summaryrefslogtreecommitdiff
path: root/llvm/test/CodeGen/X86
diff options
context:
space:
mode:
Diffstat (limited to 'llvm/test/CodeGen/X86')
-rw-r--r--llvm/test/CodeGen/X86/AMX/amx-ldtilecfg-insert.ll18
-rw-r--r--llvm/test/CodeGen/X86/i128-mul.ll178
-rw-r--r--llvm/test/CodeGen/X86/loop-strength-reduce5.ll10
-rw-r--r--llvm/test/CodeGen/X86/madd.ll22
-rw-r--r--llvm/test/CodeGen/X86/pr49451.ll6
-rw-r--r--llvm/test/CodeGen/X86/wide-scalar-shift-by-byte-multiple-legalization.ll114
-rw-r--r--llvm/test/CodeGen/X86/x86-shrink-wrapping.ll12
-rw-r--r--llvm/test/CodeGen/X86/xor.ll132
8 files changed, 248 insertions, 244 deletions
diff --git a/llvm/test/CodeGen/X86/AMX/amx-ldtilecfg-insert.ll b/llvm/test/CodeGen/X86/AMX/amx-ldtilecfg-insert.ll
index 06cf968512db..8a8e7a3b4df2 100644
--- a/llvm/test/CodeGen/X86/AMX/amx-ldtilecfg-insert.ll
+++ b/llvm/test/CodeGen/X86/AMX/amx-ldtilecfg-insert.ll
@@ -297,30 +297,30 @@ define dso_local void @test6(i16 signext %0) nounwind {
; CHECK-NEXT: movb $1, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: movb %dil, -{{[0-9]+}}(%rsp)
; CHECK-NEXT: xorl %eax, %eax
-; CHECK-NEXT: movl $buf, %ecx
-; CHECK-NEXT: movl $32, %edx
-; CHECK-NEXT: xorl %esi, %esi
+; CHECK-NEXT: xorl %ecx, %ecx
+; CHECK-NEXT: movl $buf, %edx
+; CHECK-NEXT: movl $32, %esi
; CHECK-NEXT: jmp .LBB5_1
; CHECK-NEXT: .p2align 4
; CHECK-NEXT: .LBB5_3: # %if.false
; CHECK-NEXT: # in Loop: Header=BB5_1 Depth=1
-; CHECK-NEXT: decl %esi
+; CHECK-NEXT: decl %eax
; CHECK-NEXT: .LBB5_4: # %loop.bb2
; CHECK-NEXT: # in Loop: Header=BB5_1 Depth=1
-; CHECK-NEXT: leal (%rdi,%rsi), %r8d
+; CHECK-NEXT: leal (%rdi,%rax), %r8d
; CHECK-NEXT: movw %r8w, -{{[0-9]+}}(%rsp)
-; CHECK-NEXT: cmpw $7, %si
+; CHECK-NEXT: cmpw $7, %ax
; CHECK-NEXT: ldtilecfg -{{[0-9]+}}(%rsp)
; CHECK-NEXT: tilezero %tmm0
-; CHECK-NEXT: tilestored %tmm0, (%rcx,%rdx)
+; CHECK-NEXT: tilestored %tmm0, (%rdx,%rsi)
; CHECK-NEXT: jne .LBB5_5
; CHECK-NEXT: .LBB5_1: # %loop.bb1
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT: testb %al, %al
+; CHECK-NEXT: testb %cl, %cl
; CHECK-NEXT: jne .LBB5_3
; CHECK-NEXT: # %bb.2: # %if.true
; CHECK-NEXT: # in Loop: Header=BB5_1 Depth=1
-; CHECK-NEXT: incl %esi
+; CHECK-NEXT: incl %eax
; CHECK-NEXT: jmp .LBB5_4
; CHECK-NEXT: .LBB5_5: # %exit
; CHECK-NEXT: tilerelease
diff --git a/llvm/test/CodeGen/X86/i128-mul.ll b/llvm/test/CodeGen/X86/i128-mul.ll
index 477a0dce5c81..cffd88c55bb0 100644
--- a/llvm/test/CodeGen/X86/i128-mul.ll
+++ b/llvm/test/CodeGen/X86/i128-mul.ll
@@ -111,63 +111,62 @@ define i64 @mul1(i64 %n, ptr nocapture %z, ptr nocapture %x, i64 %y) nounwind {
; X86-NOBMI-NEXT: orl %ecx, %eax
; X86-NOBMI-NEXT: je .LBB1_3
; X86-NOBMI-NEXT: # %bb.1: # %for.body.preheader
-; X86-NOBMI-NEXT: xorl %esi, %esi
+; X86-NOBMI-NEXT: xorl %eax, %eax
+; X86-NOBMI-NEXT: xorl %edx, %edx
; X86-NOBMI-NEXT: xorl %ecx, %ecx
-; X86-NOBMI-NEXT: xorl %edi, %edi
-; X86-NOBMI-NEXT: xorl %ebp, %ebp
+; X86-NOBMI-NEXT: movl $0, (%esp) # 4-byte Folded Spill
; X86-NOBMI-NEXT: .p2align 4
; X86-NOBMI-NEXT: .LBB1_2: # %for.body
; X86-NOBMI-NEXT: # =>This Inner Loop Header: Depth=1
-; X86-NOBMI-NEXT: movl %ebp, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NOBMI-NEXT: movl %edx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NOBMI-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
; X86-NOBMI-NEXT: movl {{[0-9]+}}(%esp), %eax
-; X86-NOBMI-NEXT: movl (%eax,%edi,8), %ebp
-; X86-NOBMI-NEXT: movl 4(%eax,%edi,8), %ebx
+; X86-NOBMI-NEXT: movl (%eax,%ecx,8), %edi
+; X86-NOBMI-NEXT: movl 4(%eax,%ecx,8), %ebx
; X86-NOBMI-NEXT: movl %ebx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
-; X86-NOBMI-NEXT: movl %ebp, %eax
-; X86-NOBMI-NEXT: mull {{[0-9]+}}(%esp)
-; X86-NOBMI-NEXT: movl %edx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-NOBMI-NEXT: movl %edi, %eax
+; X86-NOBMI-NEXT: movl {{[0-9]+}}(%esp), %esi
+; X86-NOBMI-NEXT: mull %esi
+; X86-NOBMI-NEXT: movl %edx, %ebp
; X86-NOBMI-NEXT: movl %eax, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
; X86-NOBMI-NEXT: movl %ebx, %eax
-; X86-NOBMI-NEXT: mull {{[0-9]+}}(%esp)
-; X86-NOBMI-NEXT: movl %eax, %ebx
-; X86-NOBMI-NEXT: addl {{[-0-9]+}}(%e{{[sb]}}p), %ebx # 4-byte Folded Reload
-; X86-NOBMI-NEXT: adcl $0, %edx
-; X86-NOBMI-NEXT: movl %edx, (%esp) # 4-byte Spill
-; X86-NOBMI-NEXT: movl %ebp, %eax
+; X86-NOBMI-NEXT: mull %esi
+; X86-NOBMI-NEXT: movl %edx, %ebx
+; X86-NOBMI-NEXT: movl %eax, %esi
+; X86-NOBMI-NEXT: addl %ebp, %esi
+; X86-NOBMI-NEXT: adcl $0, %ebx
+; X86-NOBMI-NEXT: movl %edi, %eax
; X86-NOBMI-NEXT: movl {{[0-9]+}}(%esp), %edx
; X86-NOBMI-NEXT: mull %edx
-; X86-NOBMI-NEXT: movl %eax, %ebp
-; X86-NOBMI-NEXT: addl %ebx, %ebp
