}}(%rsp), %rdx
-; SSE2-NEXT:    movq %rcx, %rbp
-; SSE2-NEXT:    adcq %r10, %rbp
-; SSE2-NEXT:    setns %bl
-; SSE2-NEXT:    testq %rcx, %rcx
-; SSE2-NEXT:    setns %cl
-; SSE2-NEXT:    cmpb %bl, %cl
-; SSE2-NEXT:    setne %r8b
-; SSE2-NEXT:    testq %r10, %r10
-; SSE2-NEXT:    setns %bl
-; SSE2-NEXT:    cmpb %bl, %cl
-; SSE2-NEXT:    sete %cl
-; SSE2-NEXT:    andb %r8b, %cl
-; SSE2-NEXT:    movzbl %cl, %ecx
-; SSE2-NEXT:    negl %ecx
-; SSE2-NEXT:    movd %ecx, %xmm1
+; SSE2-NEXT:    adcq {{[0-9]+}}(%rsp), %rcx
+; SSE2-NEXT:    seto %al
 ; SSE2-NEXT:    movzbl %al, %eax
 ; SSE2-NEXT:    negl %eax
+; SSE2-NEXT:    movd %eax, %xmm1
+; SSE2-NEXT:    movzbl %r8b, %eax
+; SSE2-NEXT:    negl %eax
 ; SSE2-NEXT:    movd %eax, %xmm0
 ; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; SSE2-NEXT:    movq %rdx, 16(%r11)
-; SSE2-NEXT:    movq %rdi, (%r11)
-; SSE2-NEXT:    movq %rbp, 24(%r11)
-; SSE2-NEXT:    movq %rsi, 8(%r11)
-; SSE2-NEXT:    popq %rbx
-; SSE2-NEXT:    popq %rbp
+; SSE2-NEXT:    movq %rdx, 16(%r10)
+; SSE2-NEXT:    movq %rdi, (%r10)
+; SSE2-NEXT:    movq %rcx, 24(%r10)
+; SSE2-NEXT:    movq %rsi, 8(%r10)
 ; SSE2-NEXT:    retq
 ;
 ; SSSE3-LABEL: saddo_v2i128:
 ; SSSE3:       # %bb.0:
-; SSSE3-NEXT:    pushq %rbp
-; SSSE3-NEXT:    pushq %rbx
-; SSSE3-NEXT:    movq {{[0-9]+}}(%rsp), %r11
 ; SSSE3-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; SSSE3-NEXT:    testq %r9, %r9
-; SSSE3-NEXT:    setns %al
-; SSSE3-NEXT:    testq %rsi, %rsi
-; SSSE3-NEXT:    setns %bl
-; SSSE3-NEXT:    cmpb %al, %bl
-; SSSE3-NEXT:    sete %bpl
 ; SSSE3-NEXT:    addq %r8, %rdi
 ; SSSE3-NEXT:    adcq %r9, %rsi
-; SSSE3-NEXT:    setns %al
-; SSSE3-NEXT:    cmpb %al, %bl
-; SSSE3-NEXT:    setne %al
-; SSSE3-NEXT:    andb %bpl, %al
+; SSSE3-NEXT:    seto %r8b
 ; SSSE3-NEXT:    addq {{[0-9]+}}(%rsp), %rdx
-; SSSE3-NEXT:    movq %rcx, %rbp
-; SSSE3-NEXT:    adcq %r10, %rbp
-; SSSE3-NEXT:    setns %bl
-; SSSE3-NEXT:    testq %rcx, %rcx
-; SSSE3-NEXT:    setns %cl
-; SSSE3-NEXT:    cmpb %bl, %cl
-; SSSE3-NEXT:    setne %r8b
-; SSSE3-NEXT:    testq %r10, %r10
-; SSSE3-NEXT:    setns %bl
-; SSSE3-NEXT:    cmpb %bl, %cl
-; SSSE3-NEXT:    sete %cl
-; SSSE3-NEXT:    andb %r8b, %cl
-; SSSE3-NEXT:    movzbl %cl, %ecx
-; SSSE3-NEXT:    negl %ecx
-; SSSE3-NEXT:    movd %ecx, %xmm1
+; SSSE3-NEXT:    adcq {{[0-9]+}}(%rsp), %rcx
+; SSSE3-NEXT:    seto %al
 ; SSSE3-NEXT:    movzbl %al, %eax
 ; SSSE3-NEXT:    negl %eax
+; SSSE3-NEXT:    movd %eax, %xmm1
+; SSSE3-NEXT:    movzbl %r8b, %eax
+; SSSE3-NEXT:    negl %eax
 ; SSSE3-NEXT:    movd %eax, %xmm0
 ; SSSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; SSSE3-NEXT:    movq %rdx, 16(%r11)
-; SSSE3-NEXT:    movq %rdi, (%r11)
-; SSSE3-NEXT:    movq %rbp, 24(%r11)
