[LegalizeTypes] Improve ExpandIntRes_XMULO codegen.

[craig.topper]

Summary

The code previously used two BUILD_PAIRs to concatenate the two UMULO
results with 0s in the lower bits to match original VT. Then it created
an ADD and a UADDO with the original bit width. Each of those operations
need to be expanded since they have illegal types.

Since we put 0s in the lower bits before the ADD, the lower half of the
ADD result will be 0. So the lower half of the UADDO result is
solely determined by the other operand. Since the UADDO need to
be split in half, we don't really needd an operation for the lower
bits. Unfortunately, we don't see that in type legalization and end up
creating something more complicated and DAG combine or
lowering aren't always able to recover it.

This patch directly generates the narrower ADD and UADDO to avoid
needing to legalize them. Now only the MUL is done on the original
type.

[RKSimon]

LGTM

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
 //
 //===----------------------------------------------------------------------===//
 //
if (N->getOpcode() == ISD::UMULO) {
     // This section expands the operation into the following sequence of
     // instructions. `iNh` here refers to a type which has half the bit width of
     // the type the original operation operated on.
     //
     // %0 = %LHS.HI != 0 && %RHS.HI != 0
     // %1 = { iNh, i1 } @umul.with.overflow.iNh(iNh %LHS.HI, iNh %RHS.LO)
     // %2 = { iNh, i1 } @umul.with.overflow.iNh(iNh %RHS.HI, iNh %LHS.LO)
     // %3 = mul nuw iN (%LHS.LOW as iN), (%RHS.LOW as iN)
-    // %4 = add iN (%1.0 as iN) << Nh, (%2.0 as iN) << Nh
-    // %5 = { iN, i1 } @uadd.with.overflow.iN( %4, %3 )
+    // %4 = add iNh %1.0, %2.0 as iN
+    // %5 = { iNh, i1 } @uadd.with.overflow.iNh(iNh %4, iNh %3.HIGH)
     //
-    // %res = { %5.0, %0 || %1.1 || %2.1 || %5.1 }
+    // %lo = %3.LO
+    // %hi = %5.0
+    // %ovf = %0 || %1.1 || %2.1 || %5.1
     SDValue LHS = N->getOperand(0), RHS = N->getOperand(1);
     SDValue LHSHigh, LHSLow, RHSHigh, RHSLow;
     GetExpandedInteger(LHS, LHSLow, LHSHigh);
     GetExpandedInteger(RHS, RHSLow, RHSHigh);
     EVT HalfVT = LHSLow.getValueType();
     EVT BitVT = N->getValueType(1);
-    SDVTList VTHalfMulO = DAG.getVTList(HalfVT, BitVT);
-    SDVTList VTFullAddO = DAG.getVTList(VT, BitVT);
+    SDVTList VTHalfWithO = DAG.getVTList(HalfVT, BitVT);
 
     SDValue HalfZero = DAG.getConstant(0, dl, HalfVT);
     SDValue Overflow = DAG.getNode(ISD::AND, dl, BitVT,
       DAG.getSetCC(dl, BitVT, LHSHigh, HalfZero, ISD::SETNE),
       DAG.getSetCC(dl, BitVT, RHSHigh, HalfZero, ISD::SETNE));
 
-    SDValue One = DAG.getNode(ISD::UMULO, dl, VTHalfMulO, LHSHigh, RHSLow);
+    SDValue One = DAG.getNode(ISD::UMULO, dl, VTHalfWithO, LHSHigh, RHSLow);
     Overflow = DAG.getNode(ISD::OR, dl, BitVT, Overflow, One.getValue(1));
-    SDValue OneInHigh = DAG.getNode(ISD::BUILD_PAIR, dl, VT, HalfZero,
-                                    One.getValue(0));
 
-    SDValue Two = DAG.getNode(ISD::UMULO, dl, VTHalfMulO, RHSHigh, LHSLow);
+    SDValue Two = DAG.getNode(ISD::UMULO, dl, VTHalfWithO, RHSHigh, LHSLow);
     Overflow = DAG.getNode(ISD::OR, dl, BitVT, Overflow, Two.getValue(1));
-    SDValue TwoInHigh = DAG.getNode(ISD::BUILD_PAIR, dl, VT, HalfZero,
-                                    Two.getValue(0));
+
+    SDValue HighSum = DAG.getNode(ISD::ADD, dl, HalfVT, One, Two);
 
     // Cannot use `UMUL_LOHI` directly, because some 32-bit targets (ARM) do not
     // know how to expand `i64,i64 = umul_lohi a, b` and abort (why isn’t this
     // operation recursively legalized?).
     //
     // Many backends understand this pattern and will convert into LOHI
     // themselves, if applicable.
     SDValue Three = DAG.getNode(ISD::MUL, dl, VT,
       DAG.getNode(ISD::ZERO_EXTEND, dl, VT, LHSLow),
       DAG.getNode(ISD::ZERO_EXTEND, dl, VT, RHSLow));
-    SDValue Four = DAG.getNode(ISD::ADD, dl, VT, OneInHigh, TwoInHigh);
-    SDValue Five = DAG.getNode(ISD::UADDO, dl, VTFullAddO, Three, Four);
-    Overflow = DAG.getNode(ISD::OR, dl, BitVT, Overflow, Five.getValue(1));
-    SplitInteger(Five, Lo, Hi);
+    SplitInteger(Three, Lo, Hi);
+
+    Hi = DAG.getNode(ISD::UADDO, dl, VTHalfWithO, Hi, HighSum);
+    Overflow = DAG.getNode(ISD::OR, dl, BitVT, Overflow, Hi.getValue(1));
     ReplaceValueWith(SDValue(N, 1), Overflow);
     return;
   }
 
   Type *RetTy = VT.getTypeForEVT(*DAG.getContext());
   EVT PtrVT = TLI.getPointerTy(DAG.getDataLayout());
   Type *PtrTy = PtrVT.getTypeForEVT(*DAG.getContext());
 

llvm/test/CodeGen/AArch64/umulo-128-legalisation-lowering.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=aarch64-unknown-linux-gnu | FileCheck %s --check-prefixes=AARCH
 
