[llvm] r301922 - [DAGCombine] (uaddo X, (addcarry Y, 0, Carry)) -> (addcarry X, Y, Carry)
Amaury Sechet via llvm-commits
llvm-commits at lists.llvm.org
Tue May 2 07:15:49 PDT 2017
Author: deadalnix
Date: Tue May 2 09:15:48 2017
New Revision: 301922
URL: http://llvm.org/viewvc/llvm-project?rev=301922&view=rev
Log:
[DAGCombine] (uaddo X, (addcarry Y, 0, Carry)) -> (addcarry X, Y, Carry)
Summary: This is a common pattern that arise when legalizing large integers operations. Only do it when Y + 1 cannot overflow as this would change the carry behavior of uaddo .
Reviewers: jyknight, nemanjai, mkuper, spatel, RKSimon, zvi, bkramer
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D32687
Modified:
llvm/trunk/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
llvm/trunk/test/CodeGen/X86/adde-carry.ll
llvm/trunk/test/CodeGen/X86/mul-i1024.ll
llvm/trunk/test/CodeGen/X86/mul-i256.ll
llvm/trunk/test/CodeGen/X86/mul-i512.ll
Modified: llvm/trunk/lib/CodeGen/SelectionDAG/DAGCombiner.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/CodeGen/SelectionDAG/DAGCombiner.cpp?rev=301922&r1=301921&r2=301922&view=diff
==============================================================================
--- llvm/trunk/lib/CodeGen/SelectionDAG/DAGCombiner.cpp (original)
+++ llvm/trunk/lib/CodeGen/SelectionDAG/DAGCombiner.cpp Tue May 2 09:15:48 2017
@@ -237,6 +237,7 @@ namespace {
SDValue visitSUB(SDNode *N);
SDValue visitADDC(SDNode *N);
SDValue visitUADDO(SDNode *N);
+ SDValue visitUADDOLike(SDValue N0, SDValue N1, SDNode *N);
SDValue visitSUBC(SDNode *N);
SDValue visitUSUBO(SDNode *N);
SDValue visitADDE(SDNode *N);
@@ -2083,6 +2084,26 @@ SDValue DAGCombiner::visitUADDO(SDNode *
return CombineTo(N, DAG.getNode(ISD::ADD, DL, VT, N0, N1),
DAG.getConstant(0, DL, CarryVT));
+ if (SDValue Combined = visitUADDOLike(N0, N1, N))
+ return Combined;
+
+ if (SDValue Combined = visitUADDOLike(N1, N0, N))
+ return Combined;
+
+ return SDValue();
+}
+
+SDValue DAGCombiner::visitUADDOLike(SDValue N0, SDValue N1, SDNode *N) {
+ // (uaddo X, (addcarry Y, 0, Carry)) -> (addcarry X, Y, Carry)
+ // If Y + 1 cannot overflow.
+ if (N1.getOpcode() == ISD::ADDCARRY && isNullConstant(N1.getOperand(1))) {
+ SDValue Y = N1.getOperand(0);
+ SDValue One = DAG.getConstant(1, SDLoc(N), Y.getValueType());
+ if (DAG.computeOverflowKind(Y, One) == SelectionDAG::OFK_Never)
+ return DAG.getNode(ISD::ADDCARRY, SDLoc(N), N->getVTList(), N0, Y,
+ N1.getOperand(2));
+ }
+
return SDValue();
}
Modified: llvm/trunk/test/CodeGen/X86/adde-carry.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/adde-carry.ll?rev=301922&r1=301921&r2=301922&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/adde-carry.ll (original)
+++ llvm/trunk/test/CodeGen/X86/adde-carry.ll Tue May 2 09:15:48 2017
@@ -86,24 +86,21 @@ entry:
define %scalar @pr31719(%scalar* nocapture readonly %this, %scalar %arg.b) {
; CHECK-LABEL: pr31719:
; CHECK: # BB#0: # %entry
-; CHECK-NEXT: addq (%rsi), %rdx
+; CHECK-NEXT: addq 8(%rsi), %rcx
; CHECK-NEXT: sbbq %r10, %r10
; CHECK-NEXT: andl $1, %r10d
-; CHECK-NEXT: addq 8(%rsi), %rcx
+; CHECK-NEXT: addq 16(%rsi), %r8
; CHECK-NEXT: sbbq %rax, %rax
; CHECK-NEXT: andl $1, %eax
-; CHECK-NEXT: addq %r10, %rcx
-; CHECK-NEXT: adcq $0, %rax
-; CHECK-NEXT: addq 16(%rsi), %r8
-; CHECK-NEXT: sbbq %r10, %r10
-; CHECK-NEXT: andl $1, %r10d
; CHECK-NEXT: addq 24(%rsi), %r9
-; CHECK-NEXT: addq %rax, %r8
-; CHECK-NEXT: adcq %r10, %r9
+; CHECK-NEXT: addq (%rsi), %rdx
+; CHECK-NEXT: adcq $0, %rcx
+; CHECK-NEXT: adcq %r8, %r10
+; CHECK-NEXT: adcq %r9, %rax
; CHECK-NEXT: movq %rdx, (%rdi)
; CHECK-NEXT: movq %rcx, 8(%rdi)
-; CHECK-NEXT: movq %r8, 16(%rdi)
-; CHECK-NEXT: movq %r9, 24(%rdi)
+; CHECK-NEXT: movq %r10, 16(%rdi)
+; CHECK-NEXT: movq %rax, 24(%rdi)
; CHECK-NEXT: movq %rdi, %rax
; CHECK-NEXT: retq
entry:
@@ -158,9 +155,8 @@ define void @muladd(%accumulator* nocapt
; CHECK-NEXT: movq %rdx, %rax
; CHECK-NEXT: mulq %rsi
; CHECK-NEXT: addq (%rdi), %rax
-; CHECK-NEXT: adcq $0, %rdx
; CHECK-NEXT: movq %rax, (%rdi)
-; CHECK-NEXT: addq 8(%rdi), %rdx
+; CHECK-NEXT: adcq 8(%rdi), %rdx
; CHECK-NEXT: movq %rdx, 8(%rdi)
; CHECK-NEXT: adcl $0, 16(%rdi)
; CHECK-NEXT: retq
Modified: llvm/trunk/test/CodeGen/X86/mul-i1024.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/mul-i1024.ll?rev=301922&r1=301921&r2=301922&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/mul-i1024.ll (original)
+++ llvm/trunk/test/CodeGen/X86/mul-i1024.ll Tue May 2 09:15:48 2017
@@ -4388,732 +4388,720 @@ define void @test_1024(i1024* %a, i1024*
; X64-NEXT: pushq %r13
; X64-NEXT: pushq %r12
; X64-NEXT: pushq %rbx
-; X64-NEXT: subq $360, %rsp # imm = 0x168
+; X64-NEXT: subq $352, %rsp # imm = 0x160
; X64-NEXT: movq %rdx, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq 48(%rdi), %r9
-; X64-NEXT: movq %r9, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq 48(%rdi), %r8
+; X64-NEXT: movq %r8, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq 40(%rdi), %rcx
; X64-NEXT: movq %rcx, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq 32(%rdi), %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rdi, %r10
-; X64-NEXT: xorl %r8d, %r8d
-; X64-NEXT: mulq %r8
-; X64-NEXT: movq %rdx, %rdi
+; X64-NEXT: movq %rdi, %r13
+; X64-NEXT: xorl %r9d, %r9d
+; X64-NEXT: mulq %r9
+; X64-NEXT: movq %rdx, %rbx
; X64-NEXT: movq %rax, %r11
; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: mulq %r8
+; X64-NEXT: mulq %r9
; X64-NEXT: movq %rax, %rcx
-; X64-NEXT: addq %rdi, %rcx
-; X64-NEXT: movq %rdx, %rbx
-; X64-NEXT: adcq $0, %rbx
+; X64-NEXT: addq %rbx, %rcx
+; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: adcq $0, %rbp
; X64-NEXT: addq %r11, %rcx
; X64-NEXT: movq %rcx, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rdi, %rbp
-; X64-NEXT: adcq $0, %rbp
-; X64-NEXT: addq %rbx, %rbp
+; X64-NEXT: adcq %rbx, %rbp
+; X64-NEXT: movq %rbx, %rcx
; X64-NEXT: sbbq %rbx, %rbx
; X64-NEXT: andl $1, %ebx
; X64-NEXT: addq %rax, %rbp
; X64-NEXT: adcq %rdx, %rbx
-; X64-NEXT: movq %r9, %rax
-; X64-NEXT: mulq %r8
+; X64-NEXT: movq %r8, %rax
+; X64-NEXT: mulq %r9
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq %rdx, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %r11, %r13
-; X64-NEXT: addq %rax, %r13
-; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: movq %rdi, %r15
-; X64-NEXT: movq %r15, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %r11, %r12
+; X64-NEXT: addq %rax, %r12
+; X64-NEXT: movq %rcx, %rax
+; X64-NEXT: movq %rcx, %r8
+; X64-NEXT: movq %r8, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: adcq %rdx, %rax
-; X64-NEXT: addq %rbp, %r13
-; X64-NEXT: movq %r13, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: addq %rbp, %r12
+; X64-NEXT: movq %r12, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: adcq %rbx, %rax
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rsi, %r8
-; X64-NEXT: movq (%r8), %rax
+; X64-NEXT: movq (%rsi), %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: xorl %ebp, %ebp
; X64-NEXT: mulq %rbp
-; X64-NEXT: movq %rax, %r14
+; X64-NEXT: movq %rax, %r10
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq 8(%r8), %rax
+; X64-NEXT: movq 8(%rsi), %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: mulq %rbp
; X64-NEXT: xorl %r9d, %r9d
-; X64-NEXT: movq %rax, %r12
-; X64-NEXT: addq %rcx, %r12
+; X64-NEXT: movq %rax, %r15
+; X64-NEXT: addq %rcx, %r15
; X64-NEXT: movq %rdx, %rbp
; X64-NEXT: adcq $0, %rbp
-; X64-NEXT: addq %r14, %r12
-; X64-NEXT: movq %rcx, %rbx
-; X64-NEXT: movq %rcx, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %rbx
-; X64-NEXT: addq %rbp, %rbx
-; X64-NEXT: sbbq %rbp, %rbp
-; X64-NEXT: andl $1, %ebp
-; X64-NEXT: addq %rax, %rbx
-; X64-NEXT: adcq %rdx, %rbp
-; X64-NEXT: movq 16(%r8), %rax
-; X64-NEXT: movq %r8, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: addq %r10, %r15
+; X64-NEXT: adcq %rcx, %rbp
+; X64-NEXT: movq %rcx, %rdi
+; X64-NEXT: movq %rdi, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: sbbq %rbx, %rbx
+; X64-NEXT: andl $1, %ebx
+; X64-NEXT: addq %rax, %rbp
+; X64-NEXT: adcq %rdx, %rbx
+; X64-NEXT: movq 16(%rsi), %rax
+; X64-NEXT: movq %rsi, %r14
+; X64-NEXT: movq %r14, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: mulq %r9
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq %rdx, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %r14, %r9
-; X64-NEXT: addq %rax, %r9
-; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: adcq %rdx, %rax
-; X64-NEXT: addq %rbx, %r9
-; X64-NEXT: adcq %rbp, %rax
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: movq %r11, %rax
+; X64-NEXT: movq %r10, %rcx
+; X64-NEXT: movq %rcx, %rsi
+; X64-NEXT: addq %rax, %rsi
+; X64-NEXT: movq %rdi, %r9
+; X64-NEXT: adcq %rdx, %r9
+; X64-NEXT: addq %rbp, %rsi
+; X64-NEXT: movq %rsi, %r10
+; X64-NEXT: adcq %rbx, %r9
; X64-NEXT: movq %r11, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: addq %r14, %rax
-; X64-NEXT: adcq %rcx, %r15
-; X64-NEXT: movq %r15, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq (%r10), %rax
+; X64-NEXT: movq %r11, %rax
+; X64-NEXT: addq %rcx, %rax
+; X64-NEXT: adcq %rdi, %r8
+; X64-NEXT: movq %r8, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %r13, %rbp
+; X64-NEXT: movq (%rbp), %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: xorl %r15d, %r15d
-; X64-NEXT: mulq %r15
+; X64-NEXT: xorl %r8d, %r8d
+; X64-NEXT: mulq %r8
; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq %rax, %rbx
-; X64-NEXT: addq %r14, %rax
-; X64-NEXT: movq %r14, %rdi
+; X64-NEXT: addq %rcx, %rax
; X64-NEXT: movq %rsi, %rax
-; X64-NEXT: adcq %rcx, %rax
-; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq 32(%r8), %rax
+; X64-NEXT: adcq %rdi, %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: mulq %r15
-; X64-NEXT: xorl %r8d, %r8d
+; X64-NEXT: movq 32(%r14), %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rdx, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rbx, %rcx
+; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rax, %r13
+; X64-NEXT: movq %rdx, (%rsp) # 8-byte Spill
+; X64-NEXT: movq %rbx, %rax
; X64-NEXT: movq %rbx, %r14
-; X64-NEXT: addq %rax, %rcx
+; X64-NEXT: addq %r13, %rax
; X64-NEXT: movq %rsi, %rax
; X64-NEXT: adcq %rdx, %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: addq %rdi, %r11
+; X64-NEXT: addq %rcx, %r11
; X64-NEXT: movq %r11, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rdi, %r11
-; X64-NEXT: movq %r11, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %rcx, %r8
+; X64-NEXT: movq %r8, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
-; X64-NEXT: adcq %r12, %rax
+; X64-NEXT: adcq %r15, %rax
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r9, %r13
-; X64-NEXT: movq %r13, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %r9, %rdi
+; X64-NEXT: adcq %r10, %r12
+; X64-NEXT: movq %r12, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %r10, %rcx
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
-; X64-NEXT: adcq %rbp, %rax
+; X64-NEXT: adcq %r9, %rax
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rbp, %r9
-; X64-NEXT: movq 8(%r10), %rax
+; X64-NEXT: movq %r9, %r10
+; X64-NEXT: movq 8(%rbp), %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %r10, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: mulq %r8
-; X64-NEXT: xorl %ecx, %ecx
-; X64-NEXT: movq %rax, %r15
-; X64-NEXT: addq %rsi, %r15
-; X64-NEXT: movq %rdx, %rbx
-; X64-NEXT: adcq $0, %rbx
-; X64-NEXT: addq %r14, %r15
-; X64-NEXT: movq %rsi, %rbp
-; X64-NEXT: movq %rsi, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %rbp
-; X64-NEXT: addq %rbx, %rbp
-; X64-NEXT: sbbq %r8, %r8
-; X64-NEXT: andl $1, %r8d
-; X64-NEXT: addq %rax, %rbp
-; X64-NEXT: adcq %rdx, %r8
-; X64-NEXT: movq 16(%r10), %rax
-; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rdx, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %r14, %rbx
-; X64-NEXT: addq %rax, %rbx
-; X64-NEXT: movq %rsi, %r10
-; X64-NEXT: adcq %rdx, %r10
-; X64-NEXT: addq %rbp, %rbx
-; X64-NEXT: adcq %r8, %r10
-; X64-NEXT: movq %r14, %rax
-; X64-NEXT: movq %r14, (%rsp) # 8-byte Spill
-; X64-NEXT: addq %r11, %rax
-; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %r15, %rcx
-; X64-NEXT: adcq %rcx, %r12
-; X64-NEXT: movq %r12, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %rbx, %rdi
+; X64-NEXT: movq %rbp, %rdi
; X64-NEXT: movq %rdi, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rbx, %r8
-; X64-NEXT: adcq %r10, %r9
-; X64-NEXT: movq %r9, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r13 # 8-byte Reload
-; X64-NEXT: movq 40(%r13), %rax
-; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: xorl %edx, %edx
; X64-NEXT: mulq %rdx
-; X64-NEXT: xorl %r11d, %r11d
-; X64-NEXT: movq %rax, %rsi
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r9 # 8-byte Reload
-; X64-NEXT: addq %r9, %rsi
-; X64-NEXT: movq %rdx, %rbx
-; X64-NEXT: adcq $0, %rbx
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: addq %rdi, %rsi
-; X64-NEXT: movq %r9, %rbp
+; X64-NEXT: xorl %r9d, %r9d
+; X64-NEXT: movq %rax, %r12