-; SSSE3-NEXT:    movq %rsi, 8(%r11)
-; SSSE3-NEXT:    popq %rbx
-; SSSE3-NEXT:    popq %rbp
+; SSSE3-NEXT:    movq %rdx, 16(%r10)
+; SSSE3-NEXT:    movq %rdi, (%r10)
+; SSSE3-NEXT:    movq %rcx, 24(%r10)
+; SSSE3-NEXT:    movq %rsi, 8(%r10)
 ; SSSE3-NEXT:    retq
 ;
 ; SSE41-LABEL: saddo_v2i128:
 ; SSE41:       # %bb.0:
-; SSE41-NEXT:    pushq %rbp
-; SSE41-NEXT:    pushq %rbx
-; SSE41-NEXT:    movq {{[0-9]+}}(%rsp), %r11
 ; SSE41-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; SSE41-NEXT:    testq %r9, %r9
-; SSE41-NEXT:    setns %al
-; SSE41-NEXT:    testq %rsi, %rsi
-; SSE41-NEXT:    setns %bl
-; SSE41-NEXT:    cmpb %al, %bl
-; SSE41-NEXT:    sete %bpl
 ; SSE41-NEXT:    addq %r8, %rdi
 ; SSE41-NEXT:    adcq %r9, %rsi
-; SSE41-NEXT:    setns %al
-; SSE41-NEXT:    cmpb %al, %bl
-; SSE41-NEXT:    setne %al
-; SSE41-NEXT:    andb %bpl, %al
+; SSE41-NEXT:    seto %r8b
 ; SSE41-NEXT:    addq {{[0-9]+}}(%rsp), %rdx
-; SSE41-NEXT:    movq %rcx, %rbp
-; SSE41-NEXT:    adcq %r10, %rbp
-; SSE41-NEXT:    setns %bl
-; SSE41-NEXT:    testq %rcx, %rcx
-; SSE41-NEXT:    setns %cl
-; SSE41-NEXT:    cmpb %bl, %cl
-; SSE41-NEXT:    setne %r8b
-; SSE41-NEXT:    testq %r10, %r10
-; SSE41-NEXT:    setns %bl
-; SSE41-NEXT:    cmpb %bl, %cl
-; SSE41-NEXT:    sete %cl
-; SSE41-NEXT:    andb %r8b, %cl
-; SSE41-NEXT:    movzbl %cl, %ecx
-; SSE41-NEXT:    negl %ecx
-; SSE41-NEXT:    movzbl %al, %eax
+; SSE41-NEXT:    adcq {{[0-9]+}}(%rsp), %rcx
+; SSE41-NEXT:    seto %al
+; SSE41-NEXT:    movzbl %al, %r9d
+; SSE41-NEXT:    negl %r9d
+; SSE41-NEXT:    movzbl %r8b, %eax
 ; SSE41-NEXT:    negl %eax
 ; SSE41-NEXT:    movd %eax, %xmm0
-; SSE41-NEXT:    pinsrd $1, %ecx, %xmm0
-; SSE41-NEXT:    movq %rdx, 16(%r11)
-; SSE41-NEXT:    movq %rdi, (%r11)
-; SSE41-NEXT:    movq %rbp, 24(%r11)
-; SSE41-NEXT:    movq %rsi, 8(%r11)
-; SSE41-NEXT:    popq %rbx
-; SSE41-NEXT:    popq %rbp
+; SSE41-NEXT:    pinsrd $1, %r9d, %xmm0
+; SSE41-NEXT:    movq %rdx, 16(%r10)
+; SSE41-NEXT:    movq %rdi, (%r10)
+; SSE41-NEXT:    movq %rcx, 24(%r10)
+; SSE41-NEXT:    movq %rsi, 8(%r10)
 ; SSE41-NEXT:    retq
 ;
 ; AVX1-LABEL: saddo_v2i128:
 ; AVX1:       # %bb.0:
-; AVX1-NEXT:    pushq %rbp
-; AVX1-NEXT:    pushq %rbx
-; AVX1-NEXT:    movq {{[0-9]+}}(%rsp), %r11
 ; AVX1-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX1-NEXT:    testq %r9, %r9
-; AVX1-NEXT:    setns %al
-; AVX1-NEXT:    testq %rsi, %rsi
-; AVX1-NEXT:    setns %bl
-; AVX1-NEXT:    cmpb %al, %bl
-; AVX1-NEXT:    sete %bpl
 ; AVX1-NEXT:    addq %r8, %rdi
 ; AVX1-NEXT:    adcq %r9, %rsi
-; AVX1-NEXT:    setns %al
-; AVX1-NEXT:    cmpb %al, %bl
-; AVX1-NEXT:    setne %al
-; AVX1-NEXT:    andb %bpl, %al
+; AVX1-NEXT:    seto %r8b
 ; AVX1-NEXT:    addq {{[0-9]+}}(%rsp), %rdx