 define { i128, i8 } @muloti_test(i128 %l, i128 %r) unnamed_addr #0 {
 ; AARCH-LABEL: muloti_test:
 ; AARCH:       // %bb.0: // %start
-; AARCH-NEXT:    mul x8, x3, x0
-; AARCH-NEXT:    umulh x9, x0, x2
-; AARCH-NEXT:    madd x11, x1, x2, x8
-; AARCH-NEXT:    add x8, x9, x11
-; AARCH-NEXT:    cmp x8, x9
-; AARCH-NEXT:    cset w9, lo
-; AARCH-NEXT:    cmp x11, #0 // =0
-; AARCH-NEXT:    csel w9, wzr, w9, eq
 ; AARCH-NEXT:    cmp x3, #0 // =0
-; AARCH-NEXT:    umulh x10, x1, x2
-; AARCH-NEXT:    cset w12, ne
+; AARCH-NEXT:    umulh x8, x1, x2
+; AARCH-NEXT:    cset w10, ne
 ; AARCH-NEXT:    cmp x1, #0 // =0
-; AARCH-NEXT:    umulh x11, x3, x0
-; AARCH-NEXT:    cset w13, ne
-; AARCH-NEXT:    cmp xzr, x10
-; AARCH-NEXT:    and w10, w13, w12
-; AARCH-NEXT:    cset w12, ne
-; AARCH-NEXT:    cmp xzr, x11
-; AARCH-NEXT:    orr w10, w10, w12
+; AARCH-NEXT:    mul x9, x3, x0
 ; AARCH-NEXT:    cset w11, ne
+; AARCH-NEXT:    cmp xzr, x8
+; AARCH-NEXT:    umulh x8, x3, x0
+; AARCH-NEXT:    madd x9, x1, x2, x9
+; AARCH-NEXT:    and w10, w11, w10
+; AARCH-NEXT:    cset w11, ne
+; AARCH-NEXT:    cmp xzr, x8
+; AARCH-NEXT:    umulh x8, x0, x2
 ; AARCH-NEXT:    orr w10, w10, w11
-; AARCH-NEXT:    orr w9, w10, w9
+; AARCH-NEXT:    cset w11, ne
+; AARCH-NEXT:    adds x1, x8, x9
+; AARCH-NEXT:    orr w8, w10, w11
+; AARCH-NEXT:    cset w9, hs
+; AARCH-NEXT:    orr w8, w8, w9
 ; AARCH-NEXT:    mul x0, x0, x2
-; AARCH-DAG:    mov x1, x8
-; AARCH-DAG:    mov w2, w9
+; AARCH-NEXT:    mov w2, w8
 ; AARCH-NEXT:    ret
 start:
   %0 = tail call { i128, i1 } @llvm.umul.with.overflow.i128(i128 %l, i128 %r) #2
   %1 = extractvalue { i128, i1 } %0, 0
   %2 = extractvalue { i128, i1 } %0, 1
   %3 = zext i1 %2 to i8
   %4 = insertvalue { i128, i8 } undef, i128 %1, 0
   %5 = insertvalue { i128, i8 } %4, i8 %3, 1

llvm/test/CodeGen/AArch64/vec_umulo.ll

; CHECK-NEXT:    ret
   %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> @umulo_v2i128(<2 x i128> %a0, <2 x i128> %a1, <2 x i128>* %p2) nounwind {
 ; CHECK-LABEL: umulo_v2i128:
 ; CHECK:       // %bb.0:
-; CHECK-NEXT:    mul x9, x7, x2
-; CHECK-NEXT:    umulh x10, x2, x6
-; CHECK-NEXT:    madd x9, x3, x6, x9
-; CHECK-NEXT:    add x15, x10, x9
-; CHECK-NEXT:    cmp x15, x10
-; CHECK-NEXT:    cset w10, lo
-; CHECK-NEXT:    cmp x9, #0 // =0
-; CHECK-NEXT:    csel w10, wzr, w10, eq
 ; CHECK-NEXT:    cmp x7, #0 // =0
-; CHECK-NEXT:    umulh x11, x3, x6
-; CHECK-NEXT:    mul x13, x5, x0
-; CHECK-NEXT:    cset w17, ne
+; CHECK-NEXT:    umulh x8, x3, x6
+; CHECK-NEXT:    cset w13, ne
 ; CHECK-NEXT:    cmp x3, #0 // =0
-; CHECK-NEXT:    umulh x12, x7, x2
-; CHECK-NEXT:    umulh x9, x0, x4
-; CHECK-NEXT:    madd x13, x1, x4, x13
-; CHECK-NEXT:    cset w18, ne
-; CHECK-NEXT:    cmp xzr, x11
+; CHECK-NEXT:    umulh x9, x7, x2
+; CHECK-NEXT:    mul x10, x7, x2
+; CHECK-NEXT:    cset w14, ne
+; CHECK-NEXT:    cmp xzr, x8
 ; CHECK-NEXT:    ldr x8, [sp]
-; CHECK-NEXT:    add x11, x9, x13
-; CHECK-NEXT:    and w17, w18, w17
-; CHECK-NEXT:    cset w18, ne
-; CHECK-NEXT:    cmp xzr, x12
-; CHECK-NEXT:    orr w12, w17, w18
-; CHECK-NEXT:    cset w17, ne
-; CHECK-NEXT:    cmp x11, x9
-; CHECK-NEXT:    orr w9, w12, w17
-; CHECK-NEXT:    cset w12, lo
-; CHECK-NEXT:    cmp x13, #0 // =0
-; CHECK-NEXT:    mul x14, x2, x6
-; CHECK-NEXT:    csel w12, wzr, w12, eq
-; CHECK-NEXT:    cmp x5, #0 // =0
-; CHECK-NEXT:    stp x14, x15, [x8, #16]
-; CHECK-NEXT:    umulh x14, x1, x4
-; CHECK-NEXT:    cset w13, ne
-; CHECK-NEXT:    cmp x1, #0 // =0
-; CHECK-NEXT:    umulh x16, x5, x0
-; CHECK-NEXT:    cset w17, ne
-; CHECK-NEXT:    cmp xzr, x14
-; CHECK-NEXT:    and w13, w17, w13
+; CHECK-NEXT:    umulh x11, x2, x6
+; CHECK-NEXT:    madd x10, x3, x6, x10
+; CHECK-NEXT:    and w13, w14, w13
 ; CHECK-NEXT:    cset w14, ne
-; CHECK-NEXT:    cmp xzr, x16
+; CHECK-NEXT:    cmp xzr, x9
 ; CHECK-NEXT:    orr w13, w13, w14
 ; CHECK-NEXT:    cset w14, ne
+; CHECK-NEXT:    adds x10, x11, x10
+; CHECK-NEXT:    mul x12, x2, x6
 ; CHECK-NEXT:    orr w13, w13, w14
-; CHECK-NEXT:    orr w12, w13, w12
+; CHECK-NEXT:    cset w14, hs
+; CHECK-NEXT:    cmp x5, #0 // =0
+; CHECK-NEXT:    umulh x17, x1, x4
+; CHECK-NEXT:    stp x12, x10, [x8, #16]
+; CHECK-NEXT:    cset w10, ne
+; CHECK-NEXT:    cmp x1, #0 // =0
+; CHECK-NEXT:    umulh x9, x5, x0
+; CHECK-NEXT:    mul x11, x5, x0
+; CHECK-NEXT:    cset w12, ne
+; CHECK-NEXT:    cmp xzr, x17
+; CHECK-NEXT:    umulh x15, x0, x4
+; CHECK-NEXT:    madd x11, x1, x4, x11
+; CHECK-NEXT:    and w10, w12, w10
+; CHECK-NEXT:    cset w12, ne
+; CHECK-NEXT:    cmp xzr, x9
+; CHECK-NEXT:    orr w9, w10, w12
+; CHECK-NEXT:    cset w10, ne
+; CHECK-NEXT:    adds x11, x15, x11
+; CHECK-NEXT:    orr w9, w9, w10
+; CHECK-NEXT:    cset w10, hs
 ; CHECK-NEXT:    orr w9, w9, w10
-; CHECK-NEXT:    fmov s0, w12
-; CHECK-NEXT:    mov v0.s[1], w9
+; CHECK-NEXT:    orr w10, w13, w14
+; CHECK-NEXT:    fmov s0, w9
+; CHECK-NEXT:    mov v0.s[1], w10
 ; CHECK-NEXT:    shl v0.2s, v0.2s, #31
-; CHECK-NEXT:    mul x15, x0, x4
+; CHECK-NEXT:    mul x16, x0, x4
 ; CHECK-NEXT:    sshr v0.2s, v0.2s, #31
-; CHECK-NEXT:    stp x15, x11, [x8]
+; CHECK-NEXT:    stp x16, x11, [x8]
 ; CHECK-NEXT:    ret
   %t = call {<2 x i128>, <2 x i1>} @llvm.umul.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/PowerPC/pr45448.ll