-; X86-NOBMI-NEXT: movl %esi, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
-; X86-NOBMI-NEXT: adcl (%esp), %edx # 4-byte Folded Reload
-; X86-NOBMI-NEXT: movl %edx, %ebx
-; X86-NOBMI-NEXT: setb (%esp) # 1-byte Folded Spill
+; X86-NOBMI-NEXT: movl %edx, %ebp
+; X86-NOBMI-NEXT: movl %eax, %edi
+; X86-NOBMI-NEXT: addl %esi, %edi
+; X86-NOBMI-NEXT: adcl %ebx, %ebp
+; X86-NOBMI-NEXT: setb %bl
; X86-NOBMI-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %eax # 4-byte Reload
; X86-NOBMI-NEXT: mull {{[0-9]+}}(%esp)
-; X86-NOBMI-NEXT: movl %eax, %esi
-; X86-NOBMI-NEXT: addl %ebx, %esi
-; X86-NOBMI-NEXT: movl %ecx, %eax
-; X86-NOBMI-NEXT: movzbl (%esp), %ebx # 1-byte Folded Reload
-; X86-NOBMI-NEXT: movl %edx, %ecx
-; X86-NOBMI-NEXT: adcl %ebx, %ecx
-; X86-NOBMI-NEXT: movl {{[0-9]+}}(%esp), %ebx
-; X86-NOBMI-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %edx # 4-byte Reload
-; X86-NOBMI-NEXT: addl {{[-0-9]+}}(%e{{[sb]}}p), %edx # 4-byte Folded Reload
-; X86-NOBMI-NEXT: adcl %eax, %ebp
-; X86-NOBMI-NEXT: adcl $0, %esi
-; X86-NOBMI-NEXT: adcl $0, %ecx
-; X86-NOBMI-NEXT: movl {{[0-9]+}}(%esp), %eax
-; X86-NOBMI-NEXT: movl %edx, (%eax,%edi,8)
-; X86-NOBMI-NEXT: movl %ebp, 4(%eax,%edi,8)
-; X86-NOBMI-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ebp # 4-byte Reload
-; X86-NOBMI-NEXT: movl {{[0-9]+}}(%esp), %edx
-; X86-NOBMI-NEXT: addl $1, %edi
-; X86-NOBMI-NEXT: adcl $0, %ebp
-; X86-NOBMI-NEXT: movl %edi, %eax
-; X86-NOBMI-NEXT: xorl %edx, %eax
-; X86-NOBMI-NEXT: movl %ebp, %edx
-; X86-NOBMI-NEXT: xorl %ebx, %edx
-; X86-NOBMI-NEXT: orl %eax, %edx
+; X86-NOBMI-NEXT: addl %ebp, %eax
+; X86-NOBMI-NEXT: movzbl %bl, %esi
+; X86-NOBMI-NEXT: movl {{[0-9]+}}(%esp), %ebp
+; X86-NOBMI-NEXT: adcl %esi, %edx
+; X86-NOBMI-NEXT: movl %ecx, %ebx
+; X86-NOBMI-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload
+; X86-NOBMI-NEXT: addl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Folded Reload
+; X86-NOBMI-NEXT: adcl {{[-0-9]+}}(%e{{[sb]}}p), %edi # 4-byte Folded Reload
+; X86-NOBMI-NEXT: adcl $0, %eax
+; X86-NOBMI-NEXT: adcl $0, %edx
+; X86-NOBMI-NEXT: movl {{[0-9]+}}(%esp), %esi
+; X86-NOBMI-NEXT: movl %ecx, (%esi,%ebx,8)
+; X86-NOBMI-NEXT: movl %ebx, %ecx
+; X86-NOBMI-NEXT: movl %edi, 4(%esi,%ebx,8)
+; X86-NOBMI-NEXT: addl $1, %ecx
+; X86-NOBMI-NEXT: movl (%esp), %edi # 4-byte Reload
+; X86-NOBMI-NEXT: adcl $0, %edi
+; X86-NOBMI-NEXT: movl %ecx, %esi
+; X86-NOBMI-NEXT: xorl {{[0-9]+}}(%esp), %esi
+; X86-NOBMI-NEXT: movl %edi, (%esp) # 4-byte Spill
+; X86-NOBMI-NEXT: xorl %ebp, %edi
+; X86-NOBMI-NEXT: orl %esi, %edi
; X86-NOBMI-NEXT: jne .LBB1_2
; X86-NOBMI-NEXT: .LBB1_3: # %for.end
; X86-NOBMI-NEXT: xorl %eax, %eax
@@ -185,66 +184,71 @@ define i64 @mul1(i64 %n, ptr nocapture %z, ptr nocapture %x, i64 %y) nounwind {
; X86-BMI-NEXT: pushl %ebx
; X86-BMI-NEXT: pushl %edi
; X86-BMI-NEXT: pushl %esi
-; X86-BMI-NEXT: subl $16, %esp
+; X86-BMI-NEXT: subl $20, %esp
; X86-BMI-NEXT: movl {{[0-9]+}}(%esp), %ecx
; X86-BMI-NEXT: movl {{[0-9]+}}(%esp), %eax
; X86-BMI-NEXT: orl %ecx, %eax
; X86-BMI-NEXT: je .LBB1_3
; X86-BMI-NEXT: # %bb.1: # %for.body.preheader
-; X86-BMI-NEXT: xorl %esi, %esi
-; X86-BMI-NEXT: xorl %edi, %edi
+; X86-BMI-NEXT: xorl %ecx, %ecx
+; X86-BMI-NEXT: xorl %eax, %eax
; X86-BMI-NEXT: xorl %ebx, %ebx
-; X86-BMI-NEXT: movl $0, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Folded Spill
+; X86-BMI-NEXT: xorl %ebp, %ebp
; X86-BMI-NEXT: .p2align 4
; X86-BMI-NEXT: .LBB1_2: # %for.body
; X86-BMI-NEXT: # =>This Inner Loop Header: Depth=1
+; X86-BMI-NEXT: movl %ecx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-BMI-NEXT: movl %eax, (%esp) # 4-byte Spill
+; X86-BMI-NEXT: movl %ebp, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
; X86-BMI-NEXT: movl {{[0-9]+}}(%esp), %eax
; X86-BMI-NEXT: movl (%eax,%ebx,8), %ecx
-; X86-BMI-NEXT: movl 4(%eax,%ebx,8), %ebp
-; X86-BMI-NEXT: movl %ebp, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
+; X86-BMI-NEXT: movl 4(%eax,%ebx,8), %esi
+; X86-BMI-NEXT: movl %esi, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
; X86-BMI-NEXT: movl %ecx, %edx
-; X86-BMI-NEXT: mulxl {{[0-9]+}}(%esp), %edx, %eax
-; X86-BMI-NEXT: movl %eax, (%esp) # 4-byte Spill
+; X86-BMI-NEXT: movl {{[0-9]+}}(%esp), %eax
+; X86-BMI-NEXT: mulxl %eax, %edx, %edi
; X86-BMI-NEXT: movl %edx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
-; X86-BMI-NEXT: movl %ebp, %edx
-; X86-BMI-NEXT: mulxl {{[0-9]+}}(%esp), %eax, %ebp
-; X86-BMI-NEXT: addl (%esp), %eax # 4-byte Folded Reload
-; X86-BMI-NEXT: adcl $0, %ebp
+; X86-BMI-NEXT: movl %esi, %edx
+; X86-BMI-NEXT: mulxl %eax, %esi, %eax
+; X86-BMI-NEXT: addl %edi, %esi
+; X86-BMI-NEXT: adcl $0, %eax
; X86-BMI-NEXT: movl %ecx, %edx