-; AVX1-NEXT:    movq %rcx, %rbp
-; AVX1-NEXT:    adcq %r10, %rbp
-; AVX1-NEXT:    setns %bl
-; AVX1-NEXT:    testq %rcx, %rcx
-; AVX1-NEXT:    setns %cl
-; AVX1-NEXT:    cmpb %bl, %cl
-; AVX1-NEXT:    setne %r8b
-; AVX1-NEXT:    testq %r10, %r10
-; AVX1-NEXT:    setns %bl
-; AVX1-NEXT:    cmpb %bl, %cl
-; AVX1-NEXT:    sete %cl
-; AVX1-NEXT:    andb %r8b, %cl
-; AVX1-NEXT:    movzbl %cl, %ecx
-; AVX1-NEXT:    negl %ecx
-; AVX1-NEXT:    movzbl %al, %eax
+; AVX1-NEXT:    adcq {{[0-9]+}}(%rsp), %rcx
+; AVX1-NEXT:    seto %al
+; AVX1-NEXT:    movzbl %al, %r9d
+; AVX1-NEXT:    negl %r9d
+; AVX1-NEXT:    movzbl %r8b, %eax
 ; AVX1-NEXT:    negl %eax
 ; AVX1-NEXT:    vmovd %eax, %xmm0
-; AVX1-NEXT:    vpinsrd $1, %ecx, %xmm0, %xmm0
-; AVX1-NEXT:    movq %rdx, 16(%r11)
-; AVX1-NEXT:    movq %rdi, (%r11)
-; AVX1-NEXT:    movq %rbp, 24(%r11)
-; AVX1-NEXT:    movq %rsi, 8(%r11)
-; AVX1-NEXT:    popq %rbx
-; AVX1-NEXT:    popq %rbp
+; AVX1-NEXT:    vpinsrd $1, %r9d, %xmm0, %xmm0
+; AVX1-NEXT:    movq %rdx, 16(%r10)
+; AVX1-NEXT:    movq %rdi, (%r10)
+; AVX1-NEXT:    movq %rcx, 24(%r10)
+; AVX1-NEXT:    movq %rsi, 8(%r10)
 ; AVX1-NEXT:    retq
 ;
 ; AVX2-LABEL: saddo_v2i128:
 ; AVX2:       # %bb.0:
-; AVX2-NEXT:    pushq %rbp
-; AVX2-NEXT:    pushq %rbx
-; AVX2-NEXT:    movq {{[0-9]+}}(%rsp), %r11
 ; AVX2-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX2-NEXT:    testq %r9, %r9
-; AVX2-NEXT:    setns %al
-; AVX2-NEXT:    testq %rsi, %rsi
-; AVX2-NEXT:    setns %bl
-; AVX2-NEXT:    cmpb %al, %bl
-; AVX2-NEXT:    sete %bpl
 ; AVX2-NEXT:    addq %r8, %rdi
 ; AVX2-NEXT:    adcq %r9, %rsi
-; AVX2-NEXT:    setns %al
-; AVX2-NEXT:    cmpb %al, %bl
-; AVX2-NEXT:    setne %al
-; AVX2-NEXT:    andb %bpl, %al
+; AVX2-NEXT:    seto %r8b
 ; AVX2-NEXT:    addq {{[0-9]+}}(%rsp), %rdx
-; AVX2-NEXT:    movq %rcx, %rbp
-; AVX2-NEXT:    adcq %r10, %rbp
-; AVX2-NEXT:    setns %bl
-; AVX2-NEXT:    testq %rcx, %rcx
-; AVX2-NEXT:    setns %cl
-; AVX2-NEXT:    cmpb %bl, %cl
-; AVX2-NEXT:    setne %r8b
-; AVX2-NEXT:    testq %r10, %r10
-; AVX2-NEXT:    setns %bl
-; AVX2-NEXT:    cmpb %bl, %cl
-; AVX2-NEXT:    sete %cl
-; AVX2-NEXT:    andb %r8b, %cl
-; AVX2-NEXT:    movzbl %cl, %ecx
-; AVX2-NEXT:    negl %ecx
-; AVX2-NEXT:    movzbl %al, %eax
+; AVX2-NEXT:    adcq {{[0-9]+}}(%rsp), %rcx
+; AVX2-NEXT:    seto %al
+; AVX2-NEXT:    movzbl %al, %r9d
+; AVX2-NEXT:    negl %r9d
+; AVX2-NEXT:    movzbl %r8b, %eax
 ; AVX2-NEXT:    negl %eax
 ; AVX2-NEXT:    vmovd %eax, %xmm0
-; AVX2-NEXT:    vpinsrd $1, %ecx, %xmm0, %xmm0
-; AVX2-NEXT:    movq %rdx, 16(%r11)
-; AVX2-NEXT:    movq %rdi, (%r11)
-; AVX2-NEXT:    movq %rbp, 24(%r11)
-; AVX2-NEXT:    movq %rsi, 8(%r11)
-; AVX2-NEXT:    popq %rbx
-; AVX2-NEXT:    popq %rbp