 ; CHECK-NEXT:  # %bb.5: # %L294
 ; CHECK-NEXT:    bc 4, 4*cr5+lt, .LBB0_7
 ; CHECK-NEXT:  .LBB0_6: # %L1057.preheader
 ; CHECK-NEXT:  .LBB0_7: # %L670
 ; CHECK-NEXT:    li r5, -3
 ; CHECK-NEXT:    cmpdi r3, 0
 ; CHECK-NEXT:    sradi r4, r3, 63
 ; CHECK-NEXT:    rldic r5, r5, 4, 32
-; CHECK-NEXT:    crnot 4*cr5+gt, eq
+; CHECK-NEXT:    crnot 4*cr5+lt, eq
 ; CHECK-NEXT:    mulhdu r3, r3, r5
 ; CHECK-NEXT:    maddld r6, r4, r5, r3
-; CHECK-NEXT:    cmpld r6, r3
-; CHECK-NEXT:    mulld r3, r4, r5
-; CHECK-NEXT:    cmpldi cr1, r3, 0
-; CHECK-NEXT:    crandc 4*cr5+lt, lt, 4*cr1+eq
+; CHECK-NEXT:    cmpld cr1, r6, r3
 ; CHECK-NEXT:    mulhdu. r3, r4, r5
-; CHECK-NEXT:    bc 4, 4*cr5+gt, .LBB0_10
+; CHECK-NEXT:    bc 4, 4*cr5+lt, .LBB0_10
 ; CHECK-NEXT:  # %bb.8: # %L670
-; CHECK-NEXT:    crorc 4*cr5+lt, 4*cr5+lt, eq
+; CHECK-NEXT:    crorc 4*cr5+lt, 4*cr1+lt, eq
 ; CHECK-NEXT:    bc 4, 4*cr5+lt, .LBB0_10
 ; CHECK-NEXT:  # %bb.9: # %L917
 ; CHECK-NEXT:  .LBB0_10: # %L994
 top:
   %0 = load i64, i64* undef, align 8
   %1 = icmp ne i64 %0, 0
   %2 = sext i64 %0 to i128
   switch i64 %0, label %pass195 [

llvm/test/CodeGen/PowerPC/umulo-128-legalisation-lowering.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=powerpc64-unknown-linux-gnu | FileCheck %s --check-prefixes=PPC64
 ; RUN: llc < %s -mtriple=powerpc-unknown-linux-gnu | FileCheck %s --check-prefixes=PPC32
 