-; X86-BMI-NEXT: mulxl {{[0-9]+}}(%esp), %ecx, %edx
-; X86-BMI-NEXT: addl %eax, %ecx
-; X86-BMI-NEXT: movl %edi, (%esp) # 4-byte Spill
-; X86-BMI-NEXT: movl %esi, %eax
-; X86-BMI-NEXT: adcl %ebp, %edx
-; X86-BMI-NEXT: movl %edx, %ebp
+; X86-BMI-NEXT: movl {{[0-9]+}}(%esp), %ecx
+; X86-BMI-NEXT: mulxl %ecx, %edi, %ebp
+; X86-BMI-NEXT: addl %esi, %edi
+; X86-BMI-NEXT: adcl %eax, %ebp
; X86-BMI-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %edx # 4-byte Reload
-; X86-BMI-NEXT: mulxl {{[0-9]+}}(%esp), %esi, %edi
+; X86-BMI-NEXT: mulxl %ecx, %ecx, %eax
; X86-BMI-NEXT: setb %dl
-; X86-BMI-NEXT: addl %ebp, %esi
+; X86-BMI-NEXT: addl %ebp, %ecx
+; X86-BMI-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ebp # 4-byte Reload
+; X86-BMI-NEXT: movl {{[0-9]+}}(%esp), %esi
; X86-BMI-NEXT: movzbl %dl, %edx
-; X86-BMI-NEXT: adcl %edx, %edi
-; X86-BMI-NEXT: movl {{[0-9]+}}(%esp), %ebp
-; X86-BMI-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %edx # 4-byte Reload
-; X86-BMI-NEXT: addl %eax, %edx
-; X86-BMI-NEXT: adcl (%esp), %ecx # 4-byte Folded Reload
-; X86-BMI-NEXT: adcl $0, %esi
-; X86-BMI-NEXT: adcl $0, %edi
-; X86-BMI-NEXT: movl {{[0-9]+}}(%esp), %eax
-; X86-BMI-NEXT: movl %edx, (%eax,%ebx,8)
-; X86-BMI-NEXT: movl %ecx, 4(%eax,%ebx,8)
-; X86-BMI-NEXT: addl $1, %ebx
-; X86-BMI-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %ecx # 4-byte Reload
+; X86-BMI-NEXT: adcl %edx, %eax
+; X86-BMI-NEXT: movl %eax, %edx
+; X86-BMI-NEXT: movl {{[-0-9]+}}(%e{{[sb]}}p), %eax # 4-byte Reload
+; X86-BMI-NEXT: addl {{[-0-9]+}}(%e{{[sb]}}p), %eax # 4-byte Folded Reload
+; X86-BMI-NEXT: adcl (%esp), %edi # 4-byte Folded Reload
; X86-BMI-NEXT: adcl $0, %ecx
-; X86-BMI-NEXT: movl %ebx, %eax
-; X86-BMI-NEXT: xorl {{[0-9]+}}(%esp), %eax
-; X86-BMI-NEXT: movl %ecx, {{[-0-9]+}}(%e{{[sb]}}p) # 4-byte Spill
-; X86-BMI-NEXT: xorl %ebp, %ecx
-; X86-BMI-NEXT: orl %eax, %ecx
+; X86-BMI-NEXT: adcl $0, %edx
+; X86-BMI-NEXT: movl %edx, (%esp) # 4-byte Spill
+; X86-BMI-NEXT: movl {{[0-9]+}}(%esp), %edx
+; X86-BMI-NEXT: movl %eax, (%edx,%ebx,8)
+; X86-BMI-NEXT: movl %edi, 4(%edx,%ebx,8)
+; X86-BMI-NEXT: movl {{[0-9]+}}(%esp), %edi
+; X86-BMI-NEXT: addl $1, %ebx
+; X86-BMI-NEXT: adcl $0, %ebp
+; X86-BMI-NEXT: movl %ebx, %edx
+; X86-BMI-NEXT: xorl %esi, %edx
+; X86-BMI-NEXT: movl %ebp, %esi
+; X86-BMI-NEXT: xorl %edi, %esi
+; X86-BMI-NEXT: orl %edx, %esi
+; X86-BMI-NEXT: movl (%esp), %eax # 4-byte Reload
; X86-BMI-NEXT: jne .LBB1_2
; X86-BMI-NEXT: .LBB1_3: # %for.end
; X86-BMI-NEXT: xorl %eax, %eax
; X86-BMI-NEXT: xorl %edx, %edx
-; X86-BMI-NEXT: addl $16, %esp
+; X86-BMI-NEXT: addl $20, %esp
; X86-BMI-NEXT: popl %esi
; X86-BMI-NEXT: popl %edi
; X86-BMI-NEXT: popl %ebx
@@ -257,12 +261,11 @@ define i64 @mul1(i64 %n, ptr nocapture %z, ptr nocapture %x, i64 %y) nounwind {
; X64-NOBMI-NEXT: je .LBB1_3
; X64-NOBMI-NEXT: # %bb.1: # %for.body.preheader
; X64-NOBMI-NEXT: movq %rdx, %r8
-; X64-NOBMI-NEXT: xorl %edx, %edx
+; X64-NOBMI-NEXT: xorl %r10d, %r10d
; X64-NOBMI-NEXT: xorl %r9d, %r9d
; X64-NOBMI-NEXT: .p2align 4
; X64-NOBMI-NEXT: .LBB1_2: # %for.body
; X64-NOBMI-NEXT: # =>This Inner Loop Header: Depth=1
-; X64-NOBMI-NEXT: movq %rdx, %r10
; X64-NOBMI-NEXT: movq %rcx, %rax
; X64-NOBMI-NEXT: mulq (%r8,%r9,8)
; X64-NOBMI-NEXT: addq %r10, %rax
@@ -270,6 +273,7 @@ define i64 @mul1(i64 %n, ptr nocapture %z, ptr nocapture %x, i64 %y) nounwind {
; X64-NOBMI-NEXT: movq %rax, (%rsi,%r9,8)
; X64-NOBMI-NEXT: incq %r9
; X64-NOBMI-NEXT: cmpq %r9, %rdi
+; X64-NOBMI-NEXT: movq %rdx, %r10
; X64-NOBMI-NEXT: jne .LBB1_2
; X64-NOBMI-NEXT: .LBB1_3: # %for.end
; X64-NOBMI-NEXT: xorl %eax, %eax
@@ -281,12 +285,11 @@ define i64 @mul1(i64 %n, ptr nocapture %z, ptr nocapture %x, i64 %y) nounwind {
; X64-BMI-NEXT: je .LBB1_3
; X64-BMI-NEXT: # %bb.1: # %for.body.preheader
; X64-BMI-NEXT: movq %rdx, %rax
-; X64-BMI-NEXT: xorl %edx, %edx
+; X64-BMI-NEXT: xorl %r9d, %r9d
; X64-BMI-NEXT: xorl %r8d, %r8d
; X64-BMI-NEXT: .p2align 4
; X64-BMI-NEXT: .LBB1_2: # %for.body
; X64-BMI-NEXT: # =>This Inner Loop Header: Depth=1
-; X64-BMI-NEXT: movq %rdx, %r9
; X64-BMI-NEXT: movq %rcx, %rdx
; X64-BMI-NEXT: mulxq (%rax,%r8,8), %r10, %rdx
; X64-BMI-NEXT: addq %r9, %r10
@@ -294,6 +297,7 @@ define i64 @mul1(i64 %n, ptr nocapture %z, ptr nocapture %x, i64 %y) nounwind {
; X64-BMI-NEXT: movq %r10, (%rsi,%r8,8)
; X64-BMI-NEXT: incq %r8
; X64-BMI-NEXT: cmpq %r8, %rdi
+; X64-BMI-NEXT: movq %rdx, %r9
; X64-BMI-NEXT: jne .LBB1_2
; X64-BMI-NEXT: .LBB1_3: # %for.end
; X64-BMI-NEXT: xorl %eax, %eax
diff --git a/llvm/test/CodeGen/X86/loop-strength-reduce5.ll b/llvm/test/CodeGen/X86/loop-strength-reduce5.ll
index ebae51fa2aa4..08003739b55d 100644