+; AVX2-NEXT:    vpinsrd $1, %r9d, %xmm0, %xmm0
+; AVX2-NEXT:    movq %rdx, 16(%r10)
+; AVX2-NEXT:    movq %rdi, (%r10)
+; AVX2-NEXT:    movq %rcx, 24(%r10)
+; AVX2-NEXT:    movq %rsi, 8(%r10)
 ; AVX2-NEXT:    retq
 ;
 ; AVX512-LABEL: saddo_v2i128:
 ; AVX512:       # %bb.0:
-; AVX512-NEXT:    pushq %r14
-; AVX512-NEXT:    pushq %rbx
 ; AVX512-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512-NEXT:    movq {{[0-9]+}}(%rsp), %r11
 ; AVX512-NEXT:    addq {{[0-9]+}}(%rsp), %rdx
-; AVX512-NEXT:    movq %rcx, %r14
-; AVX512-NEXT:    adcq %r11, %r14
-; AVX512-NEXT:    setns %bl
-; AVX512-NEXT:    testq %rcx, %rcx
-; AVX512-NEXT:    setns %cl
-; AVX512-NEXT:    cmpb %bl, %cl
-; AVX512-NEXT:    setne %bl
-; AVX512-NEXT:    testq %r11, %r11
-; AVX512-NEXT:    setns %al
-; AVX512-NEXT:    cmpb %al, %cl
-; AVX512-NEXT:    sete %al
-; AVX512-NEXT:    andb %bl, %al
+; AVX512-NEXT:    adcq {{[0-9]+}}(%rsp), %rcx
+; AVX512-NEXT:    seto %al
 ; AVX512-NEXT:    kmovd %eax, %k0
-; AVX512-NEXT:    testq %r9, %r9
-; AVX512-NEXT:    setns %al
-; AVX512-NEXT:    testq %rsi, %rsi
-; AVX512-NEXT:    setns %cl
-; AVX512-NEXT:    cmpb %al, %cl
-; AVX512-NEXT:    sete %al
 ; AVX512-NEXT:    addq %r8, %rdi
 ; AVX512-NEXT:    adcq %r9, %rsi
-; AVX512-NEXT:    setns %bl
-; AVX512-NEXT:    cmpb %bl, %cl
-; AVX512-NEXT:    setne %cl
-; AVX512-NEXT:    andb %al, %cl
-; AVX512-NEXT:    andl $1, %ecx
-; AVX512-NEXT:    kmovw %ecx, %k1
+; AVX512-NEXT:    seto %al
+; AVX512-NEXT:    andl $1, %eax
+; AVX512-NEXT:    kmovw %eax, %k1
 ; AVX512-NEXT:    kshiftlw $1, %k0, %k0
 ; AVX512-NEXT:    korw %k0, %k1, %k1
 ; AVX512-NEXT:    vpcmpeqd %xmm0, %xmm0, %xmm0
 ; AVX512-NEXT:    vmovdqa32 %xmm0, %xmm0 {%k1} {z}
 ; AVX512-NEXT:    movq %rdx, 16(%r10)
 ; AVX512-NEXT:    movq %rdi, (%r10)
-; AVX512-NEXT:    movq %r14, 24(%r10)
+; AVX512-NEXT:    movq %rcx, 24(%r10)
 ; AVX512-NEXT:    movq %rsi, 8(%r10)
-; AVX512-NEXT:    popq %rbx
-; AVX512-NEXT:    popq %r14
 ; AVX512-NEXT:    retq
   %t = call {<2 x i128>, <2 x i1>} @llvm.sadd.with.overflow.v2i128(<2 x i128> %a0, <2 x i128> %a1)
   %val = extractvalue {<2 x i128>, <2 x i1>} %t, 0
   %obit = extractvalue {<2 x i128>, <2 x i1>} %t, 1
   %res = sext <2 x i1> %obit to <2 x i32>
   store <2 x i128> %val, <2 x i128>* %p2
   ret <2 x i32> %res
 }

llvm/test/CodeGen/X86/vec_ssubo.ll

; AVX512-NEXT:    retq
   %res = sext <4 x i1> %obit to <4 x i32>
   store <4 x i1> %val, <4 x i1>* %p2
   ret <4 x i32> %res
 }
 
 define <2 x i32> @ssubo_v2i128(<2 x i128> %a0, <2 x i128> %a1, <2 x i128>* %p2) nounwind {
 ; SSE2-LABEL: ssubo_v2i128:
 ; SSE2:       # %bb.0:
-; SSE2-NEXT:    pushq %rbp
-; SSE2-NEXT:    pushq %rbx
-; SSE2-NEXT:    movq {{[0-9]+}}(%rsp), %r11