 define { i128, i8 } @muloti_test(i128 %l, i128 %r) unnamed_addr #0 {
 ; PPC64-LABEL: muloti_test:
 ; PPC64:       # %bb.0: # %start
-; PPC64-NEXT:    mulhdu. 8, 3, 6
-; PPC64-NEXT:    mcrf 1, 0
-; PPC64-NEXT:    mulld 8, 5, 4
-; PPC64-NEXT:    cmpdi 3, 0
-; PPC64-NEXT:    mulld 3, 3, 6
-; PPC64-NEXT:    cmpdi 5, 5, 0
-; PPC64-NEXT:    add 3, 3, 8
-; PPC64-NEXT:    crnor 20, 22, 2
-; PPC64-NEXT:    mulhdu 8, 4, 6
-; PPC64-NEXT:    cmpldi 3, 0
-; PPC64-NEXT:    add 3, 8, 3
-; PPC64-NEXT:    cmpld 6, 3, 8
-; PPC64-NEXT:    crandc 21, 24, 2
-; PPC64-NEXT:    crorc 20, 20, 6
-; PPC64-NEXT:    mulhdu. 5, 5, 4
-; PPC64-NEXT:    crorc 20, 20, 2
-; PPC64-NEXT:    li 7, 1
-; PPC64-NEXT:    crnor 20, 20, 21
+; PPC64-NEXT:    addic 8, 5, -1
+; PPC64-NEXT:    mulhdu 9, 5, 4
+; PPC64-NEXT:    mulld 10, 5, 4
+; PPC64-NEXT:    subfe 5, 8, 5
+; PPC64-NEXT:    mulld 8, 3, 6
+; PPC64-NEXT:    add 8, 8, 10
+; PPC64-NEXT:    addic 10, 3, -1
+; PPC64-NEXT:    mulhdu 7, 3, 6
+; PPC64-NEXT:    subfe 3, 10, 3
+; PPC64-NEXT:    and 5, 3, 5
+; PPC64-NEXT:    addic 3, 7, -1
+; PPC64-NEXT:    subfe 7, 3, 7
+; PPC64-NEXT:    or 5, 5, 7
+; PPC64-NEXT:    mulhdu 10, 4, 6
+; PPC64-NEXT:    addic 7, 9, -1
+; PPC64-NEXT:    add 3, 10, 8
+; PPC64-NEXT:    subfe 7, 7, 9
+; PPC64-NEXT:    or 5, 5, 7
+; PPC64-NEXT:    subc 7, 3, 10
+; PPC64-NEXT:    subfe 7, 3, 3
+; PPC64-NEXT:    neg 7, 7
+; PPC64-NEXT:    or 5, 5, 7
 ; PPC64-NEXT:    mulld 4, 4, 6
-; PPC64-NEXT:    bc 12, 20, .LBB0_2
-; PPC64-NEXT:  # %bb.1: # %start
-; PPC64-NEXT:    ori 5, 7, 0
-; PPC64-NEXT:    blr
-; PPC64-NEXT:  .LBB0_2: # %start
-; PPC64-NEXT:    li 5, 0
 ; PPC64-NEXT:    blr
 ;
 ; PPC32-LABEL: muloti_test:
 ; PPC32:       # %bb.0: # %start
 ; PPC32-NEXT:    mflr 0
 ; PPC32-NEXT:    stw 0, 4(1)
-; PPC32-NEXT:    stwu 1, -80(1)
-; PPC32-NEXT:    stw 26, 56(1) # 4-byte Folded Spill
+; PPC32-NEXT:    stwu 1, -64(1)
+; PPC32-NEXT:    stw 24, 32(1) # 4-byte Folded Spill
 ; PPC32-NEXT:    mfcr 12
-; PPC32-NEXT:    stw 27, 60(1) # 4-byte Folded Spill
-; PPC32-NEXT:    mr 27, 4
-; PPC32-NEXT:    stw 29, 68(1) # 4-byte Folded Spill
-; PPC32-NEXT:    mr 29, 7
-; PPC32-NEXT:    stw 30, 72(1) # 4-byte Folded Spill
-; PPC32-NEXT:    mr 30, 8
-; PPC32-NEXT:    mr 26, 3
+; PPC32-NEXT:    stw 26, 40(1) # 4-byte Folded Spill
+; PPC32-NEXT:    mr 26, 7
+; PPC32-NEXT:    stw 28, 48(1) # 4-byte Folded Spill
+; PPC32-NEXT:    mr 28, 4
+; PPC32-NEXT:    stw 29, 52(1) # 4-byte Folded Spill
+; PPC32-NEXT:    mr 29, 8
+; PPC32-NEXT:    mr 24, 3
 ; PPC32-NEXT:    li 3, 0
 ; PPC32-NEXT:    li 4, 0
 ; PPC32-NEXT:    li 7, 0
 ; PPC32-NEXT:    li 8, 0
-; PPC32-NEXT:    stw 20, 32(1) # 4-byte Folded Spill
-; PPC32-NEXT:    stw 21, 36(1) # 4-byte Folded Spill
-; PPC32-NEXT:    stw 22, 40(1) # 4-byte Folded Spill
-; PPC32-NEXT:    stw 23, 44(1) # 4-byte Folded Spill
-; PPC32-NEXT:    mr 23, 6
-; PPC32-NEXT:    stw 24, 48(1) # 4-byte Folded Spill
-; PPC32-NEXT:    mr 24, 5
-; PPC32-NEXT:    stw 25, 52(1) # 4-byte Folded Spill
-; PPC32-NEXT:    mr 25, 10
-; PPC32-NEXT:    stw 28, 64(1) # 4-byte Folded Spill
-; PPC32-NEXT:    mr 28, 9
-; PPC32-NEXT:    stw 12, 28(1)
+; PPC32-NEXT:    stw 23, 28(1) # 4-byte Folded Spill
+; PPC32-NEXT:    mr 23, 5
+; PPC32-NEXT:    stw 25, 36(1) # 4-byte Folded Spill
+; PPC32-NEXT:    mr 25, 9
+; PPC32-NEXT:    stw 27, 44(1) # 4-byte Folded Spill
+; PPC32-NEXT:    mr 27, 6
+; PPC32-NEXT:    stw 30, 56(1) # 4-byte Folded Spill
+; PPC32-NEXT:    mr 30, 10
+; PPC32-NEXT:    stw 12, 24(1)
 ; PPC32-NEXT:    bl __multi3
-; PPC32-NEXT:    mr 7, 4
-; PPC32-NEXT:    mullw 4, 24, 30
-; PPC32-NEXT:    cmpwi 5, 24, 0
-; PPC32-NEXT:    cmpwi 6, 26, 0
-; PPC32-NEXT:    cmpwi 7, 28, 0
-; PPC32-NEXT:    crnor 9, 30, 26
-; PPC32-NEXT:    mullw 8, 29, 23
-; PPC32-NEXT:    add 21, 8, 4
-; PPC32-NEXT:    mullw 11, 28, 27
-; PPC32-NEXT:    mullw 12, 26, 25
-; PPC32-NEXT:    add 11, 12, 11
-; PPC32-NEXT:    cmplwi 7, 11, 0
-; PPC32-NEXT:    mulhwu 9, 30, 23
-; PPC32-NEXT:    add 12, 9, 21
-; PPC32-NEXT:    cmplw 6, 12, 9
-; PPC32-NEXT:    mulhwu 10, 27, 25
-; PPC32-NEXT:    mullw 0, 30, 23
-; PPC32-NEXT:    mullw 22, 27, 25
-; PPC32-NEXT:    addc 4, 22, 0
-; PPC32-NEXT:    add 0, 10, 11
-; PPC32-NEXT:    adde 8, 0, 12
-; PPC32-NEXT:    addc 4, 7, 4
-; PPC32-NEXT:    adde 8, 3, 8
-; PPC32-NEXT:    xor 22, 4, 7
-; PPC32-NEXT:    xor 20, 8, 3
-; PPC32-NEXT:    or. 22, 22, 20
-; PPC32-NEXT:    mcrf 1, 0
-; PPC32-NEXT:    cmpwi 29, 0
-; PPC32-NEXT:    crnor 8, 22, 2
-; PPC32-NEXT:    mulhwu. 23, 29, 23
+; PPC32-NEXT:    mulhwu. 9, 26, 27
+; PPC32-NEXT:    mfcr 9 # cr0
