[X86][SSE2] Fix asm string for movq (Move Quadword) instruction

[aymanmus]

Summary

Replace "mov{d|q}" with "movq".

[igorb]

Could you please add encoding tests for MOVPQIto64rr.

[aymanmus]

test/MC/X86/x86-64.s

1308 ↗

(On Diff #95720)

@igorb this is a test for MOVPQIto64rr.

• Notice that I changed 6 instructions, 4 of them are marked as isCodeGenOnly=1, so basically these tests covers them all.

[zvi]

LGTM. Let Igor give the final ok.
Are the VEX/EVEX variants correct?

test/CodeGen/X86/vector-pcmp.ll

2 ↗

(On Diff #95720)

Please remove this line

test/MC/X86/x86-64.s

1304 ↗

(On Diff #95720)

Is this test-case a duplicate of the above?

[aymanmus]

@zvi, yes, the VEX/EVEX variants are correct.

[RKSimon]

test/CodeGen/X86/vector-pcmp.ll

2 ↗

(On Diff #95720)

If you can please commit the conversion to update_llc_test_checks.py (with trailing ; line removal) as an NFC pre-commit - this patch is big already!

[craig.topper]

This wasn't a mistake. We have aliases for movq for this instruction already. This was for compatibility with MacOS/Darwin's old assembler from the pre-clang days.

See this code

These are the correct encodings of the instructions so that we know how to
read correct assembly, even though we continue to emit the wrong ones for
// compatibility with Darwin's buggy assembler.
def : InstAlias<"movq\t{$src, $dst|$dst, $src}",

(MOV64toPQIrr VR128:$dst, GR64:$src), 0>;

def : InstAlias<"movq\t{$src, $dst|$dst, $src}",

(MOVPQIto64rr GR64:$dst, VR128:$src), 0>;

// Allow "vmovd" but print "vmovq" since we don't need compatibility for AVX.
def : InstAlias<"vmovd\t{$src, $dst|$dst, $src}",

(VMOV64toPQIrr VR128:$dst, GR64:$src), 0>;

def : InstAlias<"vmovd\t{$src, $dst|$dst, $src}",

(VMOVPQIto64rr GR64:$dst, VR128:$src), 0>;

[aymanmus]

@craig.topper is this still needed? Why would we insert a "bug" only to be compatible with MacOS?
What about the memory form instructions? InstAlias cannot solve this for them.

[craig.topper]

In the days before the MC layer of LLVM existed, clang would output assembly as text and feed it to the separate assembler tool. So it had to output what the MacOS assembler could parse.

I believe the assembler bug occurred around when 64-bit was introduced. There was already separate 64-bit movq to/from memory instruction before that. Then movd picked up a rex prefix that allowed it to access 64-bit registers and redudantly access 64-bit memory. I don't think there was ever a movq to memory issue because we never expect to use the rex encoded memory instruction since the older movq instruction is available.

We probably don't need to output that way anymore. But we should keep an alias around so we can process old assembly filed. So switch the instruction string to movq and change the alias to movd.

[craig.topper]

LGTM

lib/Target/X86/X86InstrSSE.td

4816 ↗

(On Diff #96295)

aliases is mispelled.

llvm/trunk/lib/Target/X86/X86InstrSSE.td

def MOVDI2PDIrr : S2I<0x6E, MRMSrcReg, (outs VR128:$dst), (ins GR32:$src),
                         (v4i32 (scalar_to_vector GR32:$src)))], IIC_SSE_MOVDQ>,
                   Sched<[WriteMove]>;
 def MOVDI2PDIrm : S2I<0x6E, MRMSrcMem, (outs VR128:$dst), (ins i32mem:$src),
                       "movd\t{$src, $dst|$dst, $src}",
                       [(set VR128:$dst,
                         (v4i32 (scalar_to_vector (loadi32 addr:$src))))],
                         IIC_SSE_MOVDQ>, Sched<[WriteLoad]>;
 def MOV64toPQIrr : RS2I<0x6E, MRMSrcReg, (outs VR128:$dst), (ins GR64:$src),
-                        "mov{d|q}\t{$src, $dst|$dst, $src}",
+                        "movq\t{$src, $dst|$dst, $src}",
                         [(set VR128:$dst,
                           (v2i64 (scalar_to_vector GR64:$src)))],
                           IIC_SSE_MOVDQ>, Sched<[WriteMove]>;
 let isCodeGenOnly = 1, ForceDisassemble = 1, hasSideEffects = 0, mayLoad = 1 in
 def MOV64toPQIrm : RS2I<0x6E, MRMSrcMem, (outs VR128:$dst), (ins i64mem:$src),
-                        "mov{d|q}\t{$src, $dst|$dst, $src}",
+                        "movq\t{$src, $dst|$dst, $src}",
                         [], IIC_SSE_MOVDQ>, Sched<[WriteLoad]>;
 let isCodeGenOnly = 1 in
 def MOV64toSDrr : RS2I<0x6E, MRMSrcReg, (outs FR64:$dst), (ins GR64:$src),
-                       "mov{d|q}\t{$src, $dst|$dst, $src}",
+                       "movq\t{$src, $dst|$dst, $src}",
                        [(set FR64:$dst, (bitconvert GR64:$src))],
                        IIC_SSE_MOVDQ>, Sched<[WriteMove]>;
 } // ExeDomain = SSEPackedInt
 