 ; SSE2-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; SSE2-NEXT:    testq %r9, %r9
-; SSE2-NEXT:    setns %al
-; SSE2-NEXT:    testq %rsi, %rsi
-; SSE2-NEXT:    setns %bl
-; SSE2-NEXT:    cmpb %al, %bl
-; SSE2-NEXT:    setne %bpl
 ; SSE2-NEXT:    subq %r8, %rdi
 ; SSE2-NEXT:    sbbq %r9, %rsi
-; SSE2-NEXT:    setns %al
-; SSE2-NEXT:    cmpb %al, %bl
-; SSE2-NEXT:    setne %al
-; SSE2-NEXT:    andb %bpl, %al
+; SSE2-NEXT:    seto %r8b
 ; SSE2-NEXT:    subq {{[0-9]+}}(%rsp), %rdx
-; SSE2-NEXT:    movq %rcx, %rbp
-; SSE2-NEXT:    sbbq %r10, %rbp
-; SSE2-NEXT:    setns %bl
-; SSE2-NEXT:    testq %rcx, %rcx
-; SSE2-NEXT:    setns %cl
-; SSE2-NEXT:    cmpb %bl, %cl
-; SSE2-NEXT:    setne %r8b
-; SSE2-NEXT:    testq %r10, %r10
-; SSE2-NEXT:    setns %bl
-; SSE2-NEXT:    cmpb %bl, %cl
-; SSE2-NEXT:    setne %cl
-; SSE2-NEXT:    andb %r8b, %cl
-; SSE2-NEXT:    movzbl %cl, %ecx
-; SSE2-NEXT:    negl %ecx
-; SSE2-NEXT:    movd %ecx, %xmm1
+; SSE2-NEXT:    sbbq {{[0-9]+}}(%rsp), %rcx
+; SSE2-NEXT:    seto %al
 ; SSE2-NEXT:    movzbl %al, %eax
 ; SSE2-NEXT:    negl %eax
+; SSE2-NEXT:    movd %eax, %xmm1
+; SSE2-NEXT:    movzbl %r8b, %eax
+; SSE2-NEXT:    negl %eax
 ; SSE2-NEXT:    movd %eax, %xmm0
 ; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; SSE2-NEXT:    movq %rdx, 16(%r11)
-; SSE2-NEXT:    movq %rdi, (%r11)
-; SSE2-NEXT:    movq %rbp, 24(%r11)
-; SSE2-NEXT:    movq %rsi, 8(%r11)
-; SSE2-NEXT:    popq %rbx
-; SSE2-NEXT:    popq %rbp
+; SSE2-NEXT:    movq %rdx, 16(%r10)
+; SSE2-NEXT:    movq %rdi, (%r10)
+; SSE2-NEXT:    movq %rcx, 24(%r10)
+; SSE2-NEXT:    movq %rsi, 8(%r10)
 ; SSE2-NEXT:    retq
 ;
 ; SSSE3-LABEL: ssubo_v2i128:
 ; SSSE3:       # %bb.0:
-; SSSE3-NEXT:    pushq %rbp
-; SSSE3-NEXT:    pushq %rbx
-; SSSE3-NEXT:    movq {{[0-9]+}}(%rsp), %r11
 ; SSSE3-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; SSSE3-NEXT:    testq %r9, %r9
-; SSSE3-NEXT:    setns %al
-; SSSE3-NEXT:    testq %rsi, %rsi
-; SSSE3-NEXT:    setns %bl
-; SSSE3-NEXT:    cmpb %al, %bl
-; SSSE3-NEXT:    setne %bpl
 ; SSSE3-NEXT:    subq %r8, %rdi
 ; SSSE3-NEXT:    sbbq %r9, %rsi
-; SSSE3-NEXT:    setns %al
-; SSSE3-NEXT:    cmpb %al, %bl
-; SSSE3-NEXT:    setne %al
-; SSSE3-NEXT:    andb %bpl, %al
+; SSSE3-NEXT:    seto %r8b
 ; SSSE3-NEXT:    subq {{[0-9]+}}(%rsp), %rdx
-; SSSE3-NEXT:    movq %rcx, %rbp
-; SSSE3-NEXT:    sbbq %r10, %rbp
-; SSSE3-NEXT:    setns %bl
-; SSSE3-NEXT:    testq %rcx, %rcx
-; SSSE3-NEXT:    setns %cl
-; SSSE3-NEXT:    cmpb %bl, %cl
-; SSSE3-NEXT:    setne %r8b
-; SSSE3-NEXT:    testq %r10, %r10
-; SSSE3-NEXT:    setns %bl
-; SSSE3-NEXT:    cmpb %bl, %cl
-; SSSE3-NEXT:    setne %cl
-; SSSE3-NEXT:    andb %r8b, %cl
-; SSSE3-NEXT:    movzbl %cl, %ecx
-; SSSE3-NEXT:    negl %ecx
-; SSSE3-NEXT:    movd %ecx, %xmm1