+; X64-NEXT: addq %rsi, %r12
+; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: adcq $0, %rbp
+; X64-NEXT: addq %r14, %r12
+; X64-NEXT: adcq %rsi, %rbp
+; X64-NEXT: movq %rsi, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: sbbq %rbx, %rbx
+; X64-NEXT: andl $1, %ebx
+; X64-NEXT: addq %rax, %rbp
+; X64-NEXT: adcq %rdx, %rbx
+; X64-NEXT: movq 16(%rdi), %rax
+; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: mulq %r9
+; X64-NEXT: xorl %edi, %edi
+; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %rdx, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %r14, %r9
+; X64-NEXT: addq %rax, %r9
+; X64-NEXT: adcq %rdx, %rsi
+; X64-NEXT: addq %rbp, %r9
+; X64-NEXT: movq %r9, %rdx
+; X64-NEXT: adcq %rbx, %rsi
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: movq %r14, %rsi
+; X64-NEXT: movq %r14, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: addq %r8, %rsi
+; X64-NEXT: movq %rsi, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %r12, %r15
+; X64-NEXT: movq %r15, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %rdx, %rcx
+; X64-NEXT: movq %rcx, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %rdx, %r15
+; X64-NEXT: adcq %rax, %r10
+; X64-NEXT: movq %r10, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %rax, %r10
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq 40(%rsi), %rax
+; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: mulq %rdi
+; X64-NEXT: xorl %r8d, %r8d
+; X64-NEXT: movq %rax, %rcx
+; X64-NEXT: movq (%rsp), %rdi # 8-byte Reload
+; X64-NEXT: addq %rdi, %rcx
+; X64-NEXT: movq %rdx, %rbp
; X64-NEXT: adcq $0, %rbp
-; X64-NEXT: addq %rbx, %rbp
+; X64-NEXT: addq %r13, %rcx
+; X64-NEXT: adcq %rdi, %rbp
; X64-NEXT: sbbq %rbx, %rbx
; X64-NEXT: andl $1, %ebx
; X64-NEXT: addq %rax, %rbp
; X64-NEXT: adcq %rdx, %rbx
-; X64-NEXT: movq 48(%r13), %rax
+; X64-NEXT: movq 48(%rsi), %rax
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: mulq %r11
+; X64-NEXT: mulq %r8
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq %rdx, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rdi, %r15
-; X64-NEXT: movq %rdi, %r11
-; X64-NEXT: addq %rax, %r15
-; X64-NEXT: movq %r9, %rdi
-; X64-NEXT: adcq %rdx, %rdi
-; X64-NEXT: addq %rbp, %r15
-; X64-NEXT: adcq %rbx, %rdi
-; X64-NEXT: addq %r11, %r14
+; X64-NEXT: movq %r13, %r8
+; X64-NEXT: addq %rax, %r8
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: adcq %rdx, %rax
+; X64-NEXT: addq %rbp, %r8
+; X64-NEXT: adcq %rbx, %rax
+; X64-NEXT: movq %r13, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: addq %r13, %r14
; X64-NEXT: movq %r14, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %rsi, %rcx
-; X64-NEXT: movq %rcx, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r15, %r8
-; X64-NEXT: movq %r8, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %rdi, %r10
+; X64-NEXT: adcq %rcx, %r12
+; X64-NEXT: movq %r12, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %r8, %r15
+; X64-NEXT: movq %r15, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %rax, %r10
; X64-NEXT: movq %r10, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r14 # 8-byte Reload
-; X64-NEXT: movq %r14, %rax
-; X64-NEXT: addq %r11, %rax
+; X64-NEXT: movq %rax, %rdx
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: addq %r13, %rax
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
-; X64-NEXT: adcq %r9, %rax
+; X64-NEXT: adcq %rdi, %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %r14, %rax
-; X64-NEXT: addq %r11, %rax
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: addq %r13, %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
-; X64-NEXT: movq %rsi, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r15 # 8-byte Folded Reload
-; X64-NEXT: movq %r15, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
-; X64-NEXT: movq %rdi, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
+; X64-NEXT: movq %rcx, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r8 # 8-byte Folded Reload
+; X64-NEXT: movq %r8, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rdx # 8-byte Folded Reload
+; X64-NEXT: movq %rdx, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: movq %rcx, %rax
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rax, %r9
-; X64-NEXT: movq %rdx, %rbx
+; X64-NEXT: movq %rax, %r8
+; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
-; X64-NEXT: movq 56(%rax), %rsi
-; X64-NEXT: movq %rsi, %rax
-; X64-NEXT: movq %rsi, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq 56(%rax), %r11
+; X64-NEXT: movq %r11, %rax
; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdi, %r10
-; X64-NEXT: movq %rdx, %rdi
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: addq %rbx, %rbp
-; X64-NEXT: adcq $0, %rdi
+; X64-NEXT: movq %rdi, %r9
+; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: movq %rax, %rbx
+; X64-NEXT: addq %rsi, %rbx
+; X64-NEXT: adcq $0, %rbp
; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rdx, %rbx
-; X64-NEXT: movq %rax, %r8
-; X64-NEXT: addq %rbp, %r8
-; X64-NEXT: adcq $0, %rbx
-; X64-NEXT: addq %rdi, %rbx
-; X64-NEXT: sbbq %rdi, %rdi
-; X64-NEXT: andl $1, %edi
-; X64-NEXT: movq %rsi, %rax
-; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rcx, %r11
-; X64-NEXT: addq %rbx, %rax
-; X64-NEXT: adcq %rdi, %rdx
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
+; X64-NEXT: mulq %rdi
+; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: movq %rax, %r10
+; X64-NEXT: addq %rbx, %r10
+; X64-NEXT: adcq %rbp, %rsi
+; X64-NEXT: sbbq %rcx, %rcx
+; X64-NEXT: andl $1, %ecx
+; X64-NEXT: movq %r11, %rax
+; X64-NEXT: mulq %rdi
+; X64-NEXT: movq %rdi, %r14
+; X64-NEXT: addq %rsi, %rax
+; X64-NEXT: adcq %rcx, %rdx
; X64-NEXT: movq {{[0-9]+}}(%rsp), %r15 # 8-byte Reload
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r15 # 8-byte Folded Reload
+; X64-NEXT: addq {{[0-9]+}}(%rsp), %r15 # 8-byte Folded Reload
; X64-NEXT: movq {{[0-9]+}}(%rsp), %r12 # 8-byte Reload
; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r12 # 8-byte Folded Reload
; X64-NEXT: addq %rax, %r15
; X64-NEXT: adcq %rdx, %r12
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: movq %r10, %rsi
-; X64-NEXT: mulq %rsi
-; X64-NEXT: movq %rdx, %r10
-; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: mulq %rsi
-; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: movq %r9, %rcx
+; X64-NEXT: mulq %rcx
+; X64-NEXT: movq %rdx, %r9
+; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: mulq %rcx
+; X64-NEXT: movq %rdx, %rbp
; X64-NEXT: movq %rax, %rbx
-; X64-NEXT: addq %r10, %rbx
+; X64-NEXT: addq %r9, %rbx
; X64-NEXT: adcq $0, %rbp
; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: movq %rdi, %r10
-; X64-NEXT: mulq %r11
-; X64-NEXT: movq %rdx, %rdi
+; X64-NEXT: mulq %r14
+; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: addq %rbx, %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %rdi
-; X64-NEXT: addq %rbp, %rdi
+; X64-NEXT: adcq %rbp, %rcx
; X64-NEXT: sbbq %rbp, %rbp
; X64-NEXT: andl $1, %ebp
-; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: movq %rcx, %rsi
-; X64-NEXT: mulq %r11
+; X64-NEXT: movq %rsi, %rbx
+; X64-NEXT: movq %rbx, %rax
+; X64-NEXT: mulq %r14
; X64-NEXT: movq %rdx, %r13
-; X64-NEXT: movq %rax, %rbx
-; X64-NEXT: addq %rdi, %rbx
+; X64-NEXT: movq %rax, %rsi
+; X64-NEXT: addq %rcx, %rsi
; X64-NEXT: adcq %rbp, %r13
-; X64-NEXT: addq {{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r13 # 8-byte Folded Reload
-; X64-NEXT: addq %r9, %rbx
-; X64-NEXT: adcq %r8, %r13
+; X64-NEXT: addq {{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r13 # 8-byte Folded Reload
+; X64-NEXT: addq %r8, %rsi
+; X64-NEXT: adcq %r10, %r13
; X64-NEXT: adcq $0, %r15
; X64-NEXT: adcq $0, %r12
-; X64-NEXT: movq %r10, %rbp
+; X64-NEXT: movq %rdi, %rbp
; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: mulq %rdi
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r9 # 8-byte Reload
+; X64-NEXT: mulq %r9
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq %rax, %r9
-; X64-NEXT: movq %rsi, %rax
-; X64-NEXT: movq %rsi, %r8
-; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %rdi
-; X64-NEXT: addq %rcx, %rdi
-; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
-; X64-NEXT: movq 24(%rax), %r14
+; X64-NEXT: movq %rax, %r10
+; X64-NEXT: movq %rbx, %rax
+; X64-NEXT: movq %rbx, %r14
+; X64-NEXT: mulq %r9
+; X64-NEXT: movq %rdx, %rdi
+; X64-NEXT: movq %rax, %rbx
+; X64-NEXT: addq %rcx, %rbx
+; X64-NEXT: adcq $0, %rdi
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
+; X64-NEXT: movq 24(%rax), %rcx
; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: mulq %r14
+; X64-NEXT: mulq %rcx
+; X64-NEXT: movq %rcx, %rbp
+; X64-NEXT: movq %rbp, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: addq %rdi, %rax
-; X64-NEXT: movq %rax, %rdi
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rsi, %rcx
-; X64-NEXT: sbbq %rsi, %rsi
-; X64-NEXT: andl $1, %esi
-; X64-NEXT: movq %r8, %rax
-; X64-NEXT: mulq %r14
+; X64-NEXT: movq %rax, %r8
+; X64-NEXT: addq %rbx, %r8
+; X64-NEXT: adcq %rdi, %rcx
+; X64-NEXT: sbbq %rdi, %rdi
+; X64-NEXT: andl $1, %edi
+; X64-NEXT: movq %r14, %rax
+; X64-NEXT: mulq %rbp
; X64-NEXT: addq %rcx, %rax
-; X64-NEXT: adcq %rsi, %rdx
+; X64-NEXT: adcq %rdi, %rdx
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbx # 8-byte Reload
+; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbp # 8-byte Reload
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r11 # 8-byte Reload
-; X64-NEXT: addq %r11, %rbp
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
-; X64-NEXT: addq %rax, %rbp
-; X64-NEXT: adcq %rdx, %rsi
-; X64-NEXT: addq %rbx, %r9
-; X64-NEXT: movq %r9, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r13, %rdi
-; X64-NEXT: movq %rdi, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r14 # 8-byte Reload
+; X64-NEXT: adcq %r14, %rbp
+; X64-NEXT: addq %rax, %rbx
+; X64-NEXT: adcq %rdx, %rbp
+; X64-NEXT: addq %rsi, %r10
+; X64-NEXT: movq %r10, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %r13, %r8
+; X64-NEXT: movq %r8, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq $0, %rbx
; X64-NEXT: adcq $0, %rbp
-; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: addq %r15, %rbp
-; X64-NEXT: adcq %r12, %rsi
-; X64-NEXT: movl $0, %r10d
-; X64-NEXT: adcq $0, %r10
-; X64-NEXT: sbbq %r15, %r15
-; X64-NEXT: andl $1, %r15d
+; X64-NEXT: addq %r15, %rbx
+; X64-NEXT: adcq %r12, %rbp
+; X64-NEXT: movl $0, %r8d
+; X64-NEXT: adcq $0, %r8
+; X64-NEXT: sbbq %r10, %r10
+; X64-NEXT: andl $1, %r10d
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdx, %r12
-; X64-NEXT: movq %rax, %r13
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r8 # 8-byte Reload
-; X64-NEXT: movq %r8, %rax
-; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdx, %rbx
+; X64-NEXT: movq %r9, %rsi
+; X64-NEXT: mulq %rsi
+; X64-NEXT: movq %rdx, %r9
+; X64-NEXT: movq %rax, %r15
+; X64-NEXT: movq %r11, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %r11, %rax
+; X64-NEXT: mulq %rsi
+; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq %rax, %rdi
-; X64-NEXT: addq %r12, %rdi
-; X64-NEXT: adcq $0, %rbx
+; X64-NEXT: addq %r9, %rdi
+; X64-NEXT: adcq $0, %rsi
; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: mulq %r14
-; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq %rax, %r9
-; X64-NEXT: addq %rdi, %r9
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rbx, %rcx
-; X64-NEXT: sbbq %rdi, %rdi
-; X64-NEXT: andl $1, %edi
-; X64-NEXT: movq %r8, %rax
-; X64-NEXT: mulq %r14
-; X64-NEXT: movq %r14, %r12
-; X64-NEXT: addq %rcx, %rax
-; X64-NEXT: adcq %rdi, %rdx
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
+; X64-NEXT: mulq %rcx
+; X64-NEXT: movq %rdx, %r9
+; X64-NEXT: movq %rax, %r13
+; X64-NEXT: addq %rdi, %r13
+; X64-NEXT: adcq %rsi, %r9
+; X64-NEXT: sbbq %rsi, %rsi
+; X64-NEXT: andl $1, %esi
+; X64-NEXT: movq %r11, %rax
+; X64-NEXT: mulq %rcx
+; X64-NEXT: movq %rcx, %r12
+; X64-NEXT: addq %r9, %rax
+; X64-NEXT: adcq %rsi, %rdx
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: addq %r11, %rcx
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
+; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: adcq %r14, %rsi
; X64-NEXT: addq %rax, %rcx
-; X64-NEXT: adcq %rdx, %rdi
-; X64-NEXT: addq %rbp, %r13
-; X64-NEXT: adcq %rsi, %r9
-; X64-NEXT: adcq %r10, %rcx
-; X64-NEXT: adcq %r15, %rdi
-; X64-NEXT: addq {{[0-9]+}}(%rsp), %r13 # 8-byte Folded Reload
+; X64-NEXT: adcq %rdx, %rsi
+; X64-NEXT: addq %rbx, %r15
+; X64-NEXT: adcq %rbp, %r13
+; X64-NEXT: adcq %r8, %rcx
+; X64-NEXT: adcq %r10, %rsi
+; X64-NEXT: addq {{[0-9]+}}(%rsp), %r15 # 8-byte Folded Reload
+; X64-NEXT: movq %r15, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r13 # 8-byte Folded Reload
; X64-NEXT: movq %r13, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r9 # 8-byte Folded Reload
-; X64-NEXT: movq %r9, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
-; X64-NEXT: movq %rcx, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
-; X64-NEXT: movq %rdi, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