 //===---------------------------------------------------------------------===//
 // Move Int Doubleword to Single Scalar
 //
 let ExeDomain = SSEPackedInt, isCodeGenOnly = 1 in {

 def VMOVPQIto64rr : VRS2I<0x7E, MRMDestReg, (outs GR64:$dst), (ins VR128:$src),
                           "movq\t{$src, $dst|$dst, $src}",
                           [(set GR64:$dst, (extractelt (v2i64 VR128:$src),
                                                         (iPTR 0)))],
                                                            IIC_SSE_MOVD_ToGP>,
                       VEX;
 
 def MOVPQIto64rr : RS2I<0x7E, MRMDestReg, (outs GR64:$dst), (ins VR128:$src),
-                        "mov{d|q}\t{$src, $dst|$dst, $src}",
+                        "movq\t{$src, $dst|$dst, $src}",
                         [(set GR64:$dst, (extractelt (v2i64 VR128:$src),
                                                          (iPTR 0)))],
                                                          IIC_SSE_MOVD_ToGP>;
 } //SchedRW
 
 let isCodeGenOnly = 1, ForceDisassemble = 1, hasSideEffects = 0, mayStore = 1 in
 def VMOVPQIto64mr : VRS2I<0x7E, MRMDestMem, (outs),
                           (ins i64mem:$dst, VR128:$src),
                           "movq\t{$src, $dst|$dst, $src}",
                           [], IIC_SSE_MOVDQ>, VEX, Sched<[WriteStore]>;
 let isCodeGenOnly = 1, ForceDisassemble = 1, hasSideEffects = 0, mayStore = 1 in
 def MOVPQIto64mr : RS2I<0x7E, MRMDestMem, (outs), (ins i64mem:$dst, VR128:$src),
-                        "mov{d|q}\t{$src, $dst|$dst, $src}",
+                        "movq\t{$src, $dst|$dst, $src}",
                         [], IIC_SSE_MOVDQ>, Sched<[WriteStore]>;
 } // ExeDomain = SSEPackedInt
 
 //===---------------------------------------------------------------------===//
 // Bitcast FR64 <-> GR64
 //
 let ExeDomain = SSEPackedInt, isCodeGenOnly = 1 in {
   let Predicates = [UseAVX] in
def VMOVSDto64mr : VRS2I<0x7E, MRMDestMem, (outs), (ins i64mem:$dst, FR64:$src),
                            [(store (i64 (bitconvert FR64:$src)), addr:$dst)],
                            IIC_SSE_MOVDQ>, VEX, Sched<[WriteStore]>;
 
   def MOV64toSDrm : S2SI<0x7E, MRMSrcMem, (outs FR64:$dst), (ins i64mem:$src),
                          "movq\t{$src, $dst|$dst, $src}",
                          [(set FR64:$dst, (bitconvert (loadi64 addr:$src)))],
                          IIC_SSE_MOVDQ>, Sched<[WriteLoad]>;
   def MOVSDto64rr : RS2I<0x7E, MRMDestReg, (outs GR64:$dst), (ins FR64:$src),
-                         "mov{d|q}\t{$src, $dst|$dst, $src}",
+                         "movq\t{$src, $dst|$dst, $src}",
                          [(set GR64:$dst, (bitconvert FR64:$src))],
                          IIC_SSE_MOVD_ToGP>, Sched<[WriteMove]>;
   def MOVSDto64mr : RS2I<0x7E, MRMDestMem, (outs), (ins i64mem:$dst, FR64:$src),
                          "movq\t{$src, $dst|$dst, $src}",
                          [(store (i64 (bitconvert FR64:$src)), addr:$dst)],
                          IIC_SSE_MOVDQ>, Sched<[WriteStore]>;
 } // ExeDomain = SSEPackedInt, isCodeGenOnly = 1
 
def : Pat<(v4i32 (X86vzmovl (bc_v4i32 (loadv4f32 addr:$src)))),
               (MOVDI2PDIrm addr:$src)>;
     def : Pat<(v4i32 (X86vzmovl (bc_v4i32 (loadv2i64 addr:$src)))),
               (MOVDI2PDIrm addr:$src)>;
     def : Pat<(v4i32 (X86vzload addr:$src)),
               (MOVDI2PDIrm addr:$src)>;
   }
 }
 
-// These are the correct encodings of the instructions so that we know how to
-// read correct assembly, even though we continue to emit the wrong ones for
-// compatibility with Darwin's buggy assembler.
-def : InstAlias<"movq\t{$src, $dst|$dst, $src}",
+// Before the MC layer of LLVM existed, clang emitted "movd" assembly instead of
+// "movq" due to MacOS parsing limitation. In order to parse old assembly, we add
+// these aliases.
+def : InstAlias<"movd\t{$src, $dst|$dst, $src}",
                 (MOV64toPQIrr VR128:$dst, GR64:$src), 0>;
-def : InstAlias<"movq\t{$src, $dst|$dst, $src}",
+def : InstAlias<"movd\t{$src, $dst|$dst, $src}",
                 (MOVPQIto64rr GR64:$dst, VR128:$src), 0>;
 // Allow "vmovd" but print "vmovq" since we don't need compatibility for AVX.
 def : InstAlias<"vmovd\t{$src, $dst|$dst, $src}",
                 (VMOV64toPQIrr VR128:$dst, GR64:$src), 0>;
 def : InstAlias<"vmovd\t{$src, $dst|$dst, $src}",
                 (VMOVPQIto64rr GR64:$dst, VR128:$src), 0>;
 