+; SSSE3-NEXT:    sbbq {{[0-9]+}}(%rsp), %rcx
+; SSSE3-NEXT:    seto %al
 ; SSSE3-NEXT:    movzbl %al, %eax
 ; SSSE3-NEXT:    negl %eax
+; SSSE3-NEXT:    movd %eax, %xmm1
+; SSSE3-NEXT:    movzbl %r8b, %eax
+; SSSE3-NEXT:    negl %eax
 ; SSSE3-NEXT:    movd %eax, %xmm0
 ; SSSE3-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; SSSE3-NEXT:    movq %rdx, 16(%r11)
-; SSSE3-NEXT:    movq %rdi, (%r11)
-; SSSE3-NEXT:    movq %rbp, 24(%r11)
-; SSSE3-NEXT:    movq %rsi, 8(%r11)
-; SSSE3-NEXT:    popq %rbx
-; SSSE3-NEXT:    popq %rbp
+; SSSE3-NEXT:    movq %rdx, 16(%r10)
+; SSSE3-NEXT:    movq %rdi, (%r10)
+; SSSE3-NEXT:    movq %rcx, 24(%r10)
+; SSSE3-NEXT:    movq %rsi, 8(%r10)
 ; SSSE3-NEXT:    retq
 ;
 ; SSE41-LABEL: ssubo_v2i128:
 ; SSE41:       # %bb.0:
-; SSE41-NEXT:    pushq %rbp
-; SSE41-NEXT:    pushq %rbx
-; SSE41-NEXT:    movq {{[0-9]+}}(%rsp), %r11
 ; SSE41-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; SSE41-NEXT:    testq %r9, %r9
-; SSE41-NEXT:    setns %al
-; SSE41-NEXT:    testq %rsi, %rsi
-; SSE41-NEXT:    setns %bl
-; SSE41-NEXT:    cmpb %al, %bl
-; SSE41-NEXT:    setne %bpl
 ; SSE41-NEXT:    subq %r8, %rdi
 ; SSE41-NEXT:    sbbq %r9, %rsi
-; SSE41-NEXT:    setns %al
-; SSE41-NEXT:    cmpb %al, %bl
-; SSE41-NEXT:    setne %al
-; SSE41-NEXT:    andb %bpl, %al
+; SSE41-NEXT:    seto %r8b
 ; SSE41-NEXT:    subq {{[0-9]+}}(%rsp), %rdx
-; SSE41-NEXT:    movq %rcx, %rbp
-; SSE41-NEXT:    sbbq %r10, %rbp
-; SSE41-NEXT:    setns %bl
-; SSE41-NEXT:    testq %rcx, %rcx
-; SSE41-NEXT:    setns %cl
-; SSE41-NEXT:    cmpb %bl, %cl
-; SSE41-NEXT:    setne %r8b
-; SSE41-NEXT:    testq %r10, %r10
-; SSE41-NEXT:    setns %bl
-; SSE41-NEXT:    cmpb %bl, %cl
-; SSE41-NEXT:    setne %cl
-; SSE41-NEXT:    andb %r8b, %cl
-; SSE41-NEXT:    movzbl %cl, %ecx
-; SSE41-NEXT:    negl %ecx
-; SSE41-NEXT:    movzbl %al, %eax
+; SSE41-NEXT:    sbbq {{[0-9]+}}(%rsp), %rcx
+; SSE41-NEXT:    seto %al
+; SSE41-NEXT:    movzbl %al, %r9d
+; SSE41-NEXT:    negl %r9d
+; SSE41-NEXT:    movzbl %r8b, %eax
 ; SSE41-NEXT:    negl %eax
 ; SSE41-NEXT:    movd %eax, %xmm0
-; SSE41-NEXT:    pinsrd $1, %ecx, %xmm0
-; SSE41-NEXT:    movq %rdx, 16(%r11)
-; SSE41-NEXT:    movq %rdi, (%r11)
-; SSE41-NEXT:    movq %rbp, 24(%r11)
-; SSE41-NEXT:    movq %rsi, 8(%r11)
-; SSE41-NEXT:    popq %rbx
-; SSE41-NEXT:    popq %rbp
+; SSE41-NEXT:    pinsrd $1, %r9d, %xmm0
+; SSE41-NEXT:    movq %rdx, 16(%r10)
+; SSE41-NEXT:    movq %rdi, (%r10)
+; SSE41-NEXT:    movq %rcx, 24(%r10)
+; SSE41-NEXT:    movq %rsi, 8(%r10)
 ; SSE41-NEXT:    retq
 ;
 ; AVX1-LABEL: ssubo_v2i128:
 ; AVX1:       # %bb.0:
-; AVX1-NEXT:    pushq %rbp
-; AVX1-NEXT:    pushq %rbx
-; AVX1-NEXT:    movq {{[0-9]+}}(%rsp), %r11