+; X64-NEXT: movq %rcx, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
+; X64-NEXT: movq %rsi, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rax, %r11
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: mulq %rsi
+; X64-NEXT: movq %rax, %r9
; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
-; X64-NEXT: movq 24(%rax), %r8
-; X64-NEXT: movq %r8, %rax
-; X64-NEXT: movq %r8, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdi, %r13
-; X64-NEXT: movq %rdx, %rdi
+; X64-NEXT: movq 24(%rax), %rbp
+; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: movq %rbp, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: mulq %rsi
+; X64-NEXT: movq %rsi, %r14
+; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq %rax, %rbx
; X64-NEXT: addq %rcx, %rbx
-; X64-NEXT: adcq $0, %rdi
-; X64-NEXT: movq %rsi, %rax
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbp # 8-byte Reload
-; X64-NEXT: mulq %rbp
+; X64-NEXT: adcq $0, %rsi
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
+; X64-NEXT: mulq %rdi
; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: movq %rax, %r15
; X64-NEXT: addq %rbx, %r15
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rdi, %rcx
+; X64-NEXT: adcq %rsi, %rcx
; X64-NEXT: sbbq %rsi, %rsi
; X64-NEXT: andl $1, %esi
-; X64-NEXT: movq %r8, %rax
-; X64-NEXT: mulq %rbp
-; X64-NEXT: movq %rbp, %r14
+; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: mulq %rdi
+; X64-NEXT: movq %rdi, %r11
; X64-NEXT: addq %rcx, %rax
; X64-NEXT: adcq %rsi, %rdx
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r8 # 8-byte Reload
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r8 # 8-byte Folded Reload
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r10 # 8-byte Reload
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r8 # 8-byte Reload
+; X64-NEXT: addq {{[0-9]+}}(%rsp), %r8 # 8-byte Folded Reload
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r10 # 8-byte Reload
; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r10 # 8-byte Folded Reload
; X64-NEXT: addq %rax, %r8
; X64-NEXT: adcq %rdx, %r10
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: movq %rsi, %rax
-; X64-NEXT: mulq %r13
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: movq %r14, %rsi
+; X64-NEXT: mulq %rsi
; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbp # 8-byte Reload
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: mulq %r13
-; X64-NEXT: movq %rdx, %rdi
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r14 # 8-byte Reload
+; X64-NEXT: movq %r14, %rax
+; X64-NEXT: mulq %rsi
+; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq %rax, %rbx
; X64-NEXT: addq %rcx, %rbx
-; X64-NEXT: adcq $0, %rdi
-; X64-NEXT: movq %rsi, %rax
-; X64-NEXT: movq %rsi, %r9
-; X64-NEXT: mulq %r14
-; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: addq %rbx, %rax
-; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: addq %rdi, %rsi
-; X64-NEXT: sbbq %rdi, %rdi
-; X64-NEXT: andl $1, %edi
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: mulq %r14
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: mulq %r11
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq %rax, %rbx
-; X64-NEXT: addq %rsi, %rbx
-; X64-NEXT: adcq %rdi, %rcx
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
-; X64-NEXT: addq %r11, %rbx
-; X64-NEXT: adcq %r15, %rcx
+; X64-NEXT: addq %rbx, %rax
+; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %rsi, %rcx
+; X64-NEXT: sbbq %rbx, %rbx
+; X64-NEXT: andl $1, %ebx
+; X64-NEXT: movq %r14, %rax
+; X64-NEXT: mulq %r11
+; X64-NEXT: movq %rdx, %r11
+; X64-NEXT: movq %rax, %rsi
+; X64-NEXT: addq %rcx, %rsi
+; X64-NEXT: adcq %rbx, %r11
+; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
+; X64-NEXT: addq %r9, %rsi
+; X64-NEXT: adcq %r15, %r11
; X64-NEXT: adcq $0, %r8
; X64-NEXT: adcq $0, %r10
-; X64-NEXT: movq %r9, %rsi
-; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: movq %rdi, %rbx
+; X64-NEXT: movq %rbx, %rax
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rdi # 8-byte Reload
; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdx, %r9
-; X64-NEXT: movq %rax, %r11
-; X64-NEXT: movq %rbp, %r14
+; X64-NEXT: movq %rdx, %rcx
+; X64-NEXT: movq %rax, %r9
; X64-NEXT: movq %r14, %rax
; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdi, %r13
+; X64-NEXT: movq %rdi, %r15
; X64-NEXT: movq %rdx, %rdi
; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: addq %r9, %rbp
+; X64-NEXT: addq %rcx, %rbp
; X64-NEXT: adcq $0, %rdi
-; X64-NEXT: movq %rsi, %rax
-; X64-NEXT: mulq %r12
-; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: movq %rbx, %rax
+; X64-NEXT: movq %r12, %r13
+; X64-NEXT: mulq %r13
+; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: addq %rbp, %rax
-; X64-NEXT: movq %rax, %r15
-; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: addq %rdi, %rsi
+; X64-NEXT: movq %rax, %rbp
+; X64-NEXT: adcq %rdi, %rcx
; X64-NEXT: sbbq %rdi, %rdi
; X64-NEXT: andl $1, %edi
; X64-NEXT: movq %r14, %rax
-; X64-NEXT: mulq %r12
-; X64-NEXT: addq %rsi, %rax
+; X64-NEXT: mulq %r13
+; X64-NEXT: addq %rcx, %rax
; X64-NEXT: adcq %rdi, %rdx
-; X64-NEXT: movq (%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbx # 8-byte Reload
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r14 # 8-byte Reload
-; X64-NEXT: addq %r14, %rsi
+; X64-NEXT: addq %r14, %rbx
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
+; X64-NEXT: addq %rax, %rbx
+; X64-NEXT: adcq %rdx, %rcx
+; X64-NEXT: addq %rsi, %r9
+; X64-NEXT: movq %r9, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %r11, %rbp
+; X64-NEXT: movq %rbp, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq $0, %rbx
+; X64-NEXT: adcq $0, %rcx
+; X64-NEXT: addq %r8, %rbx
+; X64-NEXT: adcq %r10, %rcx
+; X64-NEXT: movl $0, %r12d
+; X64-NEXT: adcq $0, %r12
+; X64-NEXT: sbbq %r9, %r9
+; X64-NEXT: andl $1, %r9d
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbp # 8-byte Reload
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rbp # 8-byte Folded Reload
-; X64-NEXT: addq %rax, %rsi
-; X64-NEXT: adcq %rdx, %rbp
-; X64-NEXT: addq %rbx, %r11
-; X64-NEXT: movq %r11, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %rcx, %r15
-; X64-NEXT: movq %r15, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: mulq %r15
+; X64-NEXT: movq %rdx, %r8
+; X64-NEXT: movq %rax, %r11
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r10 # 8-byte Reload
+; X64-NEXT: movq %r10, %rax
+; X64-NEXT: mulq %r15
+; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: movq %rax, %rdi
+; X64-NEXT: addq %r8, %rdi
; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: adcq $0, %rbp
-; X64-NEXT: addq %r8, %rsi
-; X64-NEXT: adcq %r10, %rbp
-; X64-NEXT: movl $0, %r10d
-; X64-NEXT: adcq $0, %r10
-; X64-NEXT: sbbq %r15, %r15
-; X64-NEXT: andl $1, %r15d
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: movq %rcx, %rax
+; X64-NEXT: movq %rbp, %rax
; X64-NEXT: mulq %r13
-; X64-NEXT: movq %rdx, %r9
-; X64-NEXT: movq %rax, %r11
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r8 # 8-byte Reload
-; X64-NEXT: movq %r8, %rax
+; X64-NEXT: movq %rdx, %r8
+; X64-NEXT: movq %rax, %r15
+; X64-NEXT: addq %rdi, %r15
+; X64-NEXT: adcq %rsi, %r8
+; X64-NEXT: sbbq %rsi, %rsi
+; X64-NEXT: andl $1, %esi
+; X64-NEXT: movq %r10, %rax
+; X64-NEXT: movq %r10, %rbp
; X64-NEXT: mulq %r13
-; X64-NEXT: movq %rdx, %rdi
-; X64-NEXT: movq %rax, %rbx
-; X64-NEXT: addq %r9, %rbx
-; X64-NEXT: adcq $0, %rdi
-; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: movq %r12, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: mulq %r12
-; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq %rax, %r9
-; X64-NEXT: addq %rbx, %r9
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rdi, %rcx
-; X64-NEXT: sbbq %rdi, %rdi
-; X64-NEXT: andl $1, %edi
-; X64-NEXT: movq %r8, %rax
-; X64-NEXT: mulq %r12
-; X64-NEXT: addq %rcx, %rax
-; X64-NEXT: adcq %rdi, %rdx
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: addq %r14, %rcx
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r12 # 8-byte Reload
-; X64-NEXT: movq %r12, %r13
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r13 # 8-byte Folded Reload
-; X64-NEXT: addq %rax, %rcx
-; X64-NEXT: adcq %rdx, %r13
-; X64-NEXT: addq %rsi, %r11
-; X64-NEXT: adcq %rbp, %r9
-; X64-NEXT: adcq %r10, %rcx
-; X64-NEXT: adcq %r15, %r13
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r9 # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r13 # 8-byte Folded Reload
+; X64-NEXT: addq %r8, %rax
+; X64-NEXT: adcq %rsi, %rdx
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r10 # 8-byte Reload
+; X64-NEXT: movq %r10, %rsi
+; X64-NEXT: addq %r14, %rsi
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r8 # 8-byte Reload
+; X64-NEXT: movq %r8, %rdi
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
+; X64-NEXT: addq %rax, %rsi
+; X64-NEXT: adcq %rdx, %rdi
+; X64-NEXT: addq %rbx, %r11
+; X64-NEXT: adcq %rcx, %r15
+; X64-NEXT: adcq %r12, %rsi
+; X64-NEXT: adcq %r9, %rdi
; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r15 # 8-byte Folded Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
+; X64-NEXT: addq {{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
; X64-NEXT: movq %r11, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r9 # 8-byte Folded Reload
-; X64-NEXT: movq %r9, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
-; X64-NEXT: movq %rcx, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r13 # 8-byte Folded Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r15 # 8-byte Folded Reload
+; X64-NEXT: movq %r15, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
+; X64-NEXT: movq %rsi, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
+; X64-NEXT: movq %rdi, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: adcq $0, -{{[0-9]+}}(%rsp) # 8-byte Folded Spill
; X64-NEXT: adcq $0, -{{[0-9]+}}(%rsp) # 8-byte Folded Spill
-; X64-NEXT: adcq $0, {{[0-9]+}}(%rsp) # 8-byte Folded Spill
; X64-NEXT: adcq $0, -{{[0-9]+}}(%rsp) # 8-byte Folded Spill
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: adcq $0, -{{[0-9]+}}(%rsp) # 8-byte Folded Spill
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
+; X64-NEXT: movq %rdi, %rax
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rdx, %rdi
-; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %r8, %rax
+; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: movq %rax, %r14
+; X64-NEXT: movq %rbp, %rax
; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rcx, %r11
+; X64-NEXT: movq %rcx, %r13
; X64-NEXT: movq %rdx, %rbx
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: addq %rdi, %rbp
+; X64-NEXT: movq %rax, %rcx
+; X64-NEXT: addq %rsi, %rcx
; X64-NEXT: adcq $0, %rbx
-; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: movq %rdi, %rax
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq %rax, %r10
-; X64-NEXT: addq %rbp, %r10
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rbx, %rcx
-; X64-NEXT: sbbq %rsi, %rsi
-; X64-NEXT: andl $1, %esi
-; X64-NEXT: movq %r8, %rax
+; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: movq %rax, %r11
+; X64-NEXT: addq %rcx, %r11
+; X64-NEXT: adcq %rbx, %rsi
+; X64-NEXT: sbbq %rcx, %rcx
+; X64-NEXT: andl $1, %ecx
+; X64-NEXT: movq %rbp, %rax
; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdi, %r14
-; X64-NEXT: addq %rcx, %rax
-; X64-NEXT: adcq %rsi, %rdx
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r9 # 8-byte Reload
+; X64-NEXT: movq %rdi, %r15
+; X64-NEXT: addq %rsi, %rax
+; X64-NEXT: adcq %rcx, %rdx
+; X64-NEXT: movq %r10, %r9
; X64-NEXT: addq {{[0-9]+}}(%rsp), %r9 # 8-byte Folded Reload
-; X64-NEXT: movq %r12, %r15
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r15 # 8-byte Folded Reload
+; X64-NEXT: movq %r8, %r12
+; X64-NEXT: adcq (%rsp), %r12 # 8-byte Folded Reload
; X64-NEXT: addq %rax, %r9
-; X64-NEXT: adcq %rdx, %r15
+; X64-NEXT: adcq %rdx, %r12
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: movq %r11, %rdi
-; X64-NEXT: mulq %rdi
+; X64-NEXT: movq %r13, %rbp
+; X64-NEXT: mulq %rbp
; X64-NEXT: movq %rdx, %r8
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r11 # 8-byte Reload
-; X64-NEXT: movq %r11, %rax
-; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdx, %rbx
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: addq %r8, %rbp
-; X64-NEXT: adcq $0, %rbx
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r13 # 8-byte Reload
+; X64-NEXT: movq %r13, %rax
+; X64-NEXT: mulq %rbp
+; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: movq %rax, %rbx
+; X64-NEXT: addq %r8, %rbx
+; X64-NEXT: adcq $0, %rsi
; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: movq %rcx, %r8
-; X64-NEXT: mulq %r14
+; X64-NEXT: movq %rcx, %rbp
+; X64-NEXT: mulq %r15
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: addq %rbp, %rax
+; X64-NEXT: addq %rbx, %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rbx, %rcx
-; X64-NEXT: sbbq %rbp, %rbp
-; X64-NEXT: andl $1, %ebp
-; X64-NEXT: movq %r11, %rax
-; X64-NEXT: mulq %r14
-; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: adcq %rsi, %rcx
+; X64-NEXT: sbbq %rsi, %rsi
+; X64-NEXT: andl $1, %esi
+; X64-NEXT: movq %r13, %rax
+; X64-NEXT: mulq %r15
+; X64-NEXT: movq %rdx, %r8
; X64-NEXT: movq %rax, %rbx
; X64-NEXT: addq %rcx, %rbx
-; X64-NEXT: adcq %rbp, %rsi
+; X64-NEXT: adcq %rsi, %r8
; X64-NEXT: addq {{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
-; X64-NEXT: adcq %r10, %rsi
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r8 # 8-byte Folded Reload
+; X64-NEXT: addq %r14, %rbx
+; X64-NEXT: adcq %r11, %r8
; X64-NEXT: adcq $0, %r9
-; X64-NEXT: adcq $0, %r15
-; X64-NEXT: movq %r8, %rbp
-; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: adcq $0, %r12
+; X64-NEXT: movq %rbp, %rsi
+; X64-NEXT: movq %rsi, %rax
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: mulq %rcx
; X64-NEXT: movq %rdx, %r14
-; X64-NEXT: movq %rax, %r8
-; X64-NEXT: movq %r11, %rax
+; X64-NEXT: movq %rax, %r10
+; X64-NEXT: movq %r13, %rax
; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rdx, %r12
+; X64-NEXT: movq %rdx, %rdi
; X64-NEXT: movq %rax, %rcx
; X64-NEXT: addq %r14, %rcx
-; X64-NEXT: adcq $0, %r12
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
-; X64-NEXT: movq 56(%rax), %rdi
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdi, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: adcq $0, %rdi
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
+; X64-NEXT: movq 56(%rax), %r15
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: mulq %r15
+; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: addq %rcx, %rax
-; X64-NEXT: movq %rax, %r10
-; X64-NEXT: adcq $0, %rbp
-; X64-NEXT: addq %r12, %rbp
+; X64-NEXT: movq %rax, %r11
+; X64-NEXT: adcq %rdi, %rsi
; X64-NEXT: sbbq %rcx, %rcx
; X64-NEXT: andl $1, %ecx
-; X64-NEXT: movq %r11, %rax
-; X64-NEXT: mulq %rdi
-; X64-NEXT: addq %rbp, %rax
+; X64-NEXT: movq %r13, %rax
+; X64-NEXT: mulq %r15
+; X64-NEXT: addq %rsi, %rax
; X64-NEXT: adcq %rcx, %rdx
-; X64-NEXT: movq (%rsp), %rcx # 8-byte Reload
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r11 # 8-byte Reload
-; X64-NEXT: addq %r11, %rcx
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r12 # 8-byte Reload
-; X64-NEXT: adcq %r12, %rdi
-; X64-NEXT: addq %rax, %rcx
-; X64-NEXT: adcq %rdx, %rdi
-; X64-NEXT: addq %rbx, %r8
-; X64-NEXT: movq %r8, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %rsi, %r10
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rdi # 8-byte Reload
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r13 # 8-byte Reload
+; X64-NEXT: addq %r13, %rdi
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbp # 8-byte Reload
+; X64-NEXT: adcq %rbp, %rsi
+; X64-NEXT: addq %rax, %rdi
+; X64-NEXT: adcq %rdx, %rsi
+; X64-NEXT: addq %rbx, %r10
; X64-NEXT: movq %r10, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %rcx
+; X64-NEXT: adcq %r8, %r11
+; X64-NEXT: movq %r11, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: adcq $0, %rdi
-; X64-NEXT: addq %r9, %rcx
-; X64-NEXT: adcq %r15, %rdi
-; X64-NEXT: movl $0, %r8d
-; X64-NEXT: adcq $0, %r8
-; X64-NEXT: sbbq %r9, %r9
-; X64-NEXT: andl $1, %r9d
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: movq %rsi, %rax
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbx # 8-byte Reload
-; X64-NEXT: mulq %rbx
-; X64-NEXT: movq %rdx, %r10
-; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r14 # 8-byte Reload
-; X64-NEXT: movq %r14, %rax
-; X64-NEXT: mulq %rbx
-; X64-NEXT: movq %rdx, %rbx
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: addq %r10, %rbp
-; X64-NEXT: adcq $0, %rbx
-; X64-NEXT: movq %rsi, %rax
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r10 # 8-byte Reload
-; X64-NEXT: mulq %r10
-; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %r15
-; X64-NEXT: addq %rbp, %r15
; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: addq %rbx, %rsi
-; X64-NEXT: sbbq %rbp, %rbp
-; X64-NEXT: andl $1, %ebp
-; X64-NEXT: movq %r14, %rax
-; X64-NEXT: mulq %r10
-; X64-NEXT: addq %rsi, %rax
-; X64-NEXT: adcq %rbp, %rdx
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r10 # 8-byte Reload
-; X64-NEXT: addq %r11, %r10
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r14 # 8-byte Reload
-; X64-NEXT: adcq %r12, %r14
-; X64-NEXT: addq %rax, %r10
-; X64-NEXT: adcq %rdx, %r14
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
-; X64-NEXT: addq %rcx, %rax
-; X64-NEXT: adcq %rdi, %r15
-; X64-NEXT: adcq %r8, %r10
-; X64-NEXT: adcq %r9, %r14
-; X64-NEXT: addq {{[0-9]+}}(%rsp), %rax # 8-byte Folded Reload
+; X64-NEXT: addq %r9, %rdi
+; X64-NEXT: adcq %r12, %rsi
+; X64-NEXT: movl $0, %r14d
+; X64-NEXT: adcq $0, %r14
+; X64-NEXT: sbbq %r10, %r10
+; X64-NEXT: andl $1, %r10d
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbx # 8-byte Reload
+; X64-NEXT: movq %rbx, %rax
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
+; X64-NEXT: mulq %rcx
+; X64-NEXT: movq %rdx, %r8
+; X64-NEXT: movq %rax, %r12
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r9 # 8-byte Reload
+; X64-NEXT: movq %r9, %rax
+; X64-NEXT: mulq %rcx
+; X64-NEXT: movq %rdx, %r11
+; X64-NEXT: movq %rax, %rcx
+; X64-NEXT: addq %r8, %rcx
+; X64-NEXT: adcq $0, %r11
+; X64-NEXT: movq %rbx, %rax
+; X64-NEXT: movq %r15, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: mulq %r15
+; X64-NEXT: movq %rdx, %r8
+; X64-NEXT: movq %rax, %rbx
+; X64-NEXT: addq %rcx, %rbx
+; X64-NEXT: adcq %r11, %r8
+; X64-NEXT: sbbq %rcx, %rcx
+; X64-NEXT: andl $1, %ecx
+; X64-NEXT: movq %r9, %rax
+; X64-NEXT: mulq %r15
+; X64-NEXT: addq %r8, %rax
+; X64-NEXT: adcq %rcx, %rdx
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r15 # 8-byte Reload
+; X64-NEXT: addq %r13, %r15
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r11 # 8-byte Reload
+; X64-NEXT: adcq %rbp, %r11
+; X64-NEXT: addq %rax, %r15
+; X64-NEXT: adcq %rdx, %r11
+; X64-NEXT: addq %rdi, %r12
+; X64-NEXT: adcq %rsi, %rbx
+; X64-NEXT: adcq %r14, %r15
+; X64-NEXT: adcq %r10, %r11
+; X64-NEXT: addq {{[0-9]+}}(%rsp), %r12 # 8-byte Folded Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r15 # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r10 # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r14 # 8-byte Folded Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: addq %rcx, {{[0-9]+}}(%rsp) # 8-byte Folded Spill
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: adcq %rcx, {{[0-9]+}}(%rsp) # 8-byte Folded Spill
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: adcq %rcx, {{[0-9]+}}(%rsp) # 8-byte Folded Spill
-; X64-NEXT: adcq %r13, {{[0-9]+}}(%rsp) # 8-byte Folded Spill
-; X64-NEXT: adcq $0, %rax
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
+; X64-NEXT: adcq %rcx, {{[0-9]+}}(%rsp) # 8-byte Folded Spill
+; X64-NEXT: adcq $0, %r12
+; X64-NEXT: adcq $0, %rbx
; X64-NEXT: adcq $0, %r15
-; X64-NEXT: adcq $0, %r10
-; X64-NEXT: adcq $0, %r14
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rax # 8-byte Folded Reload
-; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq $0, %r11
+; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r12 # 8-byte Folded Reload
+; X64-NEXT: movq %r12, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
+; X64-NEXT: movq %rbx, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r15 # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r10 # 8-byte Folded Reload
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r14 # 8-byte Folded Reload
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
; X64-NEXT: movl $0, %eax
; X64-NEXT: adcq $0, %rax
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
@@ -5122,124 +5110,124 @@ define void @test_1024(i1024* %a, i1024*
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movl $0, %eax
; X64-NEXT: adcq $0, %rax
-; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: sbbq %rax, %rax
; X64-NEXT: andl $1, %eax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: movq %rdi, %rax
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbp # 8-byte Reload
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rcx, %r9
-; X64-NEXT: movq %rdx, %rbx
-; X64-NEXT: movq %rax, %rcx
-; X64-NEXT: addq %rsi, %rcx
-; X64-NEXT: adcq $0, %rbx
-; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: movq %rcx, %rax
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: movq %rdx, %r8
; X64-NEXT: movq %rax, %r13
-; X64-NEXT: addq %rcx, %r13
-; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: addq %rbx, %rsi
-; X64-NEXT: sbbq %rcx, %rcx
-; X64-NEXT: andl $1, %ecx
-; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq %rsi, %rax
; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdi, %r11
-; X64-NEXT: addq %rsi, %rax
-; X64-NEXT: adcq %rcx, %rdx
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r12 # 8-byte Reload
-; X64-NEXT: addq {{[0-9]+}}(%rsp), %r12 # 8-byte Folded Reload
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r8 # 8-byte Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r8 # 8-byte Folded Reload
-; X64-NEXT: addq %rax, %r12
-; X64-NEXT: adcq %rdx, %r8
+; X64-NEXT: movq %rdi, %r14
+; X64-NEXT: movq %rdx, %rbx
+; X64-NEXT: movq %rax, %rbp
+; X64-NEXT: addq %r8, %rbp
+; X64-NEXT: adcq $0, %rbx
+; X64-NEXT: movq %rcx, %rax
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r8 # 8-byte Reload
+; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rdx, %rcx
+; X64-NEXT: movq %rax, %r12
+; X64-NEXT: addq %rbp, %r12
+; X64-NEXT: adcq %rbx, %rcx
+; X64-NEXT: sbbq %rbp, %rbp
+; X64-NEXT: andl $1, %ebp
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: mulq %r8
+; X64-NEXT: addq %rcx, %rax
+; X64-NEXT: adcq %rbp, %rdx
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r9 # 8-byte Reload
+; X64-NEXT: addq {{[0-9]+}}(%rsp), %r9 # 8-byte Folded Reload
+; X64-NEXT: movq (%rsp), %r10 # 8-byte Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r10 # 8-byte Folded Reload
+; X64-NEXT: addq %rax, %r9
+; X64-NEXT: adcq %rdx, %r10
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: mulq %r9
+; X64-NEXT: movq %r14, %rbx
+; X64-NEXT: mulq %rbx
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbp # 8-byte Reload
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: mulq %r9
-; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %rbx
-; X64-NEXT: addq %rcx, %rbx
-; X64-NEXT: adcq $0, %rsi
+; X64-NEXT: movq %rax, (%rsp) # 8-byte Spill
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: mulq %rbx
+; X64-NEXT: movq %rdx, %rbx
+; X64-NEXT: movq %rax, %rbp
+; X64-NEXT: addq %rcx, %rbp
+; X64-NEXT: adcq $0, %rbx
; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: movq %rdi, %r9
-; X64-NEXT: mulq %r11
+; X64-NEXT: movq %rdi, %r14
+; X64-NEXT: mulq %r8
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: addq %rbx, %rax
+; X64-NEXT: addq %rbp, %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rsi, %rcx
+; X64-NEXT: adcq %rbx, %rcx
; X64-NEXT: sbbq %rdi, %rdi
-; X64-NEXT: andl $1, %edi
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: mulq %r11
-; X64-NEXT: movq %rdx, %r11
-; X64-NEXT: movq %rax, %rsi
-; X64-NEXT: addq %rcx, %rsi
-; X64-NEXT: adcq %rdi, %r11
-; X64-NEXT: addq {{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
-; X64-NEXT: adcq %r13, %r11
-; X64-NEXT: adcq $0, %r12
-; X64-NEXT: adcq $0, %r8
-; X64-NEXT: movq %r8, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %r9, %rdi
-; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rdx, %r8
-; X64-NEXT: movq %rax, %r13
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: movq %rbp, %r9
-; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rdx, %rbp
-; X64-NEXT: movq %rax, %rcx
-; X64-NEXT: addq %r8, %rcx
-; X64-NEXT: adcq $0, %rbp
-; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: andl $1, %edi
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rdx, %rbx
+; X64-NEXT: movq %rax, %rbp
+; X64-NEXT: addq %rcx, %rbp
+; X64-NEXT: adcq %rdi, %rbx
+; X64-NEXT: addq {{[0-9]+}}(%rsp), %rbp # 8-byte Folded Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
+; X64-NEXT: addq %r13, %rbp
+; X64-NEXT: adcq %r12, %rbx
+; X64-NEXT: adcq $0, %r9
+; X64-NEXT: movq %r9, %r12
+; X64-NEXT: adcq $0, %r10
+; X64-NEXT: movq %r10, %r8
+; X64-NEXT: movq %r14, %rax
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rdi # 8-byte Reload
; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdx, %rbx
-; X64-NEXT: addq %rcx, %rax
-; X64-NEXT: movq %rax, %r8
-; X64-NEXT: adcq $0, %rbx
-; X64-NEXT: addq %rbp, %rbx
-; X64-NEXT: sbbq %rcx, %rcx
-; X64-NEXT: andl $1, %ecx
-; X64-NEXT: movq %r9, %rax
+; X64-NEXT: movq %rdx, %rcx
+; X64-NEXT: movq %rax, %r9
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: movq %rsi, %r10
; X64-NEXT: mulq %rdi
-; X64-NEXT: addq %rbx, %rax
-; X64-NEXT: adcq %rcx, %rdx
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: addq {{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
+; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: movq %rax, %rdi
+; X64-NEXT: addq %rcx, %rdi
+; X64-NEXT: adcq $0, %rsi
+; X64-NEXT: movq %r14, %rax
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