 //===---------------------------------------------------------------------===//

llvm/trunk/test/CodeGen/X86/GlobalISel/memop.ll

; ALL_AVX-NEXT:    retq
   %r = load float, float* %p1
   ret float %r
 }
 
 define double @test_load_double(double * %p1) {
 ; SSE-LABEL: test_load_double:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movq (%rdi), %rax
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; ALL_AVX-LABEL: test_load_double:
 ; ALL_AVX:       # BB#0:
 ; ALL_AVX-NEXT:    movq (%rdi), %rax
 ; ALL_AVX-NEXT:    vmovq %rax, %xmm0
 ; ALL_AVX-NEXT:    retq
   %r = load double, double* %p1
; ALL_AVX_GREEDY-NEXT:    retq
   store float %val, float* %p1
   ret float * %p1;
 }
 
 define double * @test_store_double(double %val, double * %p1) {
 ;
 ; SSE_FAST-LABEL: test_store_double:
 ; SSE_FAST:       # BB#0:
-; SSE_FAST-NEXT:    movd %xmm0, %rax
+; SSE_FAST-NEXT:    movq %xmm0, %rax
 ; SSE_FAST-NEXT:    movq %rax, (%rdi)
 ; SSE_FAST-NEXT:    movq %rdi, %rax
 ; SSE_FAST-NEXT:    retq
 ;
 ; SSE_GREEDY-LABEL: test_store_double:
 ; SSE_GREEDY:       # BB#0:
 ; SSE_GREEDY-NEXT:    movsd %xmm0, (%rdi)
 ; SSE_GREEDY-NEXT:    movq %rdi, %rax

llvm/trunk/test/CodeGen/X86/asm-reg-type-mismatch.ll

 }
 
 define i64 @test2() nounwind {
 entry:
   %0 = tail call i64 asm sideeffect "movq $1, $0", "={xmm7},*m,~{dirflag},~{fpsr},~{flags}"(i64* null) nounwind
   ret i64 %0
   ; CHECK: test2
   ; CHECK: movq {{.*}}, %xmm7
-  ; CHECK: movd %xmm7, %rax
+  ; CHECK: movq %xmm7, %rax
 }

llvm/trunk/test/CodeGen/X86/atomic-non-integer.ll

 ; CHECK: movd %xmm0, %eax
 ; CHECK: movl %eax, (%rdi)
   store atomic float %v, float* %fptr unordered, align 4
   ret void
 }
 
 define void @store_double(double* %fptr, double %v) {
 ; CHECK-LABEL: @store_double
-; CHECK: movd %xmm0, %rax
+; CHECK: movq %xmm0, %rax
 ; CHECK: movq %rax, (%rdi)
   store atomic double %v, double* %fptr unordered, align 8
   ret void
 }
 
 define void @store_fp128(fp128* %fptr, fp128 %v) {
 ; CHECK-LABEL: @store_fp128
 ; CHECK: callq  __sync_lock_test_and_set_16

 ; CHECK: movd %eax, %xmm0
   %v = load atomic float, float* %fptr unordered, align 4
   ret float %v
 }
 
 define double @load_double(double* %fptr) {
 ; CHECK-LABEL: @load_double
 ; CHECK: movq (%rdi), %rax
-; CHECK: movd %rax, %xmm0
+; CHECK: movq %rax, %xmm0
   %v = load atomic double, double* %fptr unordered, align 8
   ret double %v
 }
 
 define fp128 @load_fp128(fp128* %fptr) {
 ; CHECK-LABEL: @load_fp128
 ; CHECK: callq  __sync_val_compare_and_swap_16
   %v = load atomic fp128, fp128* %fptr unordered, align 16

 ; CHECK: movd %xmm0, %eax
 ; CHECK: xchgl  %eax, (%rdi)
   store atomic float %v, float* %fptr seq_cst, align 4
   ret void
 }
 
 define void @store_double_seq_cst(double* %fptr, double %v) {
 ; CHECK-LABEL: @store_double_seq_cst
-; CHECK: movd %xmm0, %rax
+; CHECK: movq %xmm0, %rax
 ; CHECK: xchgq  %rax, (%rdi)
   store atomic double %v, double* %fptr seq_cst, align 8
   ret void
 }
 
 define float @load_float_seq_cst(float* %fptr) {
 ; CHECK-LABEL: @load_float_seq_cst
 ; CHECK: movl (%rdi), %eax
 ; CHECK: movd %eax, %xmm0
   %v = load atomic float, float* %fptr seq_cst, align 4
   ret float %v
 }
 
 define double @load_double_seq_cst(double* %fptr) {
 ; CHECK-LABEL: @load_double_seq_cst
 ; CHECK: movq (%rdi), %rax
-; CHECK: movd %rax, %xmm0
+; CHECK: movq %rax, %xmm0
   %v = load atomic double, double* %fptr seq_cst, align 8
   ret double %v
 }

llvm/trunk/test/CodeGen/X86/bitcast2.ll

-; RUN: llc < %s -march=x86-64 -mattr=-avx | grep movd | count 2
+; RUN: llc < %s -march=x86-64 -mattr=-avx | grep movq | count 2
 ; RUN: llc < %s -march=x86-64 -mattr=-avx | not grep rsp
 
 define i64 @test1(double %A) {
    %B = bitcast double %A to i64
    ret i64 %B
 }
 