--- a/llvm/test/CodeGen/X86/loop-strength-reduce5.ll
+++ b/llvm/test/CodeGen/X86/loop-strength-reduce5.ll
@@ -16,11 +16,11 @@ define void @foo(i32 %N) nounwind {
; CHECK-NEXT: .p2align 4
; CHECK-NEXT: .LBB0_2: # %bb
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
-; CHECK-NEXT: movw %dx, X
-; CHECK-NEXT: movw %cx, Y
-; CHECK-NEXT: incl %edx
-; CHECK-NEXT: addl $4, %ecx
-; CHECK-NEXT: cmpl %edx, %eax
+; CHECK-NEXT: movw %cx, X
+; CHECK-NEXT: movw %dx, Y
+; CHECK-NEXT: incl %ecx
+; CHECK-NEXT: addl $4, %edx
+; CHECK-NEXT: cmpl %ecx, %eax
; CHECK-NEXT: jne .LBB0_2
; CHECK-NEXT: .LBB0_3: # %return
; CHECK-NEXT: retl
diff --git a/llvm/test/CodeGen/X86/madd.ll b/llvm/test/CodeGen/X86/madd.ll
index 2a2a4a5ca18d..209ee79d5141 100644
--- a/llvm/test/CodeGen/X86/madd.ll
+++ b/llvm/test/CodeGen/X86/madd.ll
@@ -1480,15 +1480,15 @@ define i32 @test_unsigned_short_512(ptr nocapture readonly, ptr nocapture readon
; AVX2-NEXT: vpmovzxwd {{.*#+}} ymm3 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
; AVX2-NEXT: vpmovzxwd {{.*#+}} ymm4 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
; AVX2-NEXT: vpmulld %ymm2, %ymm4, %ymm2
-; AVX2-NEXT: vpaddd %ymm1, %ymm2, %ymm1
+; AVX2-NEXT: vpaddd %ymm0, %ymm2, %ymm0
; AVX2-NEXT: vpmovzxwd {{.*#+}} ymm2 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
; AVX2-NEXT: vpmulld %ymm3, %ymm2, %ymm2
-; AVX2-NEXT: vpaddd %ymm0, %ymm2, %ymm0
+; AVX2-NEXT: vpaddd %ymm1, %ymm2, %ymm1
; AVX2-NEXT: addq $16, %rcx
; AVX2-NEXT: cmpq %rcx, %rax
; AVX2-NEXT: jne .LBB10_1
; AVX2-NEXT: # %bb.2: # %middle.block
-; AVX2-NEXT: vpaddd %ymm1, %ymm0, %ymm0
+; AVX2-NEXT: vpaddd %ymm0, %ymm1, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpaddd %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
@@ -1728,10 +1728,10 @@ define i32 @test_unsigned_short_1024(ptr nocapture readonly, ptr nocapture reado
; AVX2-NEXT: vpaddd %ymm2, %ymm4, %ymm2
; AVX2-NEXT: vpmovzxwd {{.*#+}} ymm4 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
; AVX2-NEXT: vpmulld %ymm5, %ymm4, %ymm4
-; AVX2-NEXT: vpaddd %ymm1, %ymm4, %ymm1
+; AVX2-NEXT: vpaddd %ymm0, %ymm4, %ymm0
; AVX2-NEXT: vpmovzxwd {{.*#+}} ymm4 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
; AVX2-NEXT: vpmulld %ymm6, %ymm4, %ymm4
-; AVX2-NEXT: vpaddd %ymm0, %ymm4, %ymm0
+; AVX2-NEXT: vpaddd %ymm1, %ymm4, %ymm1
; AVX2-NEXT: vpmovzxwd {{.*#+}} ymm4 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
; AVX2-NEXT: vpmulld %ymm7, %ymm4, %ymm4
; AVX2-NEXT: vpaddd %ymm3, %ymm4, %ymm3
@@ -1739,9 +1739,9 @@ define i32 @test_unsigned_short_1024(ptr nocapture readonly, ptr nocapture reado
; AVX2-NEXT: cmpq %rcx, %rax
; AVX2-NEXT: jne .LBB11_1
; AVX2-NEXT: # %bb.2: # %middle.block
-; AVX2-NEXT: vpaddd %ymm2, %ymm0, %ymm0
-; AVX2-NEXT: vpaddd %ymm3, %ymm1, %ymm1
-; AVX2-NEXT: vpaddd %ymm1, %ymm0, %ymm0
+; AVX2-NEXT: vpaddd %ymm2, %ymm1, %ymm1
+; AVX2-NEXT: vpaddd %ymm3, %ymm0, %ymm0
+; AVX2-NEXT: vpaddd %ymm0, %ymm1, %ymm0
; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
; AVX2-NEXT: vpaddd %xmm1, %xmm0, %xmm0
; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,2,3]
@@ -1765,15 +1765,15 @@ define i32 @test_unsigned_short_1024(ptr nocapture readonly, ptr nocapture reado
; AVX512-NEXT: vpmovzxwd {{.*#+}} zmm3 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero,mem[8],zero,mem[9],zero,mem[10],zero,mem[11],zero,mem[12],zero,mem[13],zero,mem[14],zero,mem[15],zero
; AVX512-NEXT: vpmovzxwd {{.*#+}} zmm4 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero,mem[8],zero,mem[9],zero,mem[10],zero,mem[11],zero,mem[12],zero,mem[13],zero,mem[14],zero,mem[15],zero
; AVX512-NEXT: vpmulld %zmm2, %zmm4, %zmm2
-; AVX512-NEXT: vpaddd %zmm1, %zmm2, %zmm1
+; AVX512-NEXT: vpaddd %zmm0, %zmm2, %zmm0
; AVX512-NEXT: vpmovzxwd {{.*#+}} zmm2 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero,mem[8],zero,mem[9],zero,mem[10],zero,mem[11],zero,mem[12],zero,mem[13],zero,mem[14],zero,mem[15],zero
; AVX512-NEXT: vpmulld %zmm3, %zmm2, %zmm2
-; AVX512-NEXT: vpaddd %zmm0, %zmm2, %zmm0
+; AVX512-NEXT: vpaddd %zmm1, %zmm2, %zmm1
; AVX512-NEXT: addq $16, %rcx
; AVX512-NEXT: cmpq %rcx, %rax
; AVX512-NEXT: jne .LBB11_1
; AVX512-NEXT: # %bb.2: # %middle.block
-; AVX512-NEXT: vpaddd %zmm1, %zmm0, %zmm0
+; AVX512-NEXT: vpaddd %zmm0, %zmm1, %zmm0
; AVX512-NEXT: vextracti64x4 $1, %zmm0, %ymm1
; AVX512-NEXT: vpaddd %zmm1, %zmm0, %zmm0
; AVX512-NEXT: vextracti128 $1, %ymm0, %xmm1
diff --git a/llvm/test/CodeGen/X86/pr49451.ll b/llvm/test/CodeGen/X86/pr49451.ll
index 173c41140ebe..1a7551f6117e 100644
--- a/llvm/test/CodeGen/X86/pr49451.ll
+++ b/llvm/test/CodeGen/X86/pr49451.ll
@@ -18,15 +18,15 @@ define void @func_6(i8 %uc_8, i64 %uli_10) nounwind {