-; X64-NEXT: addq %rax, %rdi
+; X64-NEXT: mulq %rcx
+; X64-NEXT: movq %rdx, %r14
+; X64-NEXT: addq %rdi, %rax
+; X64-NEXT: movq %rax, %rdi
+; X64-NEXT: adcq %rsi, %r14
+; X64-NEXT: sbbq %rsi, %rsi
+; X64-NEXT: andl $1, %esi
+; X64-NEXT: movq %r10, %rax
+; X64-NEXT: mulq %rcx
+; X64-NEXT: addq %r14, %rax
+; X64-NEXT: adcq %rsi, %rdx
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r14 # 8-byte Reload
+; X64-NEXT: addq %r14, %rsi
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r13 # 8-byte Reload
+; X64-NEXT: adcq %r13, %rcx
+; X64-NEXT: addq %rax, %rsi
; X64-NEXT: adcq %rdx, %rcx
-; X64-NEXT: addq %rsi, %r13
-; X64-NEXT: adcq %r11, %r8
-; X64-NEXT: movq %r8, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %rdi
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %r12, %rdi
+; X64-NEXT: addq %rbp, %r9
+; X64-NEXT: movq %r9, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %rbx, %rdi
; X64-NEXT: movq %rdi, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
-; X64-NEXT: movq %rcx, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movl $0, %r12d
-; X64-NEXT: adcq $0, %r12
+; X64-NEXT: adcq $0, %rsi
+; X64-NEXT: adcq $0, %rcx
+; X64-NEXT: addq %r12, %rsi
+; X64-NEXT: movq %rsi, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %r8, %rcx
+; X64-NEXT: movq %rcx, %r12
+; X64-NEXT: movl $0, %r10d
+; X64-NEXT: adcq $0, %r10
; X64-NEXT: sbbq %r9, %r9
; X64-NEXT: andl $1, %r9d
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbx # 8-byte Reload
@@ -5247,678 +5235,679 @@ define void @test_1024(i1024* %a, i1024*
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
; X64-NEXT: mulq %rsi
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq %rax, %r11
+; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r8 # 8-byte Reload
; X64-NEXT: movq %r8, %rax
; X64-NEXT: mulq %rsi
; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: addq %rcx, %rbp
+; X64-NEXT: movq %rax, %rdi
+; X64-NEXT: addq %rcx, %rdi
; X64-NEXT: adcq $0, %rsi
; X64-NEXT: movq %rbx, %rax
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: mulq %rdi
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbp # 8-byte Reload
+; X64-NEXT: mulq %rbp
; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: movq %rax, %rbx
-; X64-NEXT: addq %rbp, %rbx
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rsi, %rcx
+; X64-NEXT: addq %rdi, %rbx
+; X64-NEXT: adcq %rsi, %rcx
; X64-NEXT: sbbq %rsi, %rsi
; X64-NEXT: andl $1, %esi
; X64-NEXT: movq %r8, %rax
-; X64-NEXT: mulq %rdi
+; X64-NEXT: mulq %rbp
; X64-NEXT: addq %rcx, %rax
; X64-NEXT: adcq %rsi, %rdx
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: addq {{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
+; X64-NEXT: addq %r14, %rsi
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
+; X64-NEXT: adcq %r13, %rcx
; X64-NEXT: addq %rax, %rsi
; X64-NEXT: adcq %rdx, %rcx
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
-; X64-NEXT: adcq %r12, %rsi
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r13 # 8-byte Reload
+; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r13 # 8-byte Folded Reload
+; X64-NEXT: adcq %r12, %rbx
+; X64-NEXT: adcq %r10, %rsi
; X64-NEXT: adcq %r9, %rcx
-; X64-NEXT: addq {{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
+; X64-NEXT: addq {{[0-9]+}}(%rsp), %r13 # 8-byte Folded Reload
; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
-; X64-NEXT: addq %rax, {{[0-9]+}}(%rsp) # 8-byte Folded Spill
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
+; X64-NEXT: addq %rax, (%rsp) # 8-byte Folded Spill
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
+; X64-NEXT: adcq %rax, {{[0-9]+}}(%rsp) # 8-byte Folded Spill
; X64-NEXT: adcq %r15, {{[0-9]+}}(%rsp) # 8-byte Folded Spill
-; X64-NEXT: adcq %r10, %r13
+; X64-NEXT: adcq %r11, -{{[0-9]+}}(%rsp) # 8-byte Folded Spill
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r13 # 8-byte Folded Reload
; X64-NEXT: movq %r13, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r14, -{{[0-9]+}}(%rsp) # 8-byte Folded Spill
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
-; X64-NEXT: movq %r11, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
-; X64-NEXT: movq %rbx, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
+; X64-NEXT: movq %rbx, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
; X64-NEXT: movq %rsi, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
; X64-NEXT: movq %rcx, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: movq 64(%rsi), %r14
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbp # 8-byte Reload
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: mulq %r14
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r9 # 8-byte Reload
+; X64-NEXT: movq 64(%r9), %r11
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rdi # 8-byte Reload
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: mulq %r11
; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r8 # 8-byte Reload
-; X64-NEXT: movq %r8, %rax
-; X64-NEXT: mulq %r14
-; X64-NEXT: movq %rdx, %rdi
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r10 # 8-byte Reload
+; X64-NEXT: movq %r10, %rax
+; X64-NEXT: mulq %r11
+; X64-NEXT: movq %rdx, %rbp
; X64-NEXT: movq %rax, %rbx
; X64-NEXT: addq %rcx, %rbx
-; X64-NEXT: adcq $0, %rdi
-; X64-NEXT: movq 72(%rsi), %rcx
-; X64-NEXT: movq %rsi, %r13
-; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: adcq $0, %rbp
+; X64-NEXT: movq 72(%r9), %rcx
+; X64-NEXT: movq %rdi, %rax
; X64-NEXT: mulq %rcx
; X64-NEXT: movq %rcx, %rsi
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq %rax, %r10
-; X64-NEXT: addq %rbx, %r10
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rdi, %rcx
-; X64-NEXT: sbbq %rdi, %rdi
-; X64-NEXT: andl $1, %edi
-; X64-NEXT: movq %r8, %rax
+; X64-NEXT: movq %rax, %r8
+; X64-NEXT: addq %rbx, %r8
+; X64-NEXT: adcq %rbp, %rcx
+; X64-NEXT: sbbq %rbp, %rbp
+; X64-NEXT: andl $1, %ebp
+; X64-NEXT: movq %r10, %rax
; X64-NEXT: mulq %rsi
-; X64-NEXT: movq %rsi, %r8
+; X64-NEXT: movq %rsi, %r10
+; X64-NEXT: movq %r10, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: addq %rcx, %rbp
-; X64-NEXT: adcq %rdi, %rsi
-; X64-NEXT: movq %r14, %rax
+; X64-NEXT: movq %rax, %rdi
+; X64-NEXT: addq %rcx, %rdi
+; X64-NEXT: adcq %rbp, %rsi
+; X64-NEXT: movq %r11, %rax
; X64-NEXT: xorl %ecx, %ecx
; X64-NEXT: mulq %rcx
; X64-NEXT: movq %rax, %rbx
-; X64-NEXT: movq %rdx, %r11
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r15 # 8-byte Reload
-; X64-NEXT: addq %rbx, %r15
+; X64-NEXT: movq %rdx, %r14
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r12 # 8-byte Reload
-; X64-NEXT: adcq %r11, %r12
-; X64-NEXT: addq %rbp, %r15
-; X64-NEXT: adcq %rsi, %r12
+; X64-NEXT: addq %rbx, %r12
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r15 # 8-byte Reload
+; X64-NEXT: adcq %r14, %r15
+; X64-NEXT: addq %rdi, %r12
+; X64-NEXT: adcq %rsi, %r15
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: movq %r14, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: mulq %r14
-; X64-NEXT: movq %rdx, %r9
+; X64-NEXT: movq %r11, %rsi
+; X64-NEXT: movq %rsi, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: mulq %rsi
+; X64-NEXT: movq %rdx, %r11
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: mulq %r14
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbp # 8-byte Reload
+; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: mulq %rsi
; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: addq %r9, %rbp
+; X64-NEXT: movq %rax, %rdi
+; X64-NEXT: addq %r11, %rdi
; X64-NEXT: adcq $0, %rsi
; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: movq %rcx, %r14
-; X64-NEXT: movq %r8, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rcx, %r11
+; X64-NEXT: mulq %r10
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: addq %rbp, %rax
+; X64-NEXT: addq %rdi, %rax
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rsi, %rcx
+; X64-NEXT: adcq %rsi, %rcx
; X64-NEXT: sbbq %rsi, %rsi
; X64-NEXT: andl $1, %esi
-; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: movq %rdi, %r9
-; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: movq %rbp, %rdi
+; X64-NEXT: mulq %r10
; X64-NEXT: addq %rcx, %rax
; X64-NEXT: adcq %rsi, %rdx
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
+; X64-NEXT: addq {{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r13 # 8-byte Reload
+; X64-NEXT: adcq %r13, %r14
; X64-NEXT: addq %rax, %rbx
-; X64-NEXT: adcq %rdx, %r11
+; X64-NEXT: adcq %rdx, %r14
; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
-; X64-NEXT: adcq %r10, %r11
-; X64-NEXT: adcq $0, %r15
+; X64-NEXT: adcq %r8, %r14
; X64-NEXT: adcq $0, %r12
-; X64-NEXT: movq 80(%r13), %rbp
-; X64-NEXT: movq %r14, %rsi
-; X64-NEXT: movq %rsi, %rax
-; X64-NEXT: mulq %rbp
-; X64-NEXT: movq %rdx, %r8
-; X64-NEXT: movq %rax, %r14
-; X64-NEXT: movq %r9, %rax
-; X64-NEXT: mulq %rbp
+; X64-NEXT: adcq $0, %r15
+; X64-NEXT: movq %r9, %rbp
+; X64-NEXT: movq 80(%rbp), %r8
+; X64-NEXT: movq %r11, %rax
+; X64-NEXT: movq %r11, %r9
+; X64-NEXT: mulq %r8
; X64-NEXT: movq %rdx, %r10
+; X64-NEXT: movq %rax, %r11
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq %rax, %rcx
-; X64-NEXT: addq %r8, %rcx
-; X64-NEXT: adcq $0, %r10
-; X64-NEXT: movq 88(%r13), %r13
-; X64-NEXT: movq %rsi, %rax
-; X64-NEXT: mulq %r13
-; X64-NEXT: movq %rdx, %rdi
-; X64-NEXT: movq %rax, %r8
-; X64-NEXT: addq %rcx, %r8
-; X64-NEXT: adcq $0, %rdi
-; X64-NEXT: addq %r10, %rdi
+; X64-NEXT: addq %r10, %rcx
+; X64-NEXT: adcq $0, %rsi
+; X64-NEXT: movq 88(%rbp), %r10
+; X64-NEXT: movq %r9, %rax
+; X64-NEXT: mulq %r10
+; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: movq %rax, %r9
+; X64-NEXT: addq %rcx, %r9
+; X64-NEXT: adcq %rsi, %rbp
; X64-NEXT: sbbq %rsi, %rsi
; X64-NEXT: andl $1, %esi
-; X64-NEXT: movq %r9, %rax
-; X64-NEXT: mulq %r13
-; X64-NEXT: movq %rdx, %r10
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: mulq %r10
+; X64-NEXT: movq %rdx, %rdi
; X64-NEXT: movq %rax, %rcx
-; X64-NEXT: addq %rdi, %rcx
-; X64-NEXT: adcq %rsi, %r10
-; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: addq %rbp, %rcx
+; X64-NEXT: adcq %rsi, %rdi
+; X64-NEXT: movq %r8, %rax
; X64-NEXT: xorl %edx, %edx
; X64-NEXT: mulq %rdx
-; X64-NEXT: movq %rdx, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rax, %rdi
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: addq %rdi, %rsi
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
+; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: addq %rax, %rsi
+; X64-NEXT: movq %r13, %rax
; X64-NEXT: adcq %rdx, %rax
+; X64-NEXT: movq %rdx, %r13
; X64-NEXT: addq %rcx, %rsi
-; X64-NEXT: adcq %r10, %rax
-; X64-NEXT: addq %rbx, %r14
-; X64-NEXT: movq %r14, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r11, %r8
-; X64-NEXT: movq %r8, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %rdi, %rax
+; X64-NEXT: addq %rbx, %r11
+; X64-NEXT: movq %r11, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %r14, %r9
+; X64-NEXT: movq %r9, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: adcq $0, %rsi
; X64-NEXT: adcq $0, %rax
-; X64-NEXT: addq %r15, %rsi
-; X64-NEXT: movq %rsi, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r12, %rax
-; X64-NEXT: movq %rax, %r10
-; X64-NEXT: movl $0, %r15d
-; X64-NEXT: adcq $0, %r15
-; X64-NEXT: sbbq %r12, %r12
-; X64-NEXT: andl $1, %r12d
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: mulq %rbp
-; X64-NEXT: movq %rdx, %r8
-; X64-NEXT: movq %rax, %r14
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r11 # 8-byte Reload
-; X64-NEXT: movq %r11, %rax
-; X64-NEXT: mulq %rbp
+; X64-NEXT: addq %r12, %rsi
+; X64-NEXT: movq %rsi, %r14
+; X64-NEXT: adcq %r15, %rax
+; X64-NEXT: movq %rax, %r9
+; X64-NEXT: movl $0, %r12d
+; X64-NEXT: adcq $0, %r12
+; X64-NEXT: sbbq %r11, %r11
+; X64-NEXT: andl $1, %r11d
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rdi # 8-byte Reload
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rdx, %rcx
+; X64-NEXT: movq %rax, %r15
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbp # 8-byte Reload
+; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: mulq %r8
; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq %rax, %rbx
-; X64-NEXT: addq %r8, %rbx
+; X64-NEXT: addq %rcx, %rbx
; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: mulq %r13
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: mulq %r10
; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: addq %rbx, %rax
-; X64-NEXT: movq %rax, %rbx
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rsi, %rcx
+; X64-NEXT: movq %rax, %rdi
+; X64-NEXT: adcq %rsi, %rcx
; X64-NEXT: sbbq %rsi, %rsi
; X64-NEXT: andl $1, %esi
-; X64-NEXT: movq %r11, %rax
-; X64-NEXT: mulq %r13
+; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: mulq %r10
; X64-NEXT: addq %rcx, %rax
; X64-NEXT: adcq %rsi, %rdx
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: addq %rdi, %rsi
+; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
+; X64-NEXT: adcq %r13, %rcx
; X64-NEXT: addq %rax, %rsi
; X64-NEXT: adcq %rdx, %rcx