 define double @test2(i64 %A) {
    %B = bitcast i64 %A to double
    ret double %B
 }
 

llvm/trunk/test/CodeGen/X86/clear_upper_vector_element_bits.ll

 }
 
 define <16 x i8> @_clearupper16xi8b(<16 x i8>) nounwind {
 ; SSE-LABEL: _clearupper16xi8b:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    pushq %r14
 ; SSE-NEXT:    pushq %rbx
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; SSE-NEXT:    movd %xmm0, %rcx
+; SSE-NEXT:    movq %xmm0, %rcx
 ; SSE-NEXT:    movq %rcx, %r8
 ; SSE-NEXT:    movq %rcx, %r9
 ; SSE-NEXT:    movq %rcx, %r10
 ; SSE-NEXT:    movq %rcx, %rax
 ; SSE-NEXT:    movq %rcx, %rdx
 ; SSE-NEXT:    movq %rcx, %rsi
 ; SSE-NEXT:    movq %rcx, %rdi
 ; SSE-NEXT:    andb $15, %cl
 ; SSE-NEXT:    movb %cl, -{{[0-9]+}}(%rsp)
-; SSE-NEXT:    movd %xmm1, %rcx
+; SSE-NEXT:    movq %xmm1, %rcx
 ; SSE-NEXT:    shrq $56, %rdi
 ; SSE-NEXT:    andb $15, %dil
 ; SSE-NEXT:    movb %dil, -{{[0-9]+}}(%rsp)
 ; SSE-NEXT:    movq %rcx, %r11
 ; SSE-NEXT:    shrq $48, %rsi
 ; SSE-NEXT:    andb $15, %sil
 ; SSE-NEXT:    movb %sil, -{{[0-9]+}}(%rsp)
 ; SSE-NEXT:    movq %rcx, %r14

 }
 
 define <32 x i8> @_clearupper32xi8b(<32 x i8>) nounwind {
 ; SSE-LABEL: _clearupper32xi8b:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    pushq %r14
 ; SSE-NEXT:    pushq %rbx
 ; SSE-NEXT:    pshufd {{.*#+}} xmm2 = xmm0[2,3,0,1]
-; SSE-NEXT:    movd %xmm0, %rcx
+; SSE-NEXT:    movq %xmm0, %rcx
 ; SSE-NEXT:    movq %rcx, %r8
 ; SSE-NEXT:    movq %rcx, %r9
 ; SSE-NEXT:    movq %rcx, %r10
 ; SSE-NEXT:    movq %rcx, %rax
 ; SSE-NEXT:    movq %rcx, %rdx
 ; SSE-NEXT:    movq %rcx, %rsi
 ; SSE-NEXT:    movq %rcx, %rdi
 ; SSE-NEXT:    andb $15, %cl
 ; SSE-NEXT:    movb %cl, -{{[0-9]+}}(%rsp)
-; SSE-NEXT:    movd %xmm2, %rcx
+; SSE-NEXT:    movq %xmm2, %rcx
 ; SSE-NEXT:    shrq $56, %rdi
 ; SSE-NEXT:    andb $15, %dil
 ; SSE-NEXT:    movb %dil, -{{[0-9]+}}(%rsp)
 ; SSE-NEXT:    movq %rcx, %r11
 ; SSE-NEXT:    shrq $48, %rsi
 ; SSE-NEXT:    andb $15, %sil
 ; SSE-NEXT:    movb %sil, -{{[0-9]+}}(%rsp)
 ; SSE-NEXT:    movq %rcx, %r14

llvm/trunk/test/CodeGen/X86/dagcombine-cse.ll

 ; X64-NEXT:    leal (%rsi,%rdx), %eax
 ; X64-NEXT:    cltq
 ; X64-NEXT:    leal 4(%rsi,%rdx), %ecx
 ; X64-NEXT:    movslq %ecx, %rcx
 ; X64-NEXT:    movzwl (%rdi,%rcx), %ecx
 ; X64-NEXT:    shlq $32, %rcx
 ; X64-NEXT:    movl (%rdi,%rax), %eax
 ; X64-NEXT:    orq %rcx, %rax
-; X64-NEXT:    movd %rax, %xmm0
+; X64-NEXT:    movq %rax, %xmm0
 ; X64-NEXT:    movq {{.*#+}} xmm0 = xmm0[0],zero
 ; X64-NEXT:    pshuflw {{.*#+}} xmm0 = xmm0[0,1,1,2,4,5,6,7]
 ; X64-NEXT:    movd %xmm0, %eax
 ; X64-NEXT:    retq
 entry:
   %tmp7 = mul i32 %idxY, %ref_frame_stride    ; <i32> [#uses=2]
   %tmp9 = add i32 %tmp7, %idxX    ; <i32> [#uses=1]
   %tmp11 = getelementptr i8, i8* %ref_frame_ptr, i32 %tmp9    ; <i8*> [#uses=1]

llvm/trunk/test/CodeGen/X86/extractelement-index.ll

 ; AVX2-NEXT:    retq
   %b = extractelement <8 x i32> %a, i64 7
   ret i32 %b
 }
 
 define i64 @extractelement_v2i64_0(<2 x i64> %a, i256 %i) nounwind {
 ; SSE-LABEL: extractelement_v2i64_0:
 ; SSE:       # BB#0:
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: extractelement_v2i64_0:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vmovq %xmm0, %rax
 ; AVX-NEXT:    retq
   %b = extractelement <2 x i64> %a, i256 0
   ret i64 %b
 }
 