 ; AVX1-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX1-NEXT:    testq %r9, %r9
-; AVX1-NEXT:    setns %al
-; AVX1-NEXT:    testq %rsi, %rsi
-; AVX1-NEXT:    setns %bl
-; AVX1-NEXT:    cmpb %al, %bl
-; AVX1-NEXT:    setne %bpl
 ; AVX1-NEXT:    subq %r8, %rdi
 ; AVX1-NEXT:    sbbq %r9, %rsi
-; AVX1-NEXT:    setns %al
-; AVX1-NEXT:    cmpb %al, %bl
-; AVX1-NEXT:    setne %al
-; AVX1-NEXT:    andb %bpl, %al
+; AVX1-NEXT:    seto %r8b
 ; AVX1-NEXT:    subq {{[0-9]+}}(%rsp), %rdx
-; AVX1-NEXT:    movq %rcx, %rbp
-; AVX1-NEXT:    sbbq %r10, %rbp
-; AVX1-NEXT:    setns %bl
-; AVX1-NEXT:    testq %rcx, %rcx
-; AVX1-NEXT:    setns %cl
-; AVX1-NEXT:    cmpb %bl, %cl
-; AVX1-NEXT:    setne %r8b
-; AVX1-NEXT:    testq %r10, %r10
-; AVX1-NEXT:    setns %bl
-; AVX1-NEXT:    cmpb %bl, %cl
-; AVX1-NEXT:    setne %cl
-; AVX1-NEXT:    andb %r8b, %cl
-; AVX1-NEXT:    movzbl %cl, %ecx
-; AVX1-NEXT:    negl %ecx
-; AVX1-NEXT:    movzbl %al, %eax
+; AVX1-NEXT:    sbbq {{[0-9]+}}(%rsp), %rcx
+; AVX1-NEXT:    seto %al
+; AVX1-NEXT:    movzbl %al, %r9d
+; AVX1-NEXT:    negl %r9d
+; AVX1-NEXT:    movzbl %r8b, %eax
 ; AVX1-NEXT:    negl %eax
 ; AVX1-NEXT:    vmovd %eax, %xmm0
-; AVX1-NEXT:    vpinsrd $1, %ecx, %xmm0, %xmm0
-; AVX1-NEXT:    movq %rdx, 16(%r11)
-; AVX1-NEXT:    movq %rdi, (%r11)
-; AVX1-NEXT:    movq %rbp, 24(%r11)
-; AVX1-NEXT:    movq %rsi, 8(%r11)
-; AVX1-NEXT:    popq %rbx
-; AVX1-NEXT:    popq %rbp
+; AVX1-NEXT:    vpinsrd $1, %r9d, %xmm0, %xmm0
+; AVX1-NEXT:    movq %rdx, 16(%r10)
+; AVX1-NEXT:    movq %rdi, (%r10)
+; AVX1-NEXT:    movq %rcx, 24(%r10)
+; AVX1-NEXT:    movq %rsi, 8(%r10)
 ; AVX1-NEXT:    retq
 ;
 ; AVX2-LABEL: ssubo_v2i128:
 ; AVX2:       # %bb.0:
-; AVX2-NEXT:    pushq %rbp
-; AVX2-NEXT:    pushq %rbx
-; AVX2-NEXT:    movq {{[0-9]+}}(%rsp), %r11
 ; AVX2-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX2-NEXT:    testq %r9, %r9
-; AVX2-NEXT:    setns %al
-; AVX2-NEXT:    testq %rsi, %rsi
-; AVX2-NEXT:    setns %bl
-; AVX2-NEXT:    cmpb %al, %bl
-; AVX2-NEXT:    setne %bpl
 ; AVX2-NEXT:    subq %r8, %rdi
 ; AVX2-NEXT:    sbbq %r9, %rsi
-; AVX2-NEXT:    setns %al
-; AVX2-NEXT:    cmpb %al, %bl
-; AVX2-NEXT:    setne %al
-; AVX2-NEXT:    andb %bpl, %al
+; AVX2-NEXT:    seto %r8b
 ; AVX2-NEXT:    subq {{[0-9]+}}(%rsp), %rdx
-; AVX2-NEXT:    movq %rcx, %rbp
-; AVX2-NEXT:    sbbq %r10, %rbp
-; AVX2-NEXT:    setns %bl
-; AVX2-NEXT:    testq %rcx, %rcx
-; AVX2-NEXT:    setns %cl
-; AVX2-NEXT:    cmpb %bl, %cl
-; AVX2-NEXT:    setne %r8b
-; AVX2-NEXT:    testq %r10, %r10
-; AVX2-NEXT:    setns %bl
-; AVX2-NEXT:    cmpb %bl, %cl
-; AVX2-NEXT:    setne %cl
-; AVX2-NEXT:    andb %r8b, %cl
-; AVX2-NEXT:    movzbl %cl, %ecx
-; AVX2-NEXT:    negl %ecx
-; AVX2-NEXT:    movzbl %al, %eax
+; AVX2-NEXT:    sbbq {{[0-9]+}}(%rsp), %rcx