; X86-NEXT: .p2align 4
; X86-NEXT: .LBB0_1: # %for.body612
; X86-NEXT: # =>This Inner Loop Header: Depth=1
-; X86-NEXT: testb %dl, %dl
+; X86-NEXT: testb %bl, %bl
; X86-NEXT: je .LBB0_2
; X86-NEXT: # %bb.3: # %if.end1401
; X86-NEXT: # in Loop: Header=BB0_1 Depth=1
; X86-NEXT: addl %eax, %esi
; X86-NEXT: movw %si, s_2
-; X86-NEXT: movw %bx, s_0
+; X86-NEXT: movw %dx, s_0
; X86-NEXT: incl %ecx
-; X86-NEXT: incl %ebx
+; X86-NEXT: incl %edx
; X86-NEXT: cmpw $73, %cx
; X86-NEXT: jl .LBB0_1
; X86-NEXT: # %bb.4: # %for.body1703
diff --git a/llvm/test/CodeGen/X86/wide-scalar-shift-by-byte-multiple-legalization.ll b/llvm/test/CodeGen/X86/wide-scalar-shift-by-byte-multiple-legalization.ll
index 1c3d27fac420..c0962236f93d 100644
--- a/llvm/test/CodeGen/X86/wide-scalar-shift-by-byte-multiple-legalization.ll
+++ b/llvm/test/CodeGen/X86/wide-scalar-shift-by-byte-multiple-legalization.ll
@@ -12729,43 +12729,43 @@ define void @lshr_64bytes(ptr %src.ptr, ptr %byteOff.ptr, ptr %dst) nounwind {
; FALLBACK9-NEXT: pushq %rbx
; FALLBACK9-NEXT: vmovups (%rdi), %ymm0
; FALLBACK9-NEXT: vmovups 32(%rdi), %ymm1
-; FALLBACK9-NEXT: movl (%rsi), %edi
+; FALLBACK9-NEXT: movl (%rsi), %eax
; FALLBACK9-NEXT: vxorps %xmm2, %xmm2, %xmm2
; FALLBACK9-NEXT: vmovups %ymm2, -{{[0-9]+}}(%rsp)
; FALLBACK9-NEXT: vmovups %ymm2, -{{[0-9]+}}(%rsp)
; FALLBACK9-NEXT: vmovups %ymm1, -{{[0-9]+}}(%rsp)
; FALLBACK9-NEXT: vmovups %ymm0, -{{[0-9]+}}(%rsp)
-; FALLBACK9-NEXT: leal (,%rdi,8), %ecx
+; FALLBACK9-NEXT: leal (,%rax,8), %ecx
; FALLBACK9-NEXT: andl $56, %ecx
-; FALLBACK9-NEXT: andl $56, %edi
-; FALLBACK9-NEXT: movq -96(%rsp,%rdi), %rsi
-; FALLBACK9-NEXT: movq -104(%rsp,%rdi), %r9
-; FALLBACK9-NEXT: movq %r9, %rax
-; FALLBACK9-NEXT: shrdq %cl, %rsi, %rax
-; FALLBACK9-NEXT: movq -112(%rsp,%rdi), %r10
+; FALLBACK9-NEXT: andl $56, %eax
+; FALLBACK9-NEXT: movq -96(%rsp,%rax), %rdi
+; FALLBACK9-NEXT: movq -104(%rsp,%rax), %r9
+; FALLBACK9-NEXT: movq %r9, %rsi
+; FALLBACK9-NEXT: shrdq %cl, %rdi, %rsi
+; FALLBACK9-NEXT: movq -112(%rsp,%rax), %r10
; FALLBACK9-NEXT: movq %r10, %r8
; FALLBACK9-NEXT: shrdq %cl, %r9, %r8
-; FALLBACK9-NEXT: movq -80(%rsp,%rdi), %r9
-; FALLBACK9-NEXT: movq -88(%rsp,%rdi), %r11
+; FALLBACK9-NEXT: movq -80(%rsp,%rax), %r9
+; FALLBACK9-NEXT: movq -88(%rsp,%rax), %r11
; FALLBACK9-NEXT: movq %r11, %rbx
; FALLBACK9-NEXT: shrdq %cl, %r9, %rbx
-; FALLBACK9-NEXT: shrdq %cl, %r11, %rsi
-; FALLBACK9-NEXT: movq -72(%rsp,%rdi), %r11
+; FALLBACK9-NEXT: shrdq %cl, %r11, %rdi
+; FALLBACK9-NEXT: movq -72(%rsp,%rax), %r11
; FALLBACK9-NEXT: shrdq %cl, %r11, %r9
-; FALLBACK9-NEXT: movq -128(%rsp,%rdi), %r14
-; FALLBACK9-NEXT: movq -120(%rsp,%rdi), %rdi
-; FALLBACK9-NEXT: movq %rdi, %r15
+; FALLBACK9-NEXT: movq -128(%rsp,%rax), %r14
+; FALLBACK9-NEXT: movq -120(%rsp,%rax), %rax
+; FALLBACK9-NEXT: movq %rax, %r15
; FALLBACK9-NEXT: shrdq %cl, %r10, %r15
-; FALLBACK9-NEXT: shrdq %cl, %rdi, %r14
+; FALLBACK9-NEXT: shrdq %cl, %rax, %r14
; FALLBACK9-NEXT: # kill: def $cl killed $cl killed $ecx
; FALLBACK9-NEXT: shrq %cl, %r11
; FALLBACK9-NEXT: movq %r15, 8(%rdx)
; FALLBACK9-NEXT: movq %r9, 48(%rdx)
; FALLBACK9-NEXT: movq %r11, 56(%rdx)
-; FALLBACK9-NEXT: movq %rsi, 32(%rdx)
+; FALLBACK9-NEXT: movq %rdi, 32(%rdx)
; FALLBACK9-NEXT: movq %rbx, 40(%rdx)
; FALLBACK9-NEXT: movq %r8, 16(%rdx)
-; FALLBACK9-NEXT: movq %rax, 24(%rdx)
+; FALLBACK9-NEXT: movq %rsi, 24(%rdx)
; FALLBACK9-NEXT: movq %r14, (%rdx)
; FALLBACK9-NEXT: popq %rbx
; FALLBACK9-NEXT: popq %r14
@@ -12906,45 +12906,45 @@ define void @lshr_64bytes(ptr %src.ptr, ptr %byteOff.ptr, ptr %dst) nounwind {
; FALLBACK12-NEXT: pushq %rbx
; FALLBACK12-NEXT: pushq %rax
; FALLBACK12-NEXT: vmovups (%rdi), %zmm0
-; FALLBACK12-NEXT: movl (%rsi), %r10d
+; FALLBACK12-NEXT: movl (%rsi), %r9d
; FALLBACK12-NEXT: vxorps %xmm1, %xmm1, %xmm1
; FALLBACK12-NEXT: vmovups %zmm1, -{{[0-9]+}}(%rsp)
; FALLBACK12-NEXT: vmovups %zmm0, -{{[0-9]+}}(%rsp)
-; FALLBACK12-NEXT: leal (,%r10,8), %eax
+; FALLBACK12-NEXT: leal (,%r9,8), %eax
; FALLBACK12-NEXT: andl $56, %eax
-; FALLBACK12-NEXT: andl $56, %r10d
-; FALLBACK12-NEXT: movq -128(%rsp,%r10), %r9
-; FALLBACK12-NEXT: movq -120(%rsp,%r10), %r8
+; FALLBACK12-NEXT: andl $56, %r9d
+; FALLBACK12-NEXT: movq -128(%rsp,%r9), %r10
+; FALLBACK12-NEXT: movq -120(%rsp,%r9), %r8
; FALLBACK12-NEXT: movl %eax, %ecx
-; FALLBACK12-NEXT: shrq %cl, %r9
+; FALLBACK12-NEXT: shrq %cl, %r10
; FALLBACK12-NEXT: movl %eax, %esi
; FALLBACK12-NEXT: notb %sil
; FALLBACK12-NEXT: leaq (%r8,%r8), %rdi
; FALLBACK12-NEXT: movl %esi, %ecx
; FALLBACK12-NEXT: shlq %cl, %rdi
-; FALLBACK12-NEXT: orq %r9, %rdi
-; FALLBACK12-NEXT: movq -104(%rsp,%r10), %r9
-; FALLBACK12-NEXT: movq %r9, %rbx
+; FALLBACK12-NEXT: orq %r10, %rdi