 define i64 @extractelement_v2i64_1(<2 x i64> %a, i256 %i) nounwind {
 ; SSE2-LABEL: extractelement_v2i64_1:
 ; SSE2:       # BB#0:
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    movq %xmm0, %rax
 ; SSE2-NEXT:    retq
 ;
 ; SSE41-LABEL: extractelement_v2i64_1:
 ; SSE41:       # BB#0:
 ; SSE41-NEXT:    pextrq $1, %xmm0, %rax
 ; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: extractelement_v2i64_1:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vpextrq $1, %xmm0, %rax
 ; AVX-NEXT:    retq
   %b = extractelement <2 x i64> %a, i256 1
   ret i64 %b
 }
 
 define i64 @extractelement_v4i64_1(<4 x i64> %a, i256 %i) nounwind {
 ; SSE2-LABEL: extractelement_v4i64_1:
 ; SSE2:       # BB#0:
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    movq %xmm0, %rax
 ; SSE2-NEXT:    retq
 ;
 ; SSE41-LABEL: extractelement_v4i64_1:
 ; SSE41:       # BB#0:
 ; SSE41-NEXT:    pextrq $1, %xmm0, %rax
 ; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: extractelement_v4i64_1:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vpextrq $1, %xmm0, %rax
 ; AVX-NEXT:    vzeroupper
 ; AVX-NEXT:    retq
   %b = extractelement <4 x i64> %a, i256 1
   ret i64 %b
 }
 
 define i64 @extractelement_v4i64_3(<4 x i64> %a, i256 %i) nounwind {
 ; SSE2-LABEL: extractelement_v4i64_3:
 ; SSE2:       # BB#0:
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[2,3,0,1]
-; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    movq %xmm0, %rax
 ; SSE2-NEXT:    retq
 ;
 ; SSE41-LABEL: extractelement_v4i64_3:
 ; SSE41:       # BB#0:
 ; SSE41-NEXT:    pextrq $1, %xmm1, %rax
 ; SSE41-NEXT:    retq
 ;
 ; AVX1-LABEL: extractelement_v4i64_3:

llvm/trunk/test/CodeGen/X86/gather-addresses.ll

 ; RUN: llc -mtriple=x86_64-linux -mcpu=nehalem < %s | FileCheck %s --check-prefix=LIN
 ; RUN: llc -mtriple=x86_64-win32 -mcpu=nehalem < %s | FileCheck %s --check-prefix=WIN
 ; RUN: llc -mtriple=i686-win32 -mcpu=nehalem < %s | FileCheck %s --check-prefix=LIN32
 ; rdar://7398554
 
 ; When doing vector gather-scatter index calculation with 32-bit indices,
 ; use an efficient mov/shift sequence rather than shuffling each individual
 ; element out of the index vector.
 
 ; CHECK-LABEL: foo:
 ; LIN: movdqa (%rsi), %xmm0
 ; LIN: pand   (%rdx), %xmm0
 ; LIN: pextrq $1, %xmm0, %r[[REG4:.+]]
-; LIN: movd   %xmm0, %r[[REG2:.+]]
+; LIN: movq   %xmm0, %r[[REG2:.+]]
 ; LIN: movslq %e[[REG2]], %r[[REG1:.+]]
 ; LIN: sarq    $32, %r[[REG2]]
 ; LIN: movslq %e[[REG4]], %r[[REG3:.+]]
 ; LIN: sarq    $32, %r[[REG4]]
 ; LIN: movsd  (%rdi,%r[[REG1]],8), %xmm0
 ; LIN: movhpd (%rdi,%r[[REG2]],8), %xmm0
 ; LIN: movsd  (%rdi,%r[[REG3]],8), %xmm1
 ; LIN: movhpd (%rdi,%r[[REG4]],8), %xmm1
 
 ; WIN: movdqa (%rdx), %xmm0
 ; WIN: pand   (%r8), %xmm0
 ; WIN: pextrq $1, %xmm0, %r[[REG4:.+]]
-; WIN: movd   %xmm0, %r[[REG2:.+]]
+; WIN: movq   %xmm0, %r[[REG2:.+]]
 ; WIN: movslq %e[[REG2]], %r[[REG1:.+]]
 ; WIN: sarq    $32, %r[[REG2]]
 ; WIN: movslq %e[[REG4]], %r[[REG3:.+]]
 ; WIN: sarq    $32, %r[[REG4]]
 ; WIN: movsd  (%rcx,%r[[REG1]],8), %xmm0
 ; WIN: movhpd (%rcx,%r[[REG2]],8), %xmm0
 ; WIN: movsd  (%rcx,%r[[REG3]],8), %xmm1
 ; WIN: movhpd (%rcx,%r[[REG4]],8), %xmm1