+; PPC32-NEXT:    cmpwi 2, 26, 0
+; PPC32-NEXT:    stw 9, 20(1)
+; PPC32-NEXT:    cmpwi 3, 23, 0
+; PPC32-NEXT:    crnor 12, 14, 10
+; PPC32-NEXT:    cmpwi 4, 24, 0
+; PPC32-NEXT:    mulhwu. 9, 23, 29
 ; PPC32-NEXT:    mcrf 5, 0
-; PPC32-NEXT:    cmplwi 21, 0
-; PPC32-NEXT:    crandc 10, 24, 2
-; PPC32-NEXT:    cmplw 3, 0, 10
-; PPC32-NEXT:    crandc 11, 12, 30
+; PPC32-NEXT:    cmpwi 1, 25, 0
+; PPC32-NEXT:    crnor 4, 6, 18
 ; PPC32-NEXT:    mulhwu. 9, 24, 30
 ; PPC32-NEXT:    mcrf 6, 0
-; PPC32-NEXT:    cmplw 4, 7
-; PPC32-NEXT:    cmplw 7, 8, 3
-; PPC32-NEXT:    crand 12, 30, 0
-; PPC32-NEXT:    crandc 13, 28, 30
-; PPC32-NEXT:    mulhwu. 3, 26, 25
+; PPC32-NEXT:    mulhwu. 0, 25, 28
 ; PPC32-NEXT:    mcrf 7, 0
-; PPC32-NEXT:    cror 0, 12, 13
-; PPC32-NEXT:    crandc 12, 0, 6
-; PPC32-NEXT:    crorc 20, 8, 22
-; PPC32-NEXT:    crorc 20, 20, 26
-; PPC32-NEXT:    mulhwu. 3, 28, 27
-; PPC32-NEXT:    mcrf 1, 0
-; PPC32-NEXT:    crorc 25, 9, 30
-; PPC32-NEXT:    or. 3, 27, 26
-; PPC32-NEXT:    cror 24, 20, 10
-; PPC32-NEXT:    mcrf 5, 0
-; PPC32-NEXT:    crorc 25, 25, 6
-; PPC32-NEXT:    or. 3, 30, 29
-; PPC32-NEXT:    cror 25, 25, 11
-; PPC32-NEXT:    crnor 20, 2, 22
-; PPC32-NEXT:    lwz 12, 28(1)
-; PPC32-NEXT:    cror 20, 20, 25
-; PPC32-NEXT:    cror 20, 20, 24
-; PPC32-NEXT:    crnor 20, 20, 12
+; PPC32-NEXT:    or. 0, 28, 24
+; PPC32-NEXT:    mcrf 2, 0
+; PPC32-NEXT:    or. 0, 29, 26
+; PPC32-NEXT:    crnor 5, 2, 10
+; PPC32-NEXT:    mullw 10, 26, 27
+; PPC32-NEXT:    lwz 26, 20(1)
+; PPC32-NEXT:    mullw 9, 23, 29
+; PPC32-NEXT:    add 9, 10, 9
+; PPC32-NEXT:    mtcrf 128, 26 # cr0
+; PPC32-NEXT:    crorc 6, 12, 2
+; PPC32-NEXT:    crorc 20, 6, 22
+; PPC32-NEXT:    mulhwu 7, 29, 27
+; PPC32-NEXT:    add 9, 7, 9
+; PPC32-NEXT:    cmplw 9, 7
+; PPC32-NEXT:    crorc 21, 4, 26
+; PPC32-NEXT:    cror 20, 20, 0
+; PPC32-NEXT:    crorc 21, 21, 30
+; PPC32-NEXT:    mullw 11, 25, 28
+; PPC32-NEXT:    mullw 12, 24, 30
+; PPC32-NEXT:    add 10, 12, 11
+; PPC32-NEXT:    lwz 12, 24(1)
+; PPC32-NEXT:    lwz 26, 40(1) # 4-byte Folded Reload
+; PPC32-NEXT:    mulhwu 8, 28, 30
+; PPC32-NEXT:    add 10, 8, 10
+; PPC32-NEXT:    cmplw 10, 8
+; PPC32-NEXT:    cror 21, 21, 0
+; PPC32-NEXT:    cror 21, 5, 21
+; PPC32-NEXT:    cror 20, 21, 20
+; PPC32-NEXT:    mullw 0, 29, 27
+; PPC32-NEXT:    mtcrf 32, 12 # cr2
+; PPC32-NEXT:    mtcrf 16, 12 # cr3
+; PPC32-NEXT:    mtcrf 8, 12 # cr4
+; PPC32-NEXT:    lwz 29, 52(1) # 4-byte Folded Reload
+; PPC32-NEXT:    mullw 7, 28, 30
+; PPC32-NEXT:    addc 7, 7, 0
+; PPC32-NEXT:    adde 11, 10, 9
+; PPC32-NEXT:    addc 9, 4, 7
+; PPC32-NEXT:    adde 8, 3, 11
+; PPC32-NEXT:    cmplw 6, 9, 4
+; PPC32-NEXT:    cmplw 8, 3
+; PPC32-NEXT:    crand 22, 2, 24
+; PPC32-NEXT:    crandc 23, 0, 2
+; PPC32-NEXT:    cror 22, 22, 23
+; PPC32-NEXT:    crnor 20, 20, 22
 ; PPC32-NEXT:    li 3, 1
 ; PPC32-NEXT:    bc 12, 20, .LBB0_2
 ; PPC32-NEXT:  # %bb.1: # %start
 ; PPC32-NEXT:    ori 7, 3, 0
 ; PPC32-NEXT:    b .LBB0_3
 ; PPC32-NEXT:  .LBB0_2: # %start
 ; PPC32-NEXT:    li 7, 0
 ; PPC32-NEXT:  .LBB0_3: # %start
 ; PPC32-NEXT:    mr 3, 8
-; PPC32-NEXT:    mtcrf 32, 12 # cr2
-; PPC32-NEXT:    mtcrf 16, 12 # cr3
-; PPC32-NEXT:    lwz 30, 72(1) # 4-byte Folded Reload
-; PPC32-NEXT:    lwz 29, 68(1) # 4-byte Folded Reload
-; PPC32-NEXT:    lwz 28, 64(1) # 4-byte Folded Reload
-; PPC32-NEXT:    lwz 27, 60(1) # 4-byte Folded Reload
-; PPC32-NEXT:    lwz 26, 56(1) # 4-byte Folded Reload
-; PPC32-NEXT:    lwz 25, 52(1) # 4-byte Folded Reload
-; PPC32-NEXT:    lwz 24, 48(1) # 4-byte Folded Reload
-; PPC32-NEXT:    lwz 23, 44(1) # 4-byte Folded Reload
-; PPC32-NEXT:    lwz 22, 40(1) # 4-byte Folded Reload
-; PPC32-NEXT:    lwz 21, 36(1) # 4-byte Folded Reload
-; PPC32-NEXT:    lwz 20, 32(1) # 4-byte Folded Reload
-; PPC32-NEXT:    lwz 0, 84(1)
-; PPC32-NEXT:    addi 1, 1, 80
+; PPC32-NEXT:    mr 4, 9
+; PPC32-NEXT:    lwz 30, 56(1) # 4-byte Folded Reload
+; PPC32-NEXT:    lwz 28, 48(1) # 4-byte Folded Reload
+; PPC32-NEXT:    lwz 27, 44(1) # 4-byte Folded Reload
+; PPC32-NEXT:    lwz 25, 36(1) # 4-byte Folded Reload
+; PPC32-NEXT:    lwz 24, 32(1) # 4-byte Folded Reload
+; PPC32-NEXT:    lwz 23, 28(1) # 4-byte Folded Reload
+; PPC32-NEXT:    lwz 0, 68(1)
+; PPC32-NEXT:    addi 1, 1, 64
 ; PPC32-NEXT:    mtlr 0
 ; PPC32-NEXT:    blr
 start:
   %0 = tail call { i128, i1 } @llvm.umul.with.overflow.i128(i128 %l, i128 %r) #2
   %1 = extractvalue { i128, i1 } %0, 0
   %2 = extractvalue { i128, i1 } %0, 1
   %3 = zext i1 %2 to i8
   %4 = insertvalue { i128, i8 } undef, i128 %1, 0