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r14 # 8-byte Folded Reload
-; X64-NEXT: movq %r14, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r10, %rbx
-; X64-NEXT: movq %rbx, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r15, %rsi
+; X64-NEXT: addq %r14, %r15
+; X64-NEXT: movq %r15, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %r9, %rdi
+; X64-NEXT: movq %rdi, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %r12, %rsi
; X64-NEXT: movq %rsi, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r12, %rcx
+; X64-NEXT: adcq %r11, %rcx
; X64-NEXT: movq %rcx, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
-; X64-NEXT: imulq %rax, %r13
-; X64-NEXT: movq %rax, %r8
-; X64-NEXT: mulq %rbp
+; X64-NEXT: imulq %rax, %r10
+; X64-NEXT: movq %rax, %rdi
+; X64-NEXT: mulq %r8
; X64-NEXT: movq %rax, %r9
-; X64-NEXT: addq %r13, %rdx
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: imulq %rdi, %rbp
-; X64-NEXT: addq %rdx, %rbp
+; X64-NEXT: addq %r10, %rdx
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r10 # 8-byte Reload
+; X64-NEXT: imulq %r10, %r8
+; X64-NEXT: addq %rdx, %r8
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
; X64-NEXT: movq %rax, %rsi
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r12 # 8-byte Reload
-; X64-NEXT: imulq %r12, %rsi
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rax, %r10
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbx # 8-byte Reload
+; X64-NEXT: imulq %rbx, %rsi
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbp # 8-byte Reload
+; X64-NEXT: mulq %rbp
+; X64-NEXT: movq %rax, %r11
; X64-NEXT: addq %rsi, %rdx
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
-; X64-NEXT: imulq %rcx, %rax
+; X64-NEXT: imulq %rbp, %rax
; X64-NEXT: addq %rdx, %rax
-; X64-NEXT: addq %r9, %r10
-; X64-NEXT: adcq %rbp, %rax
-; X64-NEXT: movq %rax, %r9
-; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: movq %rcx, %rbp
-; X64-NEXT: mulq %r8
+; X64-NEXT: addq %r9, %r11
+; X64-NEXT: adcq %r8, %rax
+; X64-NEXT: movq %rax, %r14
+; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: mulq %rdi
; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %r12, %rax
-; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rbx, %rax
+; X64-NEXT: movq %rbx, %r8
+; X64-NEXT: mulq %rdi
; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq %rax, %rbx
; X64-NEXT: addq %rcx, %rbx
; X64-NEXT: adcq $0, %rsi
; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdx, %rbp
-; X64-NEXT: movq %rax, %r14
-; X64-NEXT: addq %rbx, %r14
-; X64-NEXT: adcq $0, %rbp
-; X64-NEXT: addq %rsi, %rbp
+; X64-NEXT: movq %r10, %rbp
+; X64-NEXT: mulq %rbp
+; X64-NEXT: movq %rdx, %rdi
+; X64-NEXT: movq %rax, %r15
+; X64-NEXT: addq %rbx, %r15
+; X64-NEXT: adcq %rsi, %rdi
; X64-NEXT: sbbq %rcx, %rcx
; X64-NEXT: andl $1, %ecx
-; X64-NEXT: movq %r12, %rax
-; X64-NEXT: mulq %rdi
+; X64-NEXT: movq %r8, %rax
+; X64-NEXT: mulq %rbp
; X64-NEXT: movq %rdx, %r12
-; X64-NEXT: movq %rax, %r8
-; X64-NEXT: addq %rbp, %r8
+; X64-NEXT: movq %rax, %r9
+; X64-NEXT: addq %rdi, %r9
; X64-NEXT: adcq %rcx, %r12
-; X64-NEXT: addq %r10, %r8
-; X64-NEXT: adcq %r9, %r12
+; X64-NEXT: addq %r11, %r9
+; X64-NEXT: adcq %r14, %r12
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdx # 8-byte Reload
; X64-NEXT: movq 120(%rdx), %rcx
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r9 # 8-byte Reload
-; X64-NEXT: imulq %r9, %rcx
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r10 # 8-byte Reload
+; X64-NEXT: imulq %r10, %rcx
; X64-NEXT: movq 112(%rdx), %rsi
-; X64-NEXT: movq %rdx, %rdi
-; X64-NEXT: movq %r9, %rax
+; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: movq %r10, %rax
; X64-NEXT: mulq %rsi
-; X64-NEXT: movq %rax, %r15
+; X64-NEXT: movq %rax, %r11
; X64-NEXT: addq %rcx, %rdx
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r10 # 8-byte Reload
-; X64-NEXT: imulq %r10, %rsi
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r8 # 8-byte Reload
+; X64-NEXT: imulq %r8, %rsi
; X64-NEXT: addq %rdx, %rsi
-; X64-NEXT: movq 96(%rdi), %rbp
-; X64-NEXT: movq 104(%rdi), %rbx
+; X64-NEXT: movq 96(%rbp), %rdi
+; X64-NEXT: movq 104(%rbp), %rbx
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
; X64-NEXT: movq %rax, %rcx
; X64-NEXT: imulq %rbx, %rcx
-; X64-NEXT: mulq %rbp
+; X64-NEXT: mulq %rdi
; X64-NEXT: movq %rax, %r13
; X64-NEXT: addq %rcx, %rdx
-; X64-NEXT: imulq %rbp, %r11
-; X64-NEXT: addq %rdx, %r11
-; X64-NEXT: addq %r15, %r13
-; X64-NEXT: adcq %rsi, %r11
-; X64-NEXT: movq %r11, %r15
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: mulq %r9
-; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
+; X64-NEXT: imulq %rdi, %rax
+; X64-NEXT: addq %rdx, %rax
+; X64-NEXT: addq %r11, %r13
+; X64-NEXT: adcq %rsi, %rax
; X64-NEXT: movq %rax, %r11
-; X64-NEXT: movq %rbx, %rax
-; X64-NEXT: mulq %r9
-; X64-NEXT: movq %rdx, %rdi
-; X64-NEXT: movq %rax, %rcx
-; X64-NEXT: addq %rsi, %rcx
-; X64-NEXT: adcq $0, %rdi
-; X64-NEXT: movq %rbp, %rax
+; X64-NEXT: movq %rdi, %rax
; X64-NEXT: mulq %r10
-; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: addq %rcx, %rbp
-; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: addq %rdi, %rsi
-; X64-NEXT: sbbq %rcx, %rcx
-; X64-NEXT: andl $1, %ecx
+; X64-NEXT: movq %rdx, %rcx
+; X64-NEXT: movq %rax, %r14
; X64-NEXT: movq %rbx, %rax
; X64-NEXT: mulq %r10
-; X64-NEXT: addq %rsi, %rax
-; X64-NEXT: adcq %rcx, %rdx
+; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: movq %rax, %rsi
+; X64-NEXT: addq %rcx, %rsi
+; X64-NEXT: adcq $0, %rbp
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rdx, %rcx
+; X64-NEXT: movq %rax, %rdi
+; X64-NEXT: addq %rsi, %rdi
+; X64-NEXT: adcq %rbp, %rcx
+; X64-NEXT: sbbq %rsi, %rsi
+; X64-NEXT: andl $1, %esi
+; X64-NEXT: movq %rbx, %rax
+; X64-NEXT: mulq %r8
+; X64-NEXT: addq %rcx, %rax
+; X64-NEXT: adcq %rsi, %rdx
; X64-NEXT: addq %r13, %rax
-; X64-NEXT: adcq %r15, %rdx
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
-; X64-NEXT: adcq %r14, %rbp
-; X64-NEXT: adcq %r8, %rax
+; X64-NEXT: adcq %r11, %rdx
+; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r14 # 8-byte Folded Reload
+; X64-NEXT: adcq %r15, %rdi
+; X64-NEXT: adcq %r9, %rax
; X64-NEXT: adcq %r12, %rdx
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
-; X64-NEXT: movq %r11, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rbp # 8-byte Folded Reload
-; X64-NEXT: movq %rbp, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r14 # 8-byte Folded Reload
+; X64-NEXT: movq %r14, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
+; X64-NEXT: movq %rdi, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rax # 8-byte Folded Reload
; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rdx # 8-byte Folded Reload
; X64-NEXT: movq %rdx, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: movq 80(%rsi), %r9
-; X64-NEXT: movq %r9, %rax
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq 88(%rsi), %r8
-; X64-NEXT: movq %rsi, %r11
-; X64-NEXT: movq %r8, %rax
-; X64-NEXT: movq %r8, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rdi, %rbx
-; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %rdi
-; X64-NEXT: addq %rcx, %rdi
-; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: movq %r9, %rax
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq 80(%rsi), %rcx
+; X64-NEXT: movq %rcx, %rax
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r8 # 8-byte Reload
+; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %rdx, %r9
+; X64-NEXT: movq 88(%rsi), %r10
+; X64-NEXT: movq %rsi, %r12
+; X64-NEXT: movq %r10, %rax
+; X64-NEXT: movq %r10, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: movq %rax, %rbx
+; X64-NEXT: addq %r9, %rbx
+; X64-NEXT: adcq $0, %rbp
+; X64-NEXT: movq %rcx, %rax
+; X64-NEXT: movq %rcx, %r9
; X64-NEXT: movq %r9, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq {{[0-9]+}}(%rsp), %r15 # 8-byte Reload
; X64-NEXT: mulq %r15
-; X64-NEXT: movq %rdx, %rbp
-; X64-NEXT: movq %rax, %r14
-; X64-NEXT: addq %rdi, %r14
-; X64-NEXT: adcq $0, %rbp
-; X64-NEXT: addq %rsi, %rbp
+; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: addq %rbx, %rax
+; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %rbp, %rsi
; X64-NEXT: sbbq %rdi, %rdi
; X64-NEXT: andl $1, %edi
-; X64-NEXT: movq %r8, %rax
+; X64-NEXT: movq %r10, %rax
; X64-NEXT: mulq %r15
-; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %rcx
-; X64-NEXT: addq %rbp, %rcx
-; X64-NEXT: adcq %rdi, %rsi
+; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: movq %rax, %rbx
+; X64-NEXT: addq %rsi, %rbx
+; X64-NEXT: adcq %rdi, %rbp
; X64-NEXT: movq %r9, %rax
-; X64-NEXT: xorl %r13d, %r13d
-; X64-NEXT: mulq %r13
-; X64-NEXT: movq %rdx, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rax, %r12
-; X64-NEXT: movq (%rsp), %r10 # 8-byte Reload
-; X64-NEXT: addq %r10, %r12
-; X64-NEXT: movq %rdx, %r8
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r9 # 8-byte Reload
-; X64-NEXT: adcq %r9, %r8
-; X64-NEXT: addq %rcx, %r12
-; X64-NEXT: adcq %rsi, %r8
-; X64-NEXT: movq %r11, %rsi
-; X64-NEXT: movq 64(%rsi), %r11
-; X64-NEXT: movq %r11, %rax
-; X64-NEXT: movq %rbx, %rdi
-; X64-NEXT: mulq %rdi
-; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq 72(%rsi), %rbx
-; X64-NEXT: movq %rbx, %rax
-; X64-NEXT: mulq %rdi
+; X64-NEXT: xorl %ecx, %ecx
+; X64-NEXT: mulq %rcx
+; X64-NEXT: movq %rdx, %r11
+; X64-NEXT: movq %r11, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %rax, %r10
+; X64-NEXT: movq %r10, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r9 # 8-byte Reload
+; X64-NEXT: addq %r9, %r10
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r13 # 8-byte Reload
+; X64-NEXT: adcq %r13, %r11
+; X64-NEXT: addq %rbx, %r10
+; X64-NEXT: adcq %rbp, %r11
+; X64-NEXT: movq %r12, %rcx
+; X64-NEXT: movq 64(%rcx), %rdi
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: addq %rcx, %rbp
-; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: movq %r11, %rax
+; X64-NEXT: movq 72(%rcx), %r14
+; X64-NEXT: movq %r14, %rax
+; X64-NEXT: mulq %r8
+; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: movq %rax, %rbx
+; X64-NEXT: addq %rsi, %rbx
+; X64-NEXT: adcq $0, %rbp
+; X64-NEXT: movq %rdi, %rax
; X64-NEXT: mulq %r15
-; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: addq %rbp, %rax
-; X64-NEXT: movq %rax, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rsi, %rcx
-; X64-NEXT: sbbq %rdi, %rdi
-; X64-NEXT: andl $1, %edi
-; X64-NEXT: movq %rbx, %rax
+; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: addq %rbx, %rax
+; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %rbp, %rsi
+; X64-NEXT: sbbq %rcx, %rcx
+; X64-NEXT: andl $1, %ecx
+; X64-NEXT: movq %r14, %rax
; X64-NEXT: mulq %r15
+; X64-NEXT: movq %rdx, %rbx
+; X64-NEXT: movq %rax, %rbp
+; X64-NEXT: addq %rsi, %rbp
+; X64-NEXT: adcq %rcx, %rbx
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: xorl %ecx, %ecx
+; X64-NEXT: mulq %rcx
+; X64-NEXT: movq %rdx, %r8
+; X64-NEXT: movq %rax, %r15
+; X64-NEXT: addq %r15, %r9
+; X64-NEXT: movq %r13, %rax
+; X64-NEXT: adcq %r8, %rax
+; X64-NEXT: addq %rbp, %r9
+; X64-NEXT: adcq %rbx, %rax
+; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r9 # 8-byte Folded Reload
+; X64-NEXT: movq %r9, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rax # 8-byte Folded Reload
+; X64-NEXT: movq %rax, %r13
+; X64-NEXT: adcq $0, %r10
+; X64-NEXT: adcq $0, %r11
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: movq %rdi, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: mulq %rsi
+; X64-NEXT: movq %rdx, %rcx
+; X64-NEXT: movq %rax, %rbx
+; X64-NEXT: movq %r14, %rax
+; X64-NEXT: movq %r14, %r12
+; X64-NEXT: movq %r12, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: mulq %rsi
+; X64-NEXT: movq %rsi, %r14
; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq %rax, %rbp
; X64-NEXT: addq %rcx, %rbp
-; X64-NEXT: adcq %rdi, %rsi
-; X64-NEXT: movq %r11, %rdi
-; X64-NEXT: movq %rdi, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: mulq %r13
-; X64-NEXT: movq %rdx, %r11
-; X64-NEXT: movq %rax, %r13
-; X64-NEXT: addq %r13, %r10
-; X64-NEXT: adcq %r11, %r9
-; X64-NEXT: addq %rbp, %r10
-; X64-NEXT: adcq %rsi, %r9
-; X64-NEXT: addq {{[0-9]+}}(%rsp), %r10 # 8-byte Folded Reload
-; X64-NEXT: movq %r10, (%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r14, %r9
-; X64-NEXT: movq %r9, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %r12
-; X64-NEXT: adcq $0, %r8
+; X64-NEXT: adcq $0, %rsi
; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: mulq %rcx
; X64-NEXT: movq %rdx, %r9
-; X64-NEXT: movq %rax, %r15
-; X64-NEXT: movq %rbx, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rbx, %rax
-; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rdx, %rbp
-; X64-NEXT: movq %rax, %rsi
-; X64-NEXT: addq %r9, %rsi
-; X64-NEXT: adcq $0, %rbp
-; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r10 # 8-byte Reload
-; X64-NEXT: mulq %r10
-; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: movq %rax, %rdi
-; X64-NEXT: addq %rsi, %rdi
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rbp, %rcx