llvm/trunk/test/CodeGen/X86/i64-to-float.ll

 ; X64-SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
 ; X64-SSE-NEXT:    pand %xmm4, %xmm0
 ; X64-SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm2[1,1,3,3]
 ; X64-SSE-NEXT:    por %xmm0, %xmm1
 ; X64-SSE-NEXT:    movdqa %xmm1, %xmm0
 ; X64-SSE-NEXT:    pandn %xmm3, %xmm0
 ; X64-SSE-NEXT:    pand {{.*}}(%rip), %xmm1
 ; X64-SSE-NEXT:    por %xmm0, %xmm1
-; X64-SSE-NEXT:    movd %xmm1, %rax
+; X64-SSE-NEXT:    movq %xmm1, %rax
 ; X64-SSE-NEXT:    xorps %xmm0, %xmm0
 ; X64-SSE-NEXT:    cvtsi2sdq %rax, %xmm0
 ; X64-SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
-; X64-SSE-NEXT:    movd %xmm1, %rax
+; X64-SSE-NEXT:    movq %xmm1, %rax
 ; X64-SSE-NEXT:    xorps %xmm1, %xmm1
 ; X64-SSE-NEXT:    cvtsi2sdq %rax, %xmm1
 ; X64-SSE-NEXT:    unpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; X64-SSE-NEXT:    retq
 ;
 ; X64-AVX-LABEL: clamp_sitofp_2i64_2f64:
 ; X64-AVX:       # BB#0:
 ; X64-AVX-NEXT:    vmovdqa {{.*#+}} xmm1 = [18446744073709551361,18446744073709551361]

llvm/trunk/test/CodeGen/X86/isint.ll

-; RUN: llc < %s -mtriple=x86_64-pc-unknown -mattr=+sse2 | FileCheck %s
-; RUN: llc < %s -mtriple=i686-pc-unknown -mattr=+sse2 | FileCheck %s
+; RUN: llc < %s -mtriple=x86_64-pc-unknown -mattr=+sse2 | FileCheck -check-prefix=CHECK -check-prefix=CHECK64 %s
 
 ; PR19059
-; RUN: llc < %s -mtriple=i686-pc-unknown -mattr=+sse2 | FileCheck -check-prefix=CHECK32 %s
+; RUN: llc < %s -mtriple=i686-pc-unknown -mattr=+sse2 | FileCheck -check-prefix=CHECK -check-prefix=CHECK32 %s
 
 define i32 @isint_return(double %d) nounwind {
 ; CHECK-LABEL: isint_return:
 ; CHECK-NOT: xor
 ; CHECK: cvt
   %i = fptosi double %d to i32
 ; CHECK-NEXT: cvt
   %e = sitofp i32 %i to double
 ; CHECK: cmpeqsd
   %c = fcmp oeq double %d, %e
 ; CHECK32-NOT: movd {{.*}}, %r{{.*}}
 ; CHECK32-NOT: andq
-; CHECK-NEXT: movd
+; CHECK32-NEXT: movd
+; CHECK64-NEXT: movq
 ; CHECK-NEXT: andl
   %z = zext i1 %c to i32
   ret i32 %z
 }
 
 define i32 @isint_float_return(float %f) nounwind {
 ; CHECK-LABEL: isint_float_return:
 ; CHECK-NOT: xor

llvm/trunk/test/CodeGen/X86/lower-bitcast.ll

; CHECK-WIDE-NEXT:    retq
   %add = add <2 x i32> %1, %2
   %3 = bitcast <2 x i32> %add to double
   ret double %3
 }
 
 define i64 @test3(i64 %A) {
 ; CHECK-LABEL: test3:
 ; CHECK:       # BB#0:
-; CHECK-NEXT:    movd %rdi, %xmm0
+; CHECK-NEXT:    movq %rdi, %xmm0
 ; CHECK-NEXT:    addps {{.*}}(%rip), %xmm0
-; CHECK-NEXT:    movd %xmm0, %rax
+; CHECK-NEXT:    movq %xmm0, %rax
 ; CHECK-NEXT:    retq
 ;
 ; CHECK-WIDE-LABEL: test3:
 ; CHECK-WIDE:       # BB#0:
-; CHECK-WIDE-NEXT:    movd %rdi, %xmm0
+; CHECK-WIDE-NEXT:    movq %rdi, %xmm0
 ; CHECK-WIDE-NEXT:    addps {{.*}}(%rip), %xmm0
-; CHECK-WIDE-NEXT:    movd %xmm0, %rax
+; CHECK-WIDE-NEXT:    movq %xmm0, %rax
 ; CHECK-WIDE-NEXT:    retq
   %1 = bitcast i64 %A to <2 x float>
   %add = fadd <2 x float> %1, <float 3.0, float 5.0>
   %2 = bitcast <2 x float> %add to i64
   ret i64 %2
 }
 
 ; FIXME: Ideally we should be able to fold the entire body of @test4 into a
 ; single paddd instruction. This is fixed with the widening legalization.
 
 define i64 @test4(i64 %A) {
 ; CHECK-LABEL: test4:
 ; CHECK:       # BB#0:
-; CHECK-NEXT:    movd %rdi, %xmm0
+; CHECK-NEXT:    movq %rdi, %xmm0
 ; CHECK-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,1,3]
 ; CHECK-NEXT:    paddd {{.*}}(%rip), %xmm0
 ; CHECK-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; CHECK-NEXT:    movd %xmm0, %rax
+; CHECK-NEXT:    movq %xmm0, %rax
 ; CHECK-NEXT:    retq
 ;
 ; CHECK-WIDE-LABEL: test4:
 ; CHECK-WIDE:       # BB#0:
-; CHECK-WIDE-NEXT:    movd %rdi, %xmm0
+; CHECK-WIDE-NEXT:    movq %rdi, %xmm0
 ; CHECK-WIDE-NEXT:    paddd {{.*}}(%rip), %xmm0
-; CHECK-WIDE-NEXT:    movd %xmm0, %rax
+; CHECK-WIDE-NEXT:    movq %xmm0, %rax
 ; CHECK-WIDE-NEXT:    retq
   %1 = bitcast i64 %A to <2 x i32>
   %add = add <2 x i32> %1, <i32 3, i32 5>
   %2 = bitcast <2 x i32> %add to i64
   ret i64 %2
 }
 