+; AVX2-NEXT:    seto %al
+; AVX2-NEXT:    movzbl %al, %r9d
+; AVX2-NEXT:    negl %r9d
+; AVX2-NEXT:    movzbl %r8b, %eax
 ; AVX2-NEXT:    negl %eax
 ; AVX2-NEXT:    vmovd %eax, %xmm0
-; AVX2-NEXT:    vpinsrd $1, %ecx, %xmm0, %xmm0
-; AVX2-NEXT:    movq %rdx, 16(%r11)
-; AVX2-NEXT:    movq %rdi, (%r11)
-; AVX2-NEXT:    movq %rbp, 24(%r11)
-; AVX2-NEXT:    movq %rsi, 8(%r11)
-; AVX2-NEXT:    popq %rbx
-; AVX2-NEXT:    popq %rbp
+; AVX2-NEXT:    vpinsrd $1, %r9d, %xmm0, %xmm0
+; AVX2-NEXT:    movq %rdx, 16(%r10)
+; AVX2-NEXT:    movq %rdi, (%r10)
+; AVX2-NEXT:    movq %rcx, 24(%r10)
+; AVX2-NEXT:    movq %rsi, 8(%r10)
 ; AVX2-NEXT:    retq
 ;
 ; AVX512-LABEL: ssubo_v2i128:
 ; AVX512:       # %bb.0:
-; AVX512-NEXT:    pushq %r14
-; AVX512-NEXT:    pushq %rbx
 ; AVX512-NEXT:    movq {{[0-9]+}}(%rsp), %r10
-; AVX512-NEXT:    movq {{[0-9]+}}(%rsp), %r11
 ; AVX512-NEXT:    subq {{[0-9]+}}(%rsp), %rdx
-; AVX512-NEXT:    movq %rcx, %r14
-; AVX512-NEXT:    sbbq %r11, %r14
-; AVX512-NEXT:    setns %bl
-; AVX512-NEXT:    testq %rcx, %rcx
-; AVX512-NEXT:    setns %cl
-; AVX512-NEXT:    cmpb %bl, %cl
-; AVX512-NEXT:    setne %bl
-; AVX512-NEXT:    testq %r11, %r11
-; AVX512-NEXT:    setns %al
-; AVX512-NEXT:    cmpb %al, %cl
-; AVX512-NEXT:    setne %al
-; AVX512-NEXT:    andb %bl, %al
+; AVX512-NEXT:    sbbq {{[0-9]+}}(%rsp), %rcx
+; AVX512-NEXT:    seto %al
 ; AVX512-NEXT:    kmovd %eax, %k0
-; AVX512-NEXT:    testq %r9, %r9
-; AVX512-NEXT:    setns %al
-; AVX512-NEXT:    testq %rsi, %rsi
-; AVX512-NEXT:    setns %cl
-; AVX512-NEXT:    cmpb %al, %cl
-; AVX512-NEXT:    setne %al
 ; AVX512-NEXT:    subq %r8, %rdi
 ; AVX512-NEXT:    sbbq %r9, %rsi
-; AVX512-NEXT:    setns %bl
-; AVX512-NEXT:    cmpb %bl, %cl
-; AVX512-NEXT:    setne %cl
-; AVX512-NEXT:    andb %al, %cl
-; AVX512-NEXT:    andl $1, %ecx
-; AVX512-NEXT:    kmovw %ecx, %k1
+; AVX512-NEXT:    seto %al
+; AVX512-NEXT:    andl $1, %eax
+; AVX512-NEXT:    kmovw %eax, %k1
 ; AVX512-NEXT:    kshiftlw $1, %k0, %k0
 ; AVX512-NEXT:    korw %k0, %k1, %k1
 ; AVX512-NEXT:    vpcmpeqd %xmm0, %xmm0, %xmm0
 ; AVX512-NEXT:    vmovdqa32 %xmm0, %xmm0 {%k1} {z}
 ; AVX512-NEXT:    movq %rdx, 16(%r10)
 ; AVX512-NEXT:    movq %rdi, (%r10)
-; AVX512-NEXT:    movq %r14, 24(%r10)
+; AVX512-NEXT:    movq %rcx, 24(%r10)
 ; AVX512-NEXT:    movq %rsi, 8(%r10)
-; AVX512-NEXT:    popq %rbx
-; AVX512-NEXT:    popq %r14
 ; AVX512-NEXT:    retq
   %t = call {<2 x i128>, <2 x i1>} @llvm.ssub.with.overflow.v2i128(<2 x i128> %a0, <2 x i128> %a1)
   %val = extractvalue {<2 x i128>, <2 x i1>} %t, 0
   %obit = extractvalue {<2 x i128>, <2 x i1>} %t, 1
   %res = sext <2 x i1> %obit to <2 x i32>
   store <2 x i128> %val, <2 x i128>* %p2
   ret <2 x i32> %res
 }