+; X64-NEXT: addq %rbp, %rdi
+; X64-NEXT: adcq %rsi, %r9
; X64-NEXT: sbbq %rsi, %rsi
; X64-NEXT: andl $1, %esi
-; X64-NEXT: movq %rbx, %rax
-; X64-NEXT: mulq %r10
-; X64-NEXT: addq %rcx, %rax
+; X64-NEXT: movq %r12, %rax
+; X64-NEXT: mulq %rcx
+; X64-NEXT: movq %rcx, %rbp
+; X64-NEXT: addq %r9, %rax
; X64-NEXT: adcq %rsi, %rdx
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r13 # 8-byte Folded Reload
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r14 # 8-byte Reload
-; X64-NEXT: adcq %r14, %r11
-; X64-NEXT: addq %rax, %r13
-; X64-NEXT: adcq %rdx, %r11
-; X64-NEXT: addq (%rsp), %r15 # 8-byte Folded Reload
-; X64-NEXT: movq %r15, (%rsp) # 8-byte Spill
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r12 # 8-byte Reload
+; X64-NEXT: addq %r12, %r15
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r8 # 8-byte Folded Reload
+; X64-NEXT: addq %rax, %r15
+; X64-NEXT: adcq %rdx, %r8
+; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
+; X64-NEXT: movq %rbx, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %r13, %rdi
; X64-NEXT: movq %rdi, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %r13
-; X64-NEXT: adcq $0, %r11
-; X64-NEXT: addq %r12, %r13
-; X64-NEXT: adcq %r8, %r11
-; X64-NEXT: movl $0, %r8d
+; X64-NEXT: adcq $0, %r15
; X64-NEXT: adcq $0, %r8
-; X64-NEXT: sbbq %r9, %r9
-; X64-NEXT: andl $1, %r9d
+; X64-NEXT: addq %r10, %r15
+; X64-NEXT: adcq %r11, %r8
+; X64-NEXT: movl $0, %r9d
+; X64-NEXT: adcq $0, %r9
+; X64-NEXT: sbbq %r13, %r13
+; X64-NEXT: andl $1, %r13d
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r15 # 8-byte Reload
-; X64-NEXT: mulq %r15
-; X64-NEXT: movq %rdx, %r12
-; X64-NEXT: movq %rax, %rdi
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbx # 8-byte Reload
-; X64-NEXT: movq %rbx, %rax
-; X64-NEXT: mulq %r15
+; X64-NEXT: movq %r14, %rsi
+; X64-NEXT: mulq %rsi
+; X64-NEXT: movq %rdx, %r14
+; X64-NEXT: movq %rax, %r10
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rdi # 8-byte Reload
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: mulq %rsi
+; X64-NEXT: movq %rsi, %r11
; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: addq %r12, %rbp
+; X64-NEXT: movq %rax, %rbx
+; X64-NEXT: addq %r14, %rbx
; X64-NEXT: adcq $0, %rsi
; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: mulq %r10
+; X64-NEXT: mulq %rbp
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: addq %rbp, %rax
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rsi, %rcx
+; X64-NEXT: addq %rbx, %rax
+; X64-NEXT: movq %rax, %rbx
+; X64-NEXT: adcq %rsi, %rcx
; X64-NEXT: sbbq %rsi, %rsi
; X64-NEXT: andl $1, %esi
-; X64-NEXT: movq %rbx, %rax
-; X64-NEXT: mulq %r10
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: movq %rdi, %r14
+; X64-NEXT: mulq %rbp
; X64-NEXT: addq %rcx, %rax
; X64-NEXT: adcq %rsi, %rdx
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
+; X64-NEXT: addq %r12, %rsi
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: adcq %r14, %rcx
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
; X64-NEXT: addq %rax, %rsi
; X64-NEXT: adcq %rdx, %rcx
-; X64-NEXT: addq %r13, %rdi
-; X64-NEXT: movq %rdi, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r11, %rbp
-; X64-NEXT: movq %rbp, {{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r8, %rsi
+; X64-NEXT: addq %r15, %r10
+; X64-NEXT: movq %r10, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %r8, %rbx
+; X64-NEXT: movq %rbx, {{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq %r9, %rsi
; X64-NEXT: movq %rsi, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq %r9, %rcx
+; X64-NEXT: adcq %r13, %rcx
; X64-NEXT: movq %rcx, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbp # 8-byte Reload
-; X64-NEXT: movq 96(%rbp), %rcx
-; X64-NEXT: imulq %rcx, %r10
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq 96(%rsi), %rcx
+; X64-NEXT: imulq %rcx, %rbp
; X64-NEXT: movq %rcx, %rax
-; X64-NEXT: mulq %r15
+; X64-NEXT: movq %r11, %rdi
+; X64-NEXT: mulq %rdi
; X64-NEXT: movq %rax, %r9
-; X64-NEXT: addq %r10, %rdx
-; X64-NEXT: movq 104(%rbp), %r8
-; X64-NEXT: imulq %r8, %r15
-; X64-NEXT: addq %rdx, %r15
-; X64-NEXT: movq 112(%rbp), %rax
-; X64-NEXT: movq %rbp, %rdi
-; X64-NEXT: movq %rax, %rsi
+; X64-NEXT: addq %rbp, %rdx
+; X64-NEXT: movq 104(%rsi), %r8
+; X64-NEXT: imulq %r8, %rdi
+; X64-NEXT: addq %rdx, %rdi
+; X64-NEXT: movq %rdi, %r10
+; X64-NEXT: movq 112(%rsi), %rax
+; X64-NEXT: movq %rsi, %rbp
+; X64-NEXT: movq %rax, %rdi
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbx # 8-byte Reload
-; X64-NEXT: imulq %rbx, %rsi
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbp # 8-byte Reload
-; X64-NEXT: mulq %rbp
-; X64-NEXT: movq %rax, %r13
-; X64-NEXT: addq %rsi, %rdx
-; X64-NEXT: movq 120(%rdi), %rdi
-; X64-NEXT: imulq %rbp, %rdi
+; X64-NEXT: imulq %rbx, %rdi
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: mulq %rsi
+; X64-NEXT: movq %rax, %r11
+; X64-NEXT: addq %rdi, %rdx
+; X64-NEXT: movq 120(%rbp), %rdi
+; X64-NEXT: imulq %rsi, %rdi
; X64-NEXT: addq %rdx, %rdi
-; X64-NEXT: addq %r9, %r13
-; X64-NEXT: adcq %r15, %rdi
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: movq %rbp, %r9
+; X64-NEXT: addq %r9, %r11
+; X64-NEXT: adcq %r10, %rdi
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: movq %rsi, %r10
; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq %rbx, %rax
+; X64-NEXT: movq %rbx, %r9
; X64-NEXT: mulq %rcx
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq %rax, %rsi
-; X64-NEXT: addq %rbp, %rsi
+; X64-NEXT: movq %rax, %rbx
+; X64-NEXT: addq %rsi, %rbx
; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: movq %r9, %rax
+; X64-NEXT: movq %r10, %rax
; X64-NEXT: mulq %r8
-; X64-NEXT: movq %rdx, %rbp
-; X64-NEXT: movq %rax, %r12
-; X64-NEXT: addq %rsi, %r12
-; X64-NEXT: adcq $0, %rbp
-; X64-NEXT: addq %rcx, %rbp
+; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: movq %rax, %r15
+; X64-NEXT: addq %rbx, %r15
+; X64-NEXT: adcq %rcx, %rsi
; X64-NEXT: sbbq %rcx, %rcx
; X64-NEXT: andl $1, %ecx
-; X64-NEXT: movq %rbx, %rax
+; X64-NEXT: movq %r9, %rax
; X64-NEXT: mulq %r8
-; X64-NEXT: movq %rdx, %r8
-; X64-NEXT: movq %rax, %r9
-; X64-NEXT: addq %rbp, %r9
-; X64-NEXT: adcq %rcx, %r8
-; X64-NEXT: addq %r13, %r9
-; X64-NEXT: adcq %rdi, %r8
+; X64-NEXT: movq %rdx, %rbx
+; X64-NEXT: movq %rax, %r8
+; X64-NEXT: addq %rsi, %r8
+; X64-NEXT: adcq %rcx, %rbx
+; X64-NEXT: addq %r11, %r8
+; X64-NEXT: adcq %rdi, %rbx
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbx # 8-byte Reload
-; X64-NEXT: imulq %rbx, %rsi
-; X64-NEXT: movq %rbx, %rax
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rax # 8-byte Reload
+; X64-NEXT: imulq %rax, %rsi
+; X64-NEXT: movq %rax, %r9
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rax, %r10
+; X64-NEXT: movq %rax, %r11
; X64-NEXT: addq %rsi, %rdx
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r11 # 8-byte Reload
-; X64-NEXT: imulq %r11, %rcx
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %r12 # 8-byte Reload
+; X64-NEXT: imulq %r12, %rcx
; X64-NEXT: addq %rdx, %rcx
-; X64-NEXT: movq %rcx, %rsi
+; X64-NEXT: movq %rcx, %rbp
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
; X64-NEXT: movq %rax, %rcx
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %r14 # 8-byte Reload
-; X64-NEXT: imulq %r14, %rcx
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbp # 8-byte Reload
-; X64-NEXT: mulq %rbp
-; X64-NEXT: movq %rax, %r13
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: imulq %rsi, %rcx
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
+; X64-NEXT: mulq %rdi
+; X64-NEXT: movq %rax, %r10
; X64-NEXT: addq %rcx, %rdx
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
-; X64-NEXT: imulq %rbp, %rax
-; X64-NEXT: addq %rdx, %rax
-; X64-NEXT: addq %r10, %r13
-; X64-NEXT: adcq %rsi, %rax
-; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: movq %rbp, %r10
-; X64-NEXT: mulq %rbx
-; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: movq %r14, %rax
-; X64-NEXT: mulq %rbx
-; X64-NEXT: movq %rdx, %rdi
+; X64-NEXT: movq %r14, %r13
+; X64-NEXT: imulq %rdi, %r13
+; X64-NEXT: addq %rdx, %r13
+; X64-NEXT: addq %r11, %r10
+; X64-NEXT: adcq %rbp, %r13
+; X64-NEXT: movq %rdi, %rax
+; X64-NEXT: movq %rdi, %r11
+; X64-NEXT: mulq %r9
+; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: movq %rax, %rdi
+; X64-NEXT: movq %rsi, %rax
+; X64-NEXT: movq %rsi, %r14
+; X64-NEXT: mulq %r9
+; X64-NEXT: movq %rdx, %r9
; X64-NEXT: movq %rax, %rcx
-; X64-NEXT: addq %rsi, %rcx
-; X64-NEXT: adcq $0, %rdi
-; X64-NEXT: movq %r10, %rax
-; X64-NEXT: mulq %r11
-; X64-NEXT: movq %rdx, %r15
-; X64-NEXT: movq %rax, %r10
-; X64-NEXT: addq %rcx, %r10
-; X64-NEXT: adcq $0, %r15
-; X64-NEXT: addq %rdi, %r15
+; X64-NEXT: addq %rbp, %rcx
+; X64-NEXT: adcq $0, %r9
+; X64-NEXT: movq %r11, %rax
+; X64-NEXT: movq %r12, %rbp
+; X64-NEXT: mulq %rbp
+; X64-NEXT: movq %rdx, %rsi
+; X64-NEXT: movq %rax, %r11
+; X64-NEXT: addq %rcx, %r11
+; X64-NEXT: adcq %r9, %rsi
; X64-NEXT: sbbq %rcx, %rcx
; X64-NEXT: andl $1, %ecx
; X64-NEXT: movq %r14, %rax
-; X64-NEXT: mulq %r11
-; X64-NEXT: addq %r15, %rax
+; X64-NEXT: mulq %rbp
+; X64-NEXT: addq %rsi, %rax
; X64-NEXT: adcq %rcx, %rdx
-; X64-NEXT: addq %r13, %rax
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rdx # 8-byte Folded Reload
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rbp # 8-byte Folded Reload
-; X64-NEXT: adcq %r12, %r10
-; X64-NEXT: adcq %r9, %rax
-; X64-NEXT: adcq %r8, %rdx
-; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rbp # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r10 # 8-byte Folded Reload
+; X64-NEXT: addq %r10, %rax
+; X64-NEXT: adcq %r13, %rdx
+; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
+; X64-NEXT: adcq %r15, %r11
+; X64-NEXT: adcq %r8, %rax
+; X64-NEXT: adcq %rbx, %rdx
+; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rax # 8-byte Folded Reload
; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rdx # 8-byte Folded Reload
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: addq {{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
-; X64-NEXT: movq (%rsp), %rbx # 8-byte Reload
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbx # 8-byte Reload
; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rbp # 8-byte Folded Reload
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r10 # 8-byte Folded Reload
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rbp # 8-byte Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rbp # 8-byte Folded Reload
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rax # 8-byte Folded Reload
; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rdx # 8-byte Folded Reload
-; X64-NEXT: addq {{[0-9]+}}(%rsp), %rcx # 8-byte Folded Reload
+; X64-NEXT: addq (%rsp), %rcx # 8-byte Folded Reload
; X64-NEXT: movq %rcx, %r8
-; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
-; X64-NEXT: movq %rdi, %r9
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
+; X64-NEXT: movq %rsi, %r9
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rbx # 8-byte Folded Reload
; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rbp # 8-byte Folded Reload
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r10 # 8-byte Folded Reload
+; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
+; X64-NEXT: adcq {{[0-9]+}}(%rsp), %r11 # 8-byte Folded Reload
; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rax # 8-byte Folded Reload
; X64-NEXT: adcq {{[0-9]+}}(%rsp), %rdx # 8-byte Folded Reload
; X64-NEXT: movq {{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: movq %rdi, (%rcx)
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: movq %rdi, 8(%rcx)
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: movq %rdi, 16(%rcx)
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: movq %rdi, 24(%rcx)
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: movq %rdi, 32(%rcx)
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: movq %rdi, 40(%rcx)
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: movq %rdi, 48(%rcx)
-; X64-NEXT: movq {{[0-9]+}}(%rsp), %rdi # 8-byte Reload
-; X64-NEXT: movq %rdi, 56(%rcx)
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq %rsi, (%rcx)
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq %rsi, 8(%rcx)
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq %rsi, 16(%rcx)
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq %rsi, 24(%rcx)
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq %rsi, 32(%rcx)
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq %rsi, 40(%rcx)
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq %rsi, 48(%rcx)
+; X64-NEXT: movq {{[0-9]+}}(%rsp), %rsi # 8-byte Reload
+; X64-NEXT: movq %rsi, 56(%rcx)
; X64-NEXT: movq %r8, 64(%rcx)
; X64-NEXT: movq %r9, 72(%rcx)
; X64-NEXT: movq %rbx, 80(%rcx)
-; X64-NEXT: movq %rsi, 88(%rcx)
-; X64-NEXT: movq %rbp, 96(%rcx)
-; X64-NEXT: movq %r10, 104(%rcx)
+; X64-NEXT: movq %rbp, 88(%rcx)
+; X64-NEXT: movq %rdi, 96(%rcx)
+; X64-NEXT: movq %r11, 104(%rcx)
; X64-NEXT: movq %rax, 112(%rcx)