 define double @test5(double %A) {

llvm/trunk/test/CodeGen/X86/mmx-bitcast.ll

 }
 
 define i64 @t5(i32 %a, i32 %b) nounwind readnone {
 ; CHECK-LABEL: t5:
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    movd %esi, %xmm0
 ; CHECK-NEXT:    movd %edi, %xmm1
 ; CHECK-NEXT:    punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
-; CHECK-NEXT:    movd %xmm1, %rax
+; CHECK-NEXT:    movq %xmm1, %rax
 ; CHECK-NEXT:    retq
   %v0 = insertelement <2 x i32> undef, i32 %a, i32 0
   %v1 = insertelement <2 x i32> %v0, i32 %b, i32 1
   %conv = bitcast <2 x i32> %v1 to i64
   ret i64 %conv
 }
 
 declare x86_mmx @llvm.x86.mmx.pslli.q(x86_mmx, i32)

llvm/trunk/test/CodeGen/X86/mmx-cvt.ll

 ; X86-NEXT:    popl %ebp
 ; X86-NEXT:    retl
 ;
 ; X64-LABEL: cvt_v2i32_v2f32:
 ; X64:       # BB#0:
 ; X64-NEXT:    movq (%rdi), %mm0
 ; X64-NEXT:    paddd %mm0, %mm0
 ; X64-NEXT:    movd %mm0, %rax
-; X64-NEXT:    movd %rax, %xmm0
+; X64-NEXT:    movq %rax, %xmm0
 ; X64-NEXT:    cvtdq2ps %xmm0, %xmm0
 ; X64-NEXT:    retq
   %2 = bitcast <1 x i64>* %0 to x86_mmx*
   %3 = load x86_mmx, x86_mmx* %2, align 8
   %4 = tail call x86_mmx @llvm.x86.mmx.padd.d(x86_mmx %3, x86_mmx %3)
   %5 = bitcast x86_mmx %4 to i64
   %6 = insertelement <2 x i64> undef, i64 %5, i32 0
   %7 = insertelement <2 x i64> %6, i64 0, i32 1

llvm/trunk/test/CodeGen/X86/mod128.ll

define i64 @mod128(i128 %x) {
   ; WIN64-NOT: xorl %r9d, %r9d
   ; WIN64-DAG: movq %rdx, 56(%rsp)
   ; WIN64-DAG: movq %rcx, 48(%rsp)
   ; WIN64-DAG: leaq 48(%rsp), %rcx
   ; WIN64-DAG: leaq 32(%rsp), %rdx
   ; WIN64-DAG: movq $0, 40(%rsp)
   ; WIN64-DAG: movq $3, 32(%rsp)
   ; WIN64: callq   __modti3
-  ; WIN64: movd    %xmm0, %rax
+  ; WIN64: movq    %xmm0, %rax
 
   %1 = srem i128 %x, 3
   %2 = trunc i128 %1 to i64
   ret i64 %2
 }

llvm/trunk/test/CodeGen/X86/movmsk.ll

 }
 
 ; PR11570
 ; FIXME: This should also use movmskps; we don't form the FGETSIGN node
 ; in this case, though.
 define void @float_call_signbit(double %n) {
 ; CHECK-LABEL: float_call_signbit:
 ; CHECK:       ## BB#0: ## %entry
-; CHECK-NEXT:    movd %xmm0, %rdi
+; CHECK-NEXT:    movq %xmm0, %rdi
 ; CHECK-NEXT:    shrq $63, %rdi
 ; CHECK-NEXT:    ## kill: %EDI<def> %EDI<kill> %RDI<kill>
 ; CHECK-NEXT:    jmp _float_call_signbit_callee ## TAILCALL
 entry:
   %t0 = bitcast double %n to i64
   %tobool.i.i.i.i = icmp slt i64 %t0, 0
   tail call void @float_call_signbit_callee(i1 zeroext %tobool.i.i.i.i)
   ret void

llvm/trunk/test/CodeGen/X86/nontemporal-2.ll

; VLX-NEXT:    retq
   store double %1, double* %dst, align 1, !nontemporal !1
   ret void
 }
 
 define void @test_extract_i64(<2 x i64> %arg, i64* %dst) {
 ; SSE2-LABEL: test_extract_i64:
 ; SSE2:       # BB#0:
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    movq %xmm0, %rax
 ; SSE2-NEXT:    movntiq %rax, (%rdi)
 ; SSE2-NEXT:    retq
 ;
 ; SSE4A-LABEL: test_extract_i64:
 ; SSE4A:       # BB#0:
 ; SSE4A-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE4A-NEXT:    movd %xmm0, %rax
+; SSE4A-NEXT:    movq %xmm0, %rax
 ; SSE4A-NEXT:    movntiq %rax, (%rdi)
 ; SSE4A-NEXT:    retq
 ;
 ; SSE41-LABEL: test_extract_i64:
 ; SSE41:       # BB#0:
 ; SSE41-NEXT:    pextrq $1, %xmm0, %rax
 ; SSE41-NEXT:    movntiq %rax, (%rdi)
 ; SSE41-NEXT:    retq