+; FALLBACK12-NEXT: movq -104(%rsp,%r9), %r10
+; FALLBACK12-NEXT: movq %r10, %rbx
; FALLBACK12-NEXT: movl %eax, %ecx
; FALLBACK12-NEXT: shrq %cl, %rbx
-; FALLBACK12-NEXT: movq -96(%rsp,%r10), %r12
+; FALLBACK12-NEXT: movq -96(%rsp,%r9), %r12
; FALLBACK12-NEXT: leaq (%r12,%r12), %r11
; FALLBACK12-NEXT: movl %esi, %ecx
; FALLBACK12-NEXT: shlq %cl, %r11
; FALLBACK12-NEXT: orq %rbx, %r11
-; FALLBACK12-NEXT: movq -112(%rsp,%r10), %rbx
+; FALLBACK12-NEXT: movq -112(%rsp,%r9), %rbx
; FALLBACK12-NEXT: movq %rbx, %r14
; FALLBACK12-NEXT: movl %eax, %ecx
; FALLBACK12-NEXT: shrq %cl, %r14
-; FALLBACK12-NEXT: addq %r9, %r9
+; FALLBACK12-NEXT: addq %r10, %r10
; FALLBACK12-NEXT: movl %esi, %ecx
-; FALLBACK12-NEXT: shlq %cl, %r9
-; FALLBACK12-NEXT: orq %r14, %r9
-; FALLBACK12-NEXT: movq -88(%rsp,%r10), %r14
+; FALLBACK12-NEXT: shlq %cl, %r10
+; FALLBACK12-NEXT: orq %r14, %r10
+; FALLBACK12-NEXT: movq -88(%rsp,%r9), %r14
; FALLBACK12-NEXT: movq %r14, %r13
; FALLBACK12-NEXT: movl %eax, %ecx
; FALLBACK12-NEXT: shrq %cl, %r13
-; FALLBACK12-NEXT: movq -80(%rsp,%r10), %rbp
+; FALLBACK12-NEXT: movq -80(%rsp,%r9), %rbp
; FALLBACK12-NEXT: leaq (%rbp,%rbp), %r15
; FALLBACK12-NEXT: movl %esi, %ecx
; FALLBACK12-NEXT: shlq %cl, %r15
@@ -12957,8 +12957,8 @@ define void @lshr_64bytes(ptr %src.ptr, ptr %byteOff.ptr, ptr %dst) nounwind {
; FALLBACK12-NEXT: orq %r12, %r14
; FALLBACK12-NEXT: movl %eax, %ecx
; FALLBACK12-NEXT: shrq %cl, %rbp
-; FALLBACK12-NEXT: movq -72(%rsp,%r10), %r10
-; FALLBACK12-NEXT: leaq (%r10,%r10), %r12
+; FALLBACK12-NEXT: movq -72(%rsp,%r9), %r9
+; FALLBACK12-NEXT: leaq (%r9,%r9), %r12
; FALLBACK12-NEXT: movl %esi, %ecx
; FALLBACK12-NEXT: shlq %cl, %r12
; FALLBACK12-NEXT: orq %rbp, %r12
@@ -12969,13 +12969,13 @@ define void @lshr_64bytes(ptr %src.ptr, ptr %byteOff.ptr, ptr %dst) nounwind {
; FALLBACK12-NEXT: shlq %cl, %rbx
; FALLBACK12-NEXT: orq %r8, %rbx
; FALLBACK12-NEXT: movl %eax, %ecx
-; FALLBACK12-NEXT: shrq %cl, %r10
-; FALLBACK12-NEXT: movq %r10, 56(%rdx)
+; FALLBACK12-NEXT: shrq %cl, %r9
+; FALLBACK12-NEXT: movq %r9, 56(%rdx)
; FALLBACK12-NEXT: movq %rbx, 8(%rdx)
; FALLBACK12-NEXT: movq %r12, 48(%rdx)
; FALLBACK12-NEXT: movq %r14, 32(%rdx)
; FALLBACK12-NEXT: movq %r15, 40(%rdx)
-; FALLBACK12-NEXT: movq %r9, 16(%rdx)
+; FALLBACK12-NEXT: movq %r10, 16(%rdx)
; FALLBACK12-NEXT: movq %r11, 24(%rdx)
; FALLBACK12-NEXT: movq %rdi, (%rdx)
; FALLBACK12-NEXT: addq $8, %rsp
@@ -13111,40 +13111,40 @@ define void @lshr_64bytes(ptr %src.ptr, ptr %byteOff.ptr, ptr %dst) nounwind {
; FALLBACK15-NEXT: pushq %r14
; FALLBACK15-NEXT: pushq %rbx
; FALLBACK15-NEXT: vmovups (%rdi), %zmm0
-; FALLBACK15-NEXT: movl (%rsi), %edi
+; FALLBACK15-NEXT: movl (%rsi), %eax
; FALLBACK15-NEXT: vxorps %xmm1, %xmm1, %xmm1
; FALLBACK15-NEXT: vmovups %zmm1, -{{[0-9]+}}(%rsp)
; FALLBACK15-NEXT: vmovups %zmm0, -{{[0-9]+}}(%rsp)
-; FALLBACK15-NEXT: leal (,%rdi,8), %ecx
+; FALLBACK15-NEXT: leal (,%rax,8), %ecx
; FALLBACK15-NEXT: andl $56, %ecx
-; FALLBACK15-NEXT: andl $56, %edi
-; FALLBACK15-NEXT: movq -96(%rsp,%rdi), %rsi
-; FALLBACK15-NEXT: movq -104(%rsp,%rdi), %r9
-; FALLBACK15-NEXT: movq %r9, %rax
-; FALLBACK15-NEXT: shrdq %cl, %rsi, %rax
-; FALLBACK15-NEXT: movq -112(%rsp,%rdi), %r10
+; FALLBACK15-NEXT: andl $56, %eax
+; FALLBACK15-NEXT: movq -96(%rsp,%rax), %rdi
+; FALLBACK15-NEXT: movq -104(%rsp,%rax), %r9
+; FALLBACK15-NEXT: movq %r9, %rsi
+; FALLBACK15-NEXT: shrdq %cl, %rdi, %rsi
+; FALLBACK15-NEXT: movq -112(%rsp,%rax), %r10
; FALLBACK15-NEXT: movq %r10, %r8
; FALLBACK15-NEXT: shrdq %cl, %r9, %r8
-; FALLBACK15-NEXT: movq -80(%rsp,%rdi), %r9
-; FALLBACK15-NEXT: movq -88(%rsp,%rdi), %r11
+; FALLBACK15-NEXT: movq -80(%rsp,%rax), %r9
+; FALLBACK15-NEXT: movq -88(%rsp,%rax), %r11
; FALLBACK15-NEXT: movq %r11, %rbx
; FALLBACK15-NEXT: shrdq %cl, %r9, %rbx
-; FALLBACK15-NEXT: shrdq %cl, %r11, %rsi
-; FALLBACK15-NEXT: movq -72(%rsp,%rdi), %r11
+; FALLBACK15-NEXT: shrdq %cl, %r11, %rdi
+; FALLBACK15-NEXT: movq -72(%rsp,%rax), %r11
; FALLBACK15-NEXT: shrdq %cl, %r11, %r9
-; FALLBACK15-NEXT: movq -128(%rsp,%rdi), %r14
-; FALLBACK15-NEXT: movq -120(%rsp,%rdi), %rdi
-; FALLBACK15-NEXT: movq %rdi, %r15
+; FALLBACK15-NEXT: movq -128(%rsp,%rax), %r14
+; FALLBACK15-NEXT: movq -120(%rsp,%rax), %rax
+; FALLBACK15-NEXT: movq %rax, %r15
; FALLBACK15-NEXT: shrdq %cl, %r10, %r15
; FALLBACK15-NEXT: shrxq %rcx, %r11, %r10
; FALLBACK15-NEXT: # kill: def $cl killed $cl killed $rcx
-; FALLBACK15-NEXT: shrdq %cl, %rdi, %r14
+; FALLBACK15-NEXT: shrdq %cl, %rax, %r14
; FALLBACK15-NEXT: movq %r15, 8(%rdx)
; FALLBACK15-NEXT: movq %r9, 48(%rdx)
-; FALLBACK15-NEXT: movq %rsi, 32(%rdx)
+; FALLBACK15-NEXT: movq %rdi, 32(%rdx)