; X64-NEXT: movq %rdx, 120(%rcx)
-; X64-NEXT: addq $360, %rsp # imm = 0x168
+; X64-NEXT: addq $352, %rsp # imm = 0x160
; X64-NEXT: popq %rbx
; X64-NEXT: popq %r12
; X64-NEXT: popq %r13
Modified: llvm/trunk/test/CodeGen/X86/mul-i256.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/mul-i256.ll?rev=301922&r1=301921&r2=301922&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/mul-i256.ll (original)
+++ llvm/trunk/test/CodeGen/X86/mul-i256.ll Tue May 2 09:15:48 2017
@@ -219,7 +219,7 @@ define void @test(i256* %a, i256* %b, i2
; X64-NEXT: .cfi_offset %r14, -24
; X64-NEXT: .Lcfi7:
; X64-NEXT: .cfi_offset %r15, -16
-; X64-NEXT: movq %rdx, %r10
+; X64-NEXT: movq %rdx, %r9
; X64-NEXT: movq (%rdi), %r14
; X64-NEXT: movq 8(%rdi), %r8
; X64-NEXT: movq 16(%rdi), %rcx
@@ -230,7 +230,7 @@ define void @test(i256* %a, i256* %b, i2
; X64-NEXT: imulq %r12, %rdi
; X64-NEXT: movq %r12, %rax
; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rax, %r9
+; X64-NEXT: movq %rax, %r10
; X64-NEXT: addq %rdi, %rdx
; X64-NEXT: imulq %r15, %rcx
; X64-NEXT: addq %rdx, %rcx
@@ -243,12 +243,12 @@ define void @test(i256* %a, i256* %b, i2
; X64-NEXT: movq 24(%rsi), %rbx
; X64-NEXT: imulq %r14, %rbx
; X64-NEXT: addq %rdx, %rbx
-; X64-NEXT: addq %r9, %r11
+; X64-NEXT: addq %r10, %r11
; X64-NEXT: adcq %rcx, %rbx
; X64-NEXT: movq %r14, %rax
; X64-NEXT: mulq %r12
; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: movq %rax, %r9
+; X64-NEXT: movq %rax, %r10
; X64-NEXT: movq %r8, %rax
; X64-NEXT: mulq %r12
; X64-NEXT: movq %rdx, %rcx
@@ -260,8 +260,7 @@ define void @test(i256* %a, i256* %b, i2
; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq %rax, %r14
; X64-NEXT: addq %rdi, %r14
-; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: addq %rcx, %rsi
+; X64-NEXT: adcq %rcx, %rsi
; X64-NEXT: sbbq %rcx, %rcx
; X64-NEXT: andl $1, %ecx
; X64-NEXT: movq %r8, %rax
@@ -270,10 +269,10 @@ define void @test(i256* %a, i256* %b, i2
; X64-NEXT: adcq %rcx, %rdx
; X64-NEXT: addq %r11, %rax
; X64-NEXT: adcq %rbx, %rdx
-; X64-NEXT: movq %r9, (%r10)
-; X64-NEXT: movq %r14, 8(%r10)
-; X64-NEXT: movq %rax, 16(%r10)
-; X64-NEXT: movq %rdx, 24(%r10)
+; X64-NEXT: movq %r10, (%r9)
+; X64-NEXT: movq %r14, 8(%r9)
+; X64-NEXT: movq %rax, 16(%r9)
+; X64-NEXT: movq %rdx, 24(%r9)
; X64-NEXT: popq %rbx
; X64-NEXT: popq %r12
; X64-NEXT: popq %r14
Modified: llvm/trunk/test/CodeGen/X86/mul-i512.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/mul-i512.ll?rev=301922&r1=301921&r2=301922&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/mul-i512.ll (original)
+++ llvm/trunk/test/CodeGen/X86/mul-i512.ll Tue May 2 09:15:48 2017
@@ -911,57 +911,57 @@ define void @test_512(i512* %a, i512* %b
; X64-NEXT: pushq %rbx
; X64-NEXT: pushq %rax
; X64-NEXT: movq %rdx, (%rsp) # 8-byte Spill
-; X64-NEXT: movq 24(%rdi), %rbp
-; X64-NEXT: movq 16(%rdi), %r11
+; X64-NEXT: movq 24(%rdi), %r11
+; X64-NEXT: movq 16(%rdi), %r14
; X64-NEXT: movq %rsi, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq (%rsi), %rdx
-; X64-NEXT: movq 8(%rsi), %r8
-; X64-NEXT: movq %r11, %rax
+; X64-NEXT: movq 8(%rsi), %rbp
+; X64-NEXT: movq %r14, %rax
; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: mulq %rsi
-; X64-NEXT: movq %rdx, %r10
+; X64-NEXT: movq %rdx, %r8
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: movq %rbp, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %r11, %rax
+; X64-NEXT: movq %r11, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: mulq %rsi
-; X64-NEXT: movq %rsi, %r9
+; X64-NEXT: movq %rsi, %r10
; X64-NEXT: movq %rdx, %rbx
; X64-NEXT: movq %rax, %rsi
-; X64-NEXT: addq %r10, %rsi
+; X64-NEXT: addq %r8, %rsi
; X64-NEXT: adcq $0, %rbx
-; X64-NEXT: movq %r11, %rax
-; X64-NEXT: movq %r11, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: mulq %r8
+; X64-NEXT: movq %r14, %rax
+; X64-NEXT: movq %r14, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: mulq %rbp
; X64-NEXT: movq %rdx, %rcx
-; X64-NEXT: movq %rax, %r12
-; X64-NEXT: addq %rsi, %r12
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rbx, %rcx
+; X64-NEXT: movq %rax, %r9
+; X64-NEXT: addq %rsi, %r9
+; X64-NEXT: adcq %rbx, %rcx
; X64-NEXT: sbbq %rbx, %rbx
; X64-NEXT: andl $1, %ebx
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: mulq %r8
+; X64-NEXT: movq %r11, %rax
+; X64-NEXT: mulq %rbp
+; X64-NEXT: movq %rbp, %r8
; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq %rax, %rbp
; X64-NEXT: addq %rcx, %rbp
; X64-NEXT: adcq %rbx, %rsi
; X64-NEXT: xorl %ecx, %ecx
-; X64-NEXT: movq %r9, %rbx
+; X64-NEXT: movq %r10, %rbx
; X64-NEXT: movq %rbx, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq %rbx, %rax
; X64-NEXT: mulq %rcx
; X64-NEXT: movq %rdx, %r13
; X64-NEXT: movq %rax, %r10
-; X64-NEXT: movq %r11, %rax
+; X64-NEXT: movq %r14, %rax
; X64-NEXT: mulq %rcx
-; X64-NEXT: movq %rdx, %r9
-; X64-NEXT: movq %r9, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: movq %rdx, %r12
+; X64-NEXT: movq %r12, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq %rax, %r15
; X64-NEXT: movq %r15, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: addq %r10, %r15
-; X64-NEXT: adcq %r13, %r9
+; X64-NEXT: adcq %r13, %r12
; X64-NEXT: addq %rbp, %r15
-; X64-NEXT: adcq %rsi, %r9
+; X64-NEXT: adcq %rsi, %r12
; X64-NEXT: movq %rdi, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq (%rdi), %r14
; X64-NEXT: movq %r14, %rax
@@ -982,8 +982,7 @@ define void @test_512(i512* %a, i512* %b
; X64-NEXT: movq %rdx, %rbx
; X64-NEXT: addq %rsi, %rax
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %rbx
-; X64-NEXT: addq %rbp, %rbx
+; X64-NEXT: adcq %rbp, %rbx
; X64-NEXT: sbbq %rdi, %rdi
; X64-NEXT: andl $1, %edi
; X64-NEXT: movq %rcx, %rax
@@ -1003,10 +1002,9 @@ define void @test_512(i512* %a, i512* %b
; X64-NEXT: addq %rbp, %r10
; X64-NEXT: adcq %rsi, %r13
; X64-NEXT: addq -{{[0-9]+}}(%rsp), %r10 # 8-byte Folded Reload
-; X64-NEXT: adcq %r12, %r13
+; X64-NEXT: adcq %r9, %r13
; X64-NEXT: adcq $0, %r15
-; X64-NEXT: adcq $0, %r9
-; X64-NEXT: movq %r9, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: adcq $0, %r12
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rsi # 8-byte Reload
; X64-NEXT: movq 16(%rsi), %r8
; X64-NEXT: movq %rcx, %rax
@@ -1014,22 +1012,21 @@ define void @test_512(i512* %a, i512* %b
; X64-NEXT: movq %r9, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: mulq %r8
; X64-NEXT: movq %rdx, %rdi
-; X64-NEXT: movq %rax, %r12
+; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
; X64-NEXT: movq %rcx, %rax
; X64-NEXT: mulq %r8
-; X64-NEXT: movq %rdx, %rbx
-; X64-NEXT: movq %rax, %rbp
-; X64-NEXT: addq %rdi, %rbp
-; X64-NEXT: adcq $0, %rbx
+; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: movq %rax, %rbx
+; X64-NEXT: addq %rdi, %rbx
+; X64-NEXT: adcq $0, %rbp
; X64-NEXT: movq 24(%rsi), %rdi
; X64-NEXT: movq %r9, %rax
; X64-NEXT: mulq %rdi
; X64-NEXT: movq %rdx, %rsi
-; X64-NEXT: addq %rbp, %rax
+; X64-NEXT: addq %rbx, %rax
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: addq %rbx, %rsi
+; X64-NEXT: adcq %rbp, %rsi
; X64-NEXT: sbbq %rbp, %rbp
; X64-NEXT: andl $1, %ebp
; X64-NEXT: movq %rcx, %rax
@@ -1047,13 +1044,12 @@ define void @test_512(i512* %a, i512* %b
; X64-NEXT: adcq %rdx, %r14
; X64-NEXT: addq %r9, %r11
; X64-NEXT: adcq %rbx, %r14
-; X64-NEXT: addq %r10, %r12
-; X64-NEXT: movq %r12, -{{[0-9]+}}(%rsp) # 8-byte Spill
+; X64-NEXT: addq %r10, -{{[0-9]+}}(%rsp) # 8-byte Folded Spill
; X64-NEXT: adcq %r13, -{{[0-9]+}}(%rsp) # 8-byte Folded Spill
; X64-NEXT: adcq $0, %r11
; X64-NEXT: adcq $0, %r14
; X64-NEXT: addq %r15, %r11
-; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %r14 # 8-byte Folded Reload
+; X64-NEXT: adcq %r12, %r14
; X64-NEXT: adcq $0, %rcx
; X64-NEXT: movq %rcx, %r13
; X64-NEXT: sbbq %r9, %r9
@@ -1075,8 +1071,7 @@ define void @test_512(i512* %a, i512* %b
; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: addq %rbx, %rax
; X64-NEXT: movq %rax, %rbx
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rsi, %rcx
+; X64-NEXT: adcq %rsi, %rcx
; X64-NEXT: sbbq %rsi, %rsi
; X64-NEXT: andl $1, %esi
; X64-NEXT: movq %r10, %rax
@@ -1102,7 +1097,7 @@ define void @test_512(i512* %a, i512* %b
; X64-NEXT: imulq %rsi, %rdi
; X64-NEXT: movq %rsi, %rax
; X64-NEXT: mulq %r8
-; X64-NEXT: movq %rax, %r11
+; X64-NEXT: movq %rax, %r10
; X64-NEXT: addq %rdi, %rdx
; X64-NEXT: movq 40(%rcx), %r9
; X64-NEXT: imulq %r9, %r8
@@ -1110,60 +1105,58 @@ define void @test_512(i512* %a, i512* %b
; X64-NEXT: movq 48(%rcx), %rax
; X64-NEXT: movq %rcx, %rbx
; X64-NEXT: movq %rax, %rdi
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r11 # 8-byte Reload
+; X64-NEXT: imulq %r11, %rdi
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rcx # 8-byte Reload
-; X64-NEXT: imulq %rcx, %rdi
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rbp # 8-byte Reload
-; X64-NEXT: mulq %rbp
+; X64-NEXT: mulq %rcx
; X64-NEXT: movq %rax, %r12
; X64-NEXT: addq %rdi, %rdx
; X64-NEXT: movq 56(%rbx), %rbx
-; X64-NEXT: imulq %rbp, %rbx
+; X64-NEXT: imulq %rcx, %rbx
; X64-NEXT: addq %rdx, %rbx
-; X64-NEXT: addq %r11, %r12
+; X64-NEXT: addq %r10, %r12
; X64-NEXT: adcq %r8, %rbx
-; X64-NEXT: movq %rbp, %rax
-; X64-NEXT: movq %rbp, %r8
+; X64-NEXT: movq %rcx, %rax
; X64-NEXT: mulq %rsi
; X64-NEXT: movq %rdx, %rdi
; X64-NEXT: movq %rax, -{{[0-9]+}}(%rsp) # 8-byte Spill
-; X64-NEXT: movq %rcx, %rax
+; X64-NEXT: movq %r11, %rax
; X64-NEXT: mulq %rsi
; X64-NEXT: movq %rdx, %rsi
; X64-NEXT: movq %rax, %rbp
; X64-NEXT: addq %rdi, %rbp
; X64-NEXT: adcq $0, %rsi
-; X64-NEXT: movq %r8, %rax
+; X64-NEXT: movq %rcx, %rax
; X64-NEXT: mulq %r9
; X64-NEXT: movq %rdx, %rdi
-; X64-NEXT: movq %rax, %r11
-; X64-NEXT: addq %rbp, %r11
-; X64-NEXT: adcq $0, %rdi
-; X64-NEXT: addq %rsi, %rdi
+; X64-NEXT: movq %rax, %r15
+; X64-NEXT: addq %rbp, %r15
+; X64-NEXT: adcq %rsi, %rdi
; X64-NEXT: sbbq %rsi, %rsi
; X64-NEXT: andl $1, %esi
-; X64-NEXT: movq %rcx, %rax
+; X64-NEXT: movq %r11, %rax
; X64-NEXT: mulq %r9
-; X64-NEXT: movq %rdx, %r14
-; X64-NEXT: movq %rax, %r15
-; X64-NEXT: addq %rdi, %r15
-; X64-NEXT: adcq %rsi, %r14
-; X64-NEXT: addq %r12, %r15
-; X64-NEXT: adcq %rbx, %r14
+; X64-NEXT: movq %rdx, %r11
+; X64-NEXT: movq %rax, %r14
+; X64-NEXT: addq %rdi, %r14
+; X64-NEXT: adcq %rsi, %r11
+; X64-NEXT: addq %r12, %r14
+; X64-NEXT: adcq %rbx, %r11
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rdx # 8-byte Reload
; X64-NEXT: movq 56(%rdx), %rcx
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r8 # 8-byte Reload
-; X64-NEXT: imulq %r8, %rcx
-; X64-NEXT: movq 48(%rdx), %rbp
-; X64-NEXT: movq %rdx, %rbx
-; X64-NEXT: movq %r8, %rax
-; X64-NEXT: mulq %rbp
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r10 # 8-byte Reload
+; X64-NEXT: imulq %r10, %rcx
+; X64-NEXT: movq 48(%rdx), %rbx
+; X64-NEXT: movq %rdx, %rbp
+; X64-NEXT: movq %r10, %rax
+; X64-NEXT: mulq %rbx
; X64-NEXT: movq %rax, %rsi
; X64-NEXT: addq %rcx, %rdx
-; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r10 # 8-byte Reload
-; X64-NEXT: imulq %r10, %rbp
-; X64-NEXT: addq %rdx, %rbp
-; X64-NEXT: movq 32(%rbx), %rdi
-; X64-NEXT: movq 40(%rbx), %r12
+; X64-NEXT: movq -{{[0-9]+}}(%rsp), %r8 # 8-byte Reload
+; X64-NEXT: imulq %r8, %rbx
+; X64-NEXT: addq %rdx, %rbx
+; X64-NEXT: movq 32(%rbp), %rdi
+; X64-NEXT: movq 40(%rbp), %r12
; X64-NEXT: movq -{{[0-9]+}}(%rsp), %rax # 8-byte Reload
; X64-NEXT: movq %rax, %rcx
; X64-NEXT: imulq %r12, %rcx
@@ -1174,36 +1167,35 @@ define void @test_512(i512* %a, i512* %b
; X64-NEXT: imulq %rdi, %r13
; X64-NEXT: addq %rdx, %r13
; X64-NEXT: addq %rsi, %r9
-; X64-NEXT: adcq %rbp, %r13
+; X64-NEXT: adcq %rbx, %r13
; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: mulq %r8
+; X64-NEXT: mulq %r10
; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: movq %rax, %rsi
; X64-NEXT: movq %r12, %rax
-; X64-NEXT: mulq %r8
+; X64-NEXT: mulq %r10
; X64-NEXT: movq %rdx, %rbx
; X64-NEXT: movq %rax, %rbp
; X64-NEXT: addq %rcx, %rbp
; X64-NEXT: adcq $0, %rbx
; X64-NEXT: movq %rdi, %rax
-; X64-NEXT: mulq %r10
+; X64-NEXT: mulq %r8
; X64-NEXT: movq %rdx, %rcx
; X64-NEXT: movq %rax, %rdi
; X64-NEXT: addq %rbp, %rdi
-; X64-NEXT: adcq $0, %rcx
-; X64-NEXT: addq %rbx, %rcx
-; X64-NEXT: sbbq %rbp, %rbp
-; X64-NEXT: andl $1, %ebp
+; X64-NEXT: adcq %rbx, %rcx
+; X64-NEXT: sbbq %rbx, %rbx
+; X64-NEXT: andl $1, %ebx
; X64-NEXT: movq %r12, %rax
-; X64-NEXT: mulq %r10
+; X64-NEXT: mulq %r8
; X64-NEXT: addq %rcx, %rax
-; X64-NEXT: adcq %rbp, %rdx
+; X64-NEXT: adcq %rbx, %rdx
; X64-NEXT: addq %r9, %rax
; X64-NEXT: adcq %r13, %rdx
; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
-; X64-NEXT: adcq %r11, %rdi
-; X64-NEXT: adcq %r15, %rax
-; X64-NEXT: adcq %r14, %rdx
+; X64-NEXT: adcq %r15, %rdi
+; X64-NEXT: adcq %r14, %rax
+; X64-NEXT: adcq %r11, %rdx
; X64-NEXT: addq -{{[0-9]+}}(%rsp), %rsi # 8-byte Folded Reload
; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rdi # 8-byte Folded Reload
; X64-NEXT: adcq -{{[0-9]+}}(%rsp), %rax # 8-byte Folded Reload
More information about the llvm-commits
mailing list