llvm/test/CodeGen/SPARC/umulo-128-legalisation-lowering.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=sparc-unknown-linux-gnu | FileCheck %s --check-prefixes=SPARC
 ; RUN: llc < %s -mtriple=sparc64-unknown-linux-gnu | FileCheck %s --check-prefixes=SPARC64
 
 define { i128, i8 } @muloti_test(i128 %l, i128 %r) unnamed_addr #0 {
 ; SPARC-LABEL: muloti_test:
 ; SPARC:         .cfi_startproc
 ; SPARC-NEXT:  ! %bb.0: ! %start
-; SPARC-NEXT:    save %sp, -128, %sp
+; SPARC-NEXT:    save %sp, -120, %sp
 ; SPARC-NEXT:    .cfi_def_cfa_register %fp
 ; SPARC-NEXT:    .cfi_window_save
 ; SPARC-NEXT:    .cfi_register %o7, %i7
-; SPARC-NEXT:    ld [%fp+92], %l3
+; SPARC-NEXT:    ld [%fp+92], %l4
 ; SPARC-NEXT:    ld [%fp+96], %g2
 ; SPARC-NEXT:    umul %i2, %i5, %g3
 ; SPARC-NEXT:    rd %y, %g4
-; SPARC-NEXT:    st %g4, [%fp+-20] ! 4-byte Folded Spill
+; SPARC-NEXT:    st %g4, [%fp+-12] ! 4-byte Folded Spill
 ; SPARC-NEXT:    umul %i4, %i3, %g4
 ; SPARC-NEXT:    rd %y, %l0
-; SPARC-NEXT:    st %l0, [%fp+-24] ! 4-byte Folded Spill
+; SPARC-NEXT:    st %l0, [%fp+-16] ! 4-byte Folded Spill
 ; SPARC-NEXT:    st %g2, [%sp+96]
-; SPARC-NEXT:    st %i5, [%fp+-8] ! 4-byte Folded Spill
 ; SPARC-NEXT:    umul %i5, %i3, %l0
-; SPARC-NEXT:    rd %y, %l5
-; SPARC-NEXT:    st %l3, [%sp+92]
-; SPARC-NEXT:    umul %l3, %i1, %l4
-; SPARC-NEXT:    rd %y, %i5
-; SPARC-NEXT:    st %i5, [%fp+-12] ! 4-byte Folded Spill
-; SPARC-NEXT:    add %g4, %g3, %l2
-; SPARC-NEXT:    mov %i0, %i5
-; SPARC-NEXT:    umul %i0, %g2, %g3
-; SPARC-NEXT:    rd %y, %i0
-; SPARC-NEXT:    st %i0, [%fp+-16] ! 4-byte Folded Spill
-; SPARC-NEXT:    add %l5, %l2, %l1
-; SPARC-NEXT:    st %i1, [%fp+-4] ! 4-byte Folded Spill
-; SPARC-NEXT:    umul %i1, %g2, %g2
 ; SPARC-NEXT:    rd %y, %l6
-; SPARC-NEXT:    add %g3, %l4, %i0
-; SPARC-NEXT:    add %l6, %i0, %l7
-; SPARC-NEXT:    addcc %g2, %l0, %l4
+; SPARC-NEXT:    st %l4, [%sp+92]
+; SPARC-NEXT:    umul %l4, %i1, %l2
+; SPARC-NEXT:    rd %y, %l1
+; SPARC-NEXT:    st %l1, [%fp+-4] ! 4-byte Folded Spill
+; SPARC-NEXT:    add %g4, %g3, %g3
+; SPARC-NEXT:    umul %i0, %g2, %g4
+; SPARC-NEXT:    rd %y, %l1
+; SPARC-NEXT:    st %l1, [%fp+-8] ! 4-byte Folded Spill
+; SPARC-NEXT:    add %l6, %g3, %l3
+; SPARC-NEXT:    umul %i1, %g2, %g2
+; SPARC-NEXT:    rd %y, %l1
+; SPARC-NEXT:    add %g4, %l2, %g3
+; SPARC-NEXT:    add %l1, %g3, %l2
+; SPARC-NEXT:    addcc %g2, %l0, %l7
 ; SPARC-NEXT:    mov %g0, %l0
-; SPARC-NEXT:    addxcc %l7, %l1, %i1
+; SPARC-NEXT:    addxcc %l2, %l3, %l5
 ; SPARC-NEXT:    mov %l0, %o0
 ; SPARC-NEXT:    mov %l0, %o1
 ; SPARC-NEXT:    mov %i2, %o2
 ; SPARC-NEXT:    mov %i3, %o3
 ; SPARC-NEXT:    mov %l0, %o4
 ; SPARC-NEXT:    call __multi3
 ; SPARC-NEXT:    mov %l0, %o5
-; SPARC-NEXT:    addcc %o1, %l4, %i3
-; SPARC-NEXT:    addxcc %o0, %i1, %g2
+; SPARC-NEXT:    addcc %o1, %l7, %i3
+; SPARC-NEXT:    addxcc %o0, %l5, %g2
 ; SPARC-NEXT:    mov 1, %g3
 ; SPARC-NEXT:    cmp %g2, %o0
 ; SPARC-NEXT:    bcs .LBB0_2
-; SPARC-NEXT:    mov %g3, %g4
+; SPARC-NEXT:    mov %g3, %o4
 ; SPARC-NEXT:  ! %bb.1: ! %start
-; SPARC-NEXT:    mov %l0, %g4
+; SPARC-NEXT:    mov %l0, %o4
 ; SPARC-NEXT:  .LBB0_2: ! %start
 ; SPARC-NEXT:    cmp %i3, %o1
 ; SPARC-NEXT:    bcs .LBB0_4
-; SPARC-NEXT:    mov %g3, %o4
+; SPARC-NEXT:    mov %g3, %g4
 ; SPARC-NEXT:  ! %bb.3: ! %start
-; SPARC-NEXT:    mov %l0, %o4
+; SPARC-NEXT:    mov %l0, %g4
 ; SPARC-NEXT:  .LBB0_4: ! %start
 ; SPARC-NEXT:    cmp %g2, %o0
 ; SPARC-NEXT:    be .LBB0_6
 ; SPARC-NEXT:    nop
 ; SPARC-NEXT:  ! %bb.5: ! %start
-; SPARC-NEXT:    mov %g4, %o4
+; SPARC-NEXT:    mov %o4, %g4
 ; SPARC-NEXT:  .LBB0_6: ! %start
-; SPARC-NEXT:    xor %g2, %o0, %i1
-; SPARC-NEXT:    xor %i3, %o1, %g4
-; SPARC-NEXT:    or %g4, %i1, %i1
-; SPARC-NEXT:    cmp %i1, 0
-; SPARC-NEXT:    be .LBB0_8
-; SPARC-NEXT:    mov %l0, %g4
+; SPARC-NEXT:    cmp %i2, 0
+; SPARC-NEXT:    bne .LBB0_8
+; SPARC-NEXT:    mov %g3, %i2
 ; SPARC-NEXT:  ! %bb.7: ! %start
-; SPARC-NEXT:    mov %o4, %g4
+; SPARC-NEXT:    mov %l0, %i2
 ; SPARC-NEXT:  .LBB0_8: ! %start
-; SPARC-NEXT:    cmp %l1, %l5
-; SPARC-NEXT:    mov %g3, %l1
-; SPARC-NEXT:    bcs .LBB0_10
-; SPARC-NEXT:    mov %i5, %i1
+; SPARC-NEXT:    cmp %i4, 0
+; SPARC-NEXT:    bne .LBB0_10
+; SPARC-NEXT:    mov %g3, %o1
 ; SPARC-NEXT:  ! %bb.9: ! %start
-; SPARC-NEXT:    mov %l0, %l1
+; SPARC-NEXT:    mov %l0, %o1
 ; SPARC-NEXT:  .LBB0_10: ! %start
-; SPARC-NEXT:    cmp %l2, 0
-; SPARC-NEXT:    be .LBB0_12
-; SPARC-NEXT:    mov %l0, %o0
+; SPARC-NEXT:    ld [%fp+-16], %l5 ! 4-byte Folded Reload
+; SPARC-NEXT:    cmp %l5, 0
+; SPARC-NEXT:    bne .LBB0_12
+; SPARC-NEXT:    mov %g3, %o0
 ; SPARC-NEXT:  ! %bb.11: ! %start
-; SPARC-NEXT:    mov %l1, %o0
+; SPARC-NEXT:    mov %l0, %o0
 ; SPARC-NEXT:  .LBB0_12: ! %start