; FALLBACK15-NEXT: movq %rbx, 40(%rdx)
; FALLBACK15-NEXT: movq %r8, 16(%rdx)
-; FALLBACK15-NEXT: movq %rax, 24(%rdx)
+; FALLBACK15-NEXT: movq %rsi, 24(%rdx)
; FALLBACK15-NEXT: movq %r14, (%rdx)
; FALLBACK15-NEXT: movq %r10, 56(%rdx)
; FALLBACK15-NEXT: popq %rbx
diff --git a/llvm/test/CodeGen/X86/x86-shrink-wrapping.ll b/llvm/test/CodeGen/X86/x86-shrink-wrapping.ll
index 9fbbba2ed3b4..37620ecf8c1b 100644
--- a/llvm/test/CodeGen/X86/x86-shrink-wrapping.ll
+++ b/llvm/test/CodeGen/X86/x86-shrink-wrapping.ll
@@ -1185,10 +1185,10 @@ define i32 @useLEAForPrologue(i32 %d, i32 %a, i8 %c) #3 {
; ENABLE-NEXT: .p2align 4
; ENABLE-NEXT: LBB14_2: ## %for.body
; ENABLE-NEXT: ## =>This Inner Loop Header: Depth=1
-; ENABLE-NEXT: movl %esi, %eax
+; ENABLE-NEXT: cmpl %esi, %edi
+; ENABLE-NEXT: setl %al
; ENABLE-NEXT: xorl %esi, %esi
-; ENABLE-NEXT: cmpl %eax, %edi
-; ENABLE-NEXT: setl %sil
+; ENABLE-NEXT: movb %al, %sil
; ENABLE-NEXT: incb %dl
; ENABLE-NEXT: cmpb $45, %dl
; ENABLE-NEXT: jl LBB14_2
@@ -1220,10 +1220,10 @@ define i32 @useLEAForPrologue(i32 %d, i32 %a, i8 %c) #3 {
; DISABLE-NEXT: .p2align 4
; DISABLE-NEXT: LBB14_2: ## %for.body
; DISABLE-NEXT: ## =>This Inner Loop Header: Depth=1
-; DISABLE-NEXT: movl %esi, %eax
+; DISABLE-NEXT: cmpl %esi, %edi
+; DISABLE-NEXT: setl %al
; DISABLE-NEXT: xorl %esi, %esi
-; DISABLE-NEXT: cmpl %eax, %edi
-; DISABLE-NEXT: setl %sil
+; DISABLE-NEXT: movb %al, %sil
; DISABLE-NEXT: incb %dl
; DISABLE-NEXT: cmpb $45, %dl
; DISABLE-NEXT: jl LBB14_2
diff --git a/llvm/test/CodeGen/X86/xor.ll b/llvm/test/CodeGen/X86/xor.ll
index 59fbf7183abc..2bef66825d8c 100644
--- a/llvm/test/CodeGen/X86/xor.ll
+++ b/llvm/test/CodeGen/X86/xor.ll
@@ -62,12 +62,12 @@ define i32 @test4(i32 %a, i32 %b) nounwind {
; X86-NEXT: .p2align 4
; X86-NEXT: .LBB3_1: # %bb
; X86-NEXT: # =>This Inner Loop Header: Depth=1
-; X86-NEXT: movl %ecx, %edx
; X86-NEXT: xorl %ecx, %eax
-; X86-NEXT: movl %eax, %ecx
-; X86-NEXT: notl %ecx
-; X86-NEXT: andl %edx, %ecx
-; X86-NEXT: addl %ecx, %ecx
+; X86-NEXT: movl %eax, %edx
+; X86-NEXT: notl %edx
+; X86-NEXT: andl %ecx, %edx
+; X86-NEXT: addl %edx, %edx
+; X86-NEXT: movl %edx, %ecx
; X86-NEXT: jne .LBB3_1
; X86-NEXT: # %bb.2: # %bb12
; X86-NEXT: retl
@@ -78,12 +78,12 @@ define i32 @test4(i32 %a, i32 %b) nounwind {
; X64-LIN-NEXT: .p2align 4
; X64-LIN-NEXT: .LBB3_1: # %bb
; X64-LIN-NEXT: # =>This Inner Loop Header: Depth=1
-; X64-LIN-NEXT: movl %esi, %ecx
; X64-LIN-NEXT: xorl %esi, %eax
-; X64-LIN-NEXT: movl %eax, %esi
-; X64-LIN-NEXT: notl %esi
-; X64-LIN-NEXT: andl %ecx, %esi
-; X64-LIN-NEXT: addl %esi, %esi
+; X64-LIN-NEXT: movl %eax, %ecx
+; X64-LIN-NEXT: notl %ecx
+; X64-LIN-NEXT: andl %esi, %ecx
+; X64-LIN-NEXT: addl %ecx, %ecx
+; X64-LIN-NEXT: movl %ecx, %esi
; X64-LIN-NEXT: jne .LBB3_1
; X64-LIN-NEXT: # %bb.2: # %bb12
; X64-LIN-NEXT: retq
@@ -94,12 +94,12 @@ define i32 @test4(i32 %a, i32 %b) nounwind {
; X64-WIN-NEXT: .p2align 4
; X64-WIN-NEXT: .LBB3_1: # %bb
; X64-WIN-NEXT: # =>This Inner Loop Header: Depth=1
-; X64-WIN-NEXT: movl %edx, %ecx
; X64-WIN-NEXT: xorl %edx, %eax
-; X64-WIN-NEXT: movl %eax, %edx
-; X64-WIN-NEXT: notl %edx
-; X64-WIN-NEXT: andl %ecx, %edx
-; X64-WIN-NEXT: addl %edx, %edx
+; X64-WIN-NEXT: movl %eax, %ecx
+; X64-WIN-NEXT: notl %ecx
+; X64-WIN-NEXT: andl %edx, %ecx
+; X64-WIN-NEXT: addl %ecx, %ecx
+; X64-WIN-NEXT: movl %ecx, %edx
; X64-WIN-NEXT: jne .LBB3_1
; X64-WIN-NEXT: # %bb.2: # %bb12
; X64-WIN-NEXT: retq
@@ -126,13 +126,13 @@ define i16 @test5(i16 %a, i16 %b) nounwind {
; X86-NEXT: .p2align 4
; X86-NEXT: .LBB4_1: # %bb
; X86-NEXT: # =>This Inner Loop Header: Depth=1
-; X86-NEXT: movl %ecx, %edx
-; X86-NEXT: xorl %edx, %eax
-; X86-NEXT: movl %eax, %ecx
-; X86-NEXT: notl %ecx
-; X86-NEXT: andl %edx, %ecx
-; X86-NEXT: addl %ecx, %ecx
-; X86-NEXT: testw %cx, %cx
+; X86-NEXT: xorl %ecx, %eax
+; X86-NEXT: movl %eax, %edx
+; X86-NEXT: notl %edx
+; X86-NEXT: andl %ecx, %edx
+; X86-NEXT: addl %edx, %edx
+; X86-NEXT: testw %dx, %dx
+; X86-NEXT: movl %edx, %ecx
; X86-NEXT: jne .LBB4_1
; X86-NEXT: # %bb.2: # %bb12
; X86-NEXT: # kill: def $ax killed $ax killed $eax
@@ -144,13 +144,13 @@ define i16 @test5(i16 %a, i16 %b) nounwind {
; X64-LIN-NEXT: .p2align 4
; X64-LIN-NEXT: .LBB4_1: # %bb
; X64-LIN-NEXT: # =>This Inner Loop Header: Depth=1
-; X64-LIN-NEXT: movl %esi, %ecx
-; X64-LIN-NEXT: xorl %ecx, %eax
-; X64-LIN-NEXT: movl %eax, %esi
-; X64-LIN-NEXT: notl %esi
-; X64-LIN-NEXT: andl %ecx, %esi
-; X64-LIN-NEXT: addl %esi, %esi
-; X64-LIN-NEXT: testw %si, %si
+; X64-LIN-NEXT: xorl %esi, %eax
+; X64-LIN-NEXT: movl %eax, %ecx
+; X64-LIN-NEXT: notl %ecx
+; X64-LIN-NEXT: andl %esi, %ecx
+; X64-LIN-NEXT: addl %ecx, %ecx
+; X64-LIN-NEXT: testw %cx, %cx