-; SPARC-NEXT:    cmp %i2, 0
+; SPARC-NEXT:    ld [%fp+-12], %l5 ! 4-byte Folded Reload
+; SPARC-NEXT:    cmp %l5, 0
 ; SPARC-NEXT:    bne .LBB0_14
-; SPARC-NEXT:    mov %g3, %i2
+; SPARC-NEXT:    mov %g3, %l5
 ; SPARC-NEXT:  ! %bb.13: ! %start
-; SPARC-NEXT:    mov %l0, %i2
+; SPARC-NEXT:    mov %l0, %l5
 ; SPARC-NEXT:  .LBB0_14: ! %start
-; SPARC-NEXT:    cmp %i4, 0
-; SPARC-NEXT:    bne .LBB0_16
-; SPARC-NEXT:    mov %g3, %o1
+; SPARC-NEXT:    cmp %l3, %l6
+; SPARC-NEXT:    bcs .LBB0_16
+; SPARC-NEXT:    mov %g3, %l3
 ; SPARC-NEXT:  ! %bb.15: ! %start
-; SPARC-NEXT:    mov %l0, %o1
+; SPARC-NEXT:    mov %l0, %l3
 ; SPARC-NEXT:  .LBB0_16: ! %start
-; SPARC-NEXT:    ld [%fp+-24], %i5 ! 4-byte Folded Reload
-; SPARC-NEXT:    cmp %i5, 0
+; SPARC-NEXT:    cmp %l4, 0
 ; SPARC-NEXT:    bne .LBB0_18
-; SPARC-NEXT:    mov %g3, %l5
+; SPARC-NEXT:    mov %g3, %l4
 ; SPARC-NEXT:  ! %bb.17: ! %start
-; SPARC-NEXT:    mov %l0, %l5
+; SPARC-NEXT:    mov %l0, %l4
 ; SPARC-NEXT:  .LBB0_18: ! %start
-; SPARC-NEXT:    ld [%fp+-20], %i5 ! 4-byte Folded Reload
-; SPARC-NEXT:    cmp %i5, 0
+; SPARC-NEXT:    cmp %i0, 0
 ; SPARC-NEXT:    bne .LBB0_20
-; SPARC-NEXT:    mov %g3, %l1
+; SPARC-NEXT:    mov %g3, %l7
 ; SPARC-NEXT:  ! %bb.19: ! %start
-; SPARC-NEXT:    mov %l0, %l1
+; SPARC-NEXT:    mov %l0, %l7
 ; SPARC-NEXT:  .LBB0_20: ! %start
-; SPARC-NEXT:    cmp %l7, %l6
-; SPARC-NEXT:    bcs .LBB0_22
+; SPARC-NEXT:    ld [%fp+-8], %l6 ! 4-byte Folded Reload
+; SPARC-NEXT:    cmp %l6, 0
+; SPARC-NEXT:    bne .LBB0_22
 ; SPARC-NEXT:    mov %g3, %l6
 ; SPARC-NEXT:  ! %bb.21: ! %start
 ; SPARC-NEXT:    mov %l0, %l6
 ; SPARC-NEXT:  .LBB0_22: ! %start
-; SPARC-NEXT:    cmp %i0, 0
-; SPARC-NEXT:    be .LBB0_24
-; SPARC-NEXT:    mov %l0, %l2
+; SPARC-NEXT:    and %o1, %i2, %i2
+; SPARC-NEXT:    ld [%fp+-4], %o1 ! 4-byte Folded Reload
+; SPARC-NEXT:    cmp %o1, 0
+; SPARC-NEXT:    and %l7, %l4, %o1
+; SPARC-NEXT:    bne .LBB0_24
+; SPARC-NEXT:    mov %g3, %l4
 ; SPARC-NEXT:  ! %bb.23: ! %start
-; SPARC-NEXT:    mov %l6, %l2
+; SPARC-NEXT:    mov %l0, %l4
 ; SPARC-NEXT:  .LBB0_24: ! %start
-; SPARC-NEXT:    cmp %l3, 0
-; SPARC-NEXT:    bne .LBB0_26
-; SPARC-NEXT:    mov %g3, %l3
+; SPARC-NEXT:    or %i2, %o0, %l7
+; SPARC-NEXT:    cmp %l2, %l1
+; SPARC-NEXT:    or %o1, %l6, %l2
+; SPARC-NEXT:    bcs .LBB0_26
+; SPARC-NEXT:    mov %g3, %i2
 ; SPARC-NEXT:  ! %bb.25: ! %start
-; SPARC-NEXT:    mov %l0, %l3
+; SPARC-NEXT:    mov %l0, %i2
 ; SPARC-NEXT:  .LBB0_26: ! %start
-; SPARC-NEXT:    cmp %i1, 0
+; SPARC-NEXT:    or %l7, %l5, %l1
+; SPARC-NEXT:    or %i5, %i4, %i4
+; SPARC-NEXT:    cmp %i4, 0
+; SPARC-NEXT:    or %l2, %l4, %l2
 ; SPARC-NEXT:    bne .LBB0_28
-; SPARC-NEXT:    mov %g3, %l4
+; SPARC-NEXT:    mov %g3, %i4
 ; SPARC-NEXT:  ! %bb.27: ! %start
-; SPARC-NEXT:    mov %l0, %l4
+; SPARC-NEXT:    mov %l0, %i4
 ; SPARC-NEXT:  .LBB0_28: ! %start
-; SPARC-NEXT:    and %o1, %i2, %i2
-; SPARC-NEXT:    ld [%fp+-16], %i0 ! 4-byte Folded Reload
+; SPARC-NEXT:    or %l1, %l3, %i5
+; SPARC-NEXT:    or %i1, %i0, %i0
 ; SPARC-NEXT:    cmp %i0, 0
-; SPARC-NEXT:    and %l4, %l3, %l4
 ; SPARC-NEXT:    bne .LBB0_30
-; SPARC-NEXT:    mov %g3, %l6
+; SPARC-NEXT:    or %l2, %i2, %i0
 ; SPARC-NEXT:  ! %bb.29: ! %start
-; SPARC-NEXT:    mov %l0, %l6
-; SPARC-NEXT:  .LBB0_30: ! %start
-; SPARC-NEXT:    or %i2, %l5, %l3
-; SPARC-NEXT:    ld [%fp+-12], %i0 ! 4-byte Folded Reload
-; SPARC-NEXT:    cmp %i0, 0
-; SPARC-NEXT:    or %l4, %l6, %i2
-; SPARC-NEXT:    bne .LBB0_32
-; SPARC-NEXT:    mov %g3, %l4
-; SPARC-NEXT:  ! %bb.31: ! %start
-; SPARC-NEXT:    mov %l0, %l4
-; SPARC-NEXT:  .LBB0_32: ! %start
-; SPARC-NEXT:    or %l3, %l1, %l1
-; SPARC-NEXT:    ld [%fp+-8], %i0 ! 4-byte Folded Reload
-; SPARC-NEXT:    or %i0, %i4, %i0
-; SPARC-NEXT:    cmp %i0, 0
-; SPARC-NEXT:    or %i2, %l4, %i5
-; SPARC-NEXT:    bne .LBB0_34
-; SPARC-NEXT:    mov %g3, %i2
-; SPARC-NEXT:  ! %bb.33: ! %start
-; SPARC-NEXT:    mov %l0, %i2
-; SPARC-NEXT:  .LBB0_34: ! %start
-; SPARC-NEXT:    or %l1, %o0, %i4
-; SPARC-NEXT:    ld [%fp+-4], %i0 ! 4-byte Folded Reload
-; SPARC-NEXT:    or %i0, %i1, %i0
-; SPARC-NEXT:    cmp %i0, 0
-; SPARC-NEXT:    bne .LBB0_36
-; SPARC-NEXT:    or %i5, %l2, %i0
-; SPARC-NEXT:  ! %bb.35: ! %start
 ; SPARC-NEXT:    mov %l0, %g3
-; SPARC-NEXT:  .LBB0_36: ! %start
-; SPARC-NEXT:    and %g3, %i2, %i1
+; SPARC-NEXT:  .LBB0_30: ! %start
+; SPARC-NEXT:    and %g3, %i4, %i1
 ; SPARC-NEXT:    or %i1, %i0, %i0
-; SPARC-NEXT:    or %i0, %i4, %i0
+; SPARC-NEXT:    or %i0, %i5, %i0
 ; SPARC-NEXT:    or %i0, %g4, %i0
 ; SPARC-NEXT:    and %i0, 1, %i4
 ; SPARC-NEXT:    mov %g2, %i0
 ; SPARC-NEXT:    mov %i3, %i1
 ; SPARC-NEXT:    mov %o2, %i2
 ; SPARC-NEXT:    ret
 ; SPARC-NEXT:    restore %g0, %o3, %o3
 ;

 ; SPARC64-NEXT:    mov 0, %o0
 ; SPARC64-NEXT:    mov %i1, %o1
 ; SPARC64-NEXT:    mov %o0, %o2
 ; SPARC64-NEXT:    call __multi3
 ; SPARC64-NEXT:    mov %i3, %o3
 ; SPARC64-NEXT:    add %o0, %i5, %i1
 ; SPARC64-NEXT:    mov %g0, %i3
 ; SPARC64-NEXT:    cmp %i1, %o0
-; SPARC64-NEXT:    mov %i3, %g2
-; SPARC64-NEXT:    movcs %xcc, 1, %g2
-; SPARC64-NEXT:    cmp %i5, 0
-; SPARC64-NEXT:    move %xcc, 0, %g2
+; SPARC64-NEXT:    mov %i3, %i5
+; SPARC64-NEXT:    movcs %xcc, 1, %i5
 ; SPARC64-NEXT:    cmp %i4, 0
 ; SPARC64-NEXT:    mov %i3, %i4
 ; SPARC64-NEXT:    movne %xcc, 1, %i4
 ; SPARC64-NEXT:    cmp %l0, 0
-; SPARC64-NEXT:    mov %i3, %i5
-; SPARC64-NEXT:    movne %xcc, 1, %i5
+; SPARC64-NEXT:    mov %i3, %g2
+; SPARC64-NEXT:    movne %xcc, 1, %g2
 ; SPARC64-NEXT:    cmp %i2, 0
 ; SPARC64-NEXT:    mov %i3, %i2
 ; SPARC64-NEXT:    movne %xcc, 1, %i2
 ; SPARC64-NEXT:    cmp %i0, 0
 ; SPARC64-NEXT:    movne %xcc, 1, %i3
 ; SPARC64-NEXT:    and %i3, %i2, %i0
-; SPARC64-NEXT:    or %i0, %i5, %i0
-; SPARC64-NEXT:    or %i0, %i4, %i0
 ; SPARC64-NEXT:    or %i0, %g2, %i0
+; SPARC64-NEXT:    or %i0, %i4, %i0
+; SPARC64-NEXT:    or %i0, %i5, %i0
 ; SPARC64-NEXT:    srl %i0, 0, %i2
 ; SPARC64-NEXT:    mov %i1, %i0
 ; SPARC64-NEXT:    ret
 ; SPARC64-NEXT:    restore %g0, %o1, %o1
 start:
   %0 = tail call { i128, i1 } @llvm.umul.with.overflow.i128(i128 %l, i128 %r) #2
   %1 = extractvalue { i128, i1 } %0, 0
   %2 = extractvalue { i128, i1 } %0, 1