+; X64-LIN-NEXT: movl %ecx, %esi
; X64-LIN-NEXT: jne .LBB4_1
; X64-LIN-NEXT: # %bb.2: # %bb12
; X64-LIN-NEXT: # kill: def $ax killed $ax killed $eax
@@ -163,13 +163,13 @@ define i16 @test5(i16 %a, i16 %b) nounwind {
; X64-WIN-NEXT: .p2align 4
; X64-WIN-NEXT: .LBB4_1: # %bb
; X64-WIN-NEXT: # =>This Inner Loop Header: Depth=1
-; X64-WIN-NEXT: movl %edx, %ecx
-; X64-WIN-NEXT: xorl %ecx, %eax
-; X64-WIN-NEXT: movl %eax, %edx
-; X64-WIN-NEXT: notl %edx
-; X64-WIN-NEXT: andl %ecx, %edx
-; X64-WIN-NEXT: addl %edx, %edx
-; X64-WIN-NEXT: testw %dx, %dx
+; X64-WIN-NEXT: xorl %edx, %eax
+; X64-WIN-NEXT: movl %eax, %ecx
+; X64-WIN-NEXT: notl %ecx
+; X64-WIN-NEXT: andl %edx, %ecx
+; X64-WIN-NEXT: addl %ecx, %ecx
+; X64-WIN-NEXT: testw %cx, %cx
+; X64-WIN-NEXT: movl %ecx, %edx
; X64-WIN-NEXT: jne .LBB4_1
; X64-WIN-NEXT: # %bb.2: # %bb12
; X64-WIN-NEXT: # kill: def $ax killed $ax killed $eax
@@ -197,12 +197,12 @@ define i8 @test6(i8 %a, i8 %b) nounwind {
; X86-NEXT: .p2align 4
; X86-NEXT: .LBB5_1: # %bb
; X86-NEXT: # =>This Inner Loop Header: Depth=1
-; X86-NEXT: movl %ecx, %edx
; X86-NEXT: xorb %cl, %al
-; X86-NEXT: movl %eax, %ecx
-; X86-NEXT: notb %cl
-; X86-NEXT: andb %dl, %cl
-; X86-NEXT: addb %cl, %cl
+; X86-NEXT: movl %eax, %edx
+; X86-NEXT: notb %dl
+; X86-NEXT: andb %cl, %dl
+; X86-NEXT: addb %dl, %dl
+; X86-NEXT: movl %edx, %ecx
; X86-NEXT: jne .LBB5_1
; X86-NEXT: # %bb.2: # %bb12
; X86-NEXT: retl
@@ -213,12 +213,12 @@ define i8 @test6(i8 %a, i8 %b) nounwind {
; X64-LIN-NEXT: .p2align 4
; X64-LIN-NEXT: .LBB5_1: # %bb
; X64-LIN-NEXT: # =>This Inner Loop Header: Depth=1
-; X64-LIN-NEXT: movl %esi, %ecx
; X64-LIN-NEXT: xorb %sil, %al
-; X64-LIN-NEXT: movl %eax, %esi
-; X64-LIN-NEXT: notb %sil
-; X64-LIN-NEXT: andb %cl, %sil
-; X64-LIN-NEXT: addb %sil, %sil
+; X64-LIN-NEXT: movl %eax, %ecx
+; X64-LIN-NEXT: notb %cl
+; X64-LIN-NEXT: andb %sil, %cl
+; X64-LIN-NEXT: addb %cl, %cl
+; X64-LIN-NEXT: movl %ecx, %esi
; X64-LIN-NEXT: jne .LBB5_1
; X64-LIN-NEXT: # %bb.2: # %bb12
; X64-LIN-NEXT: # kill: def $al killed $al killed $eax
@@ -230,12 +230,12 @@ define i8 @test6(i8 %a, i8 %b) nounwind {
; X64-WIN-NEXT: .p2align 4
; X64-WIN-NEXT: .LBB5_1: # %bb
; X64-WIN-NEXT: # =>This Inner Loop Header: Depth=1
-; X64-WIN-NEXT: movl %edx, %ecx
; X64-WIN-NEXT: xorb %dl, %al
-; X64-WIN-NEXT: movl %eax, %edx
-; X64-WIN-NEXT: notb %dl
-; X64-WIN-NEXT: andb %cl, %dl
-; X64-WIN-NEXT: addb %dl, %dl
+; X64-WIN-NEXT: movl %eax, %ecx
+; X64-WIN-NEXT: notb %cl
+; X64-WIN-NEXT: andb %dl, %cl
+; X64-WIN-NEXT: addb %cl, %cl
+; X64-WIN-NEXT: movl %ecx, %edx
; X64-WIN-NEXT: jne .LBB5_1
; X64-WIN-NEXT: # %bb.2: # %bb12
; X64-WIN-NEXT: retq
@@ -262,12 +262,12 @@ define i32 @test7(i32 %a, i32 %b) nounwind {
; X86-NEXT: .p2align 4
; X86-NEXT: .LBB6_1: # %bb
; X86-NEXT: # =>This Inner Loop Header: Depth=1
-; X86-NEXT: movl %ecx, %edx
; X86-NEXT: xorl %ecx, %eax
-; X86-NEXT: movl %eax, %ecx
-; X86-NEXT: xorl $2147483646, %ecx # imm = 0x7FFFFFFE
-; X86-NEXT: andl %edx, %ecx
-; X86-NEXT: addl %ecx, %ecx
+; X86-NEXT: movl %eax, %edx
+; X86-NEXT: xorl $2147483646, %edx # imm = 0x7FFFFFFE
+; X86-NEXT: andl %ecx, %edx
+; X86-NEXT: addl %edx, %edx
+; X86-NEXT: movl %edx, %ecx
; X86-NEXT: jne .LBB6_1
; X86-NEXT: # %bb.2: # %bb12
; X86-NEXT: retl
@@ -278,12 +278,12 @@ define i32 @test7(i32 %a, i32 %b) nounwind {
; X64-LIN-NEXT: .p2align 4
; X64-LIN-NEXT: .LBB6_1: # %bb
; X64-LIN-NEXT: # =>This Inner Loop Header: Depth=1
-; X64-LIN-NEXT: movl %esi, %ecx
; X64-LIN-NEXT: xorl %esi, %eax
-; X64-LIN-NEXT: movl %eax, %esi
-; X64-LIN-NEXT: xorl $2147483646, %esi # imm = 0x7FFFFFFE
-; X64-LIN-NEXT: andl %ecx, %esi
-; X64-LIN-NEXT: addl %esi, %esi
+; X64-LIN-NEXT: movl %eax, %ecx
+; X64-LIN-NEXT: xorl $2147483646, %ecx # imm = 0x7FFFFFFE
+; X64-LIN-NEXT: andl %esi, %ecx
+; X64-LIN-NEXT: addl %ecx, %ecx
+; X64-LIN-NEXT: movl %ecx, %esi
; X64-LIN-NEXT: jne .LBB6_1
; X64-LIN-NEXT: # %bb.2: # %bb12
; X64-LIN-NEXT: retq
@@ -294,12 +294,12 @@ define i32 @test7(i32 %a, i32 %b) nounwind {
; X64-WIN-NEXT: .p2align 4
; X64-WIN-NEXT: .LBB6_1: # %bb
; X64-WIN-NEXT: # =>This Inner Loop Header: Depth=1
-; X64-WIN-NEXT: movl %edx, %ecx
; X64-WIN-NEXT: xorl %edx, %eax
-; X64-WIN-NEXT: movl %eax, %edx
-; X64-WIN-NEXT: xorl $2147483646, %edx # imm = 0x7FFFFFFE
-; X64-WIN-NEXT: andl %ecx, %edx
-; X64-WIN-NEXT: addl %edx, %edx
+; X64-WIN-NEXT: movl %eax, %ecx
+; X64-WIN-NEXT: xorl $2147483646, %ecx # imm = 0x7FFFFFFE
+; X64-WIN-NEXT: andl %edx, %ecx
+; X64-WIN-NEXT: addl %ecx, %ecx
+; X64-WIN-NEXT: movl %ecx, %edx
; X64-WIN-NEXT: jne .LBB6_1
; X64-WIN-NEXT: # %bb.2: # %bb12
; X64-WIN-NEXT: retq
generated by cgit v1.2.3 (git 2.25.1) at 2026-05-31 09:12:50 +0000