llvm/trunk/test/CodeGen/X86/pr18344.ll

 ; X86-NEXT:    popl %edi
 ; X86-NEXT:    popl %ebx
 ; X86-NEXT:    retl
 ;
 ; X64-LABEL: FFT:
 ; X64:       # BB#0: # %begin
 ; X64-NEXT:    movdqu (%rdx), %xmm0
 ; X64-NEXT:    pslld $4, %xmm0
-; X64-NEXT:    movd %xmm0, %rax
+; X64-NEXT:    movq %xmm0, %rax
 ; X64-NEXT:    movslq %eax, %r8
 ; X64-NEXT:    sarq $32, %rax
 ; X64-NEXT:    pextrq $1, %xmm0, %rdx
 ; X64-NEXT:    movslq %edx, %rcx
 ; X64-NEXT:    sarq $32, %rdx
 ; X64-NEXT:    movss {{.*#+}} xmm0 = mem[0],zero,zero,zero
 ; X64-NEXT:    movss {{.*#+}} xmm1 = mem[0],zero,zero,zero
 ; X64-NEXT:    movss {{.*#+}} xmm2 = mem[0],zero,zero,zero

llvm/trunk/test/CodeGen/X86/pr21792.ll

 ; CHECK:       # BB#0: # %entry
 ; CHECK-NEXT:    pushq %rax
 ; CHECK-NEXT:  .Lcfi0:
 ; CHECK-NEXT:    .cfi_def_cfa_offset 16
 ; CHECK-NEXT:    pand {{.*}}(%rip), %xmm0
 ; CHECK-NEXT:    pextrq $1, %xmm0, %rdx
 ; CHECK-NEXT:    movq %rdx, %rcx
 ; CHECK-NEXT:    shrq $32, %rcx
-; CHECK-NEXT:    movd %xmm0, %rax
+; CHECK-NEXT:    movq %xmm0, %rax
 ; CHECK-NEXT:    movq %rax, %r9
 ; CHECK-NEXT:    shrq $32, %r9
 ; CHECK-NEXT:    andl $2032, %eax # imm = 0x7F0
 ; CHECK-NEXT:    leaq stuff(%rax), %rdi
 ; CHECK-NEXT:    leaq stuff(%r9), %rsi
 ; CHECK-NEXT:    andl $2032, %edx # imm = 0x7F0
 ; CHECK-NEXT:    leaq stuff(%rdx), %rdx
 ; CHECK-NEXT:    leaq stuff(%rcx), %rcx

llvm/trunk/test/CodeGen/X86/pr30511.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-pc-linux-gnu | FileCheck %s
 
 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
 target triple = "x86_64-pc-linux-gnu"
 
 define i64 @PR30511(<2 x double> %a) {
 ; CHECK-LABEL: PR30511:
 ; CHECK:       # BB#0:
 ; CHECK-NEXT:    addpd {{.*}}(%rip), %xmm0
 ; CHECK-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2,2,3]
 ; CHECK-NEXT:    cvtdq2pd %xmm0, %xmm0
 ; CHECK-NEXT:    mulpd {{.*}}(%rip), %xmm0
-; CHECK-NEXT:    movd %xmm0, %rax
+; CHECK-NEXT:    movq %xmm0, %rax
 ; CHECK-NEXT:    retq
   %1 = fadd <2 x double> %a, <double 0x4338000000000000, double 0x4338000000000000>
   %2 = bitcast <2 x double> %1 to <2 x i64>
   %3 = trunc <2 x i64> %2 to <2 x i32>
   %4 = sitofp <2 x i32> %3 to <2 x double>
   %5 = fmul <2 x double> %4, <double 0x3E76800000000000, double 0x3E76800000000000>
   %6 = extractelement <2 x double> %5, i32 0
   %7 = bitcast double %6 to i64
   ret i64 %7
 }
 

llvm/trunk/test/CodeGen/X86/pshufb-mask-comments.ll

; CHECK-NEXT:    retq
   %4 = tail call <16 x i8> @llvm.x86.ssse3.pshuf.b.128(<16 x i8> %V, <16 x i8> %3)
   ret <16 x i8> %4
 }
 
 define <16 x i8> @test5(<16 x i8> %V) {
 ; CHECK-LABEL: test5:
 ; CHECK:       # BB#0:
 ; CHECK-NEXT:    movl $1, %eax
-; CHECK-NEXT:    movd %rax, %xmm1
+; CHECK-NEXT:    movq %rax, %xmm1
 ; CHECK-NEXT:    movdqa %xmm1, (%rax)
 ; CHECK-NEXT:    movdqa {{.*#+}} xmm1 = [1,1]
 ; CHECK-NEXT:    movdqa %xmm1, (%rax)
 ; CHECK-NEXT:    pshufb %xmm1, %xmm0
 ; CHECK-NEXT:    retq
   store <2 x i64> <i64 1, i64 0>, <2 x i64>* undef, align 16
   %l = load <2 x i64>, <2 x i64>* undef, align 16
   %shuffle = shufflevector <2 x i64> %l, <2 x i64> undef, <2 x i32> zeroinitializer

llvm/trunk/test/CodeGen/X86/ret-mmx.ll

 ; CHECK-NEXT:    retq
   ret <1 x i64> <i64 1>
 }
 
 define <2 x i32> @t3() nounwind {
 ; CHECK-LABEL: t3:
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    movl $1, %eax
-; CHECK-NEXT:    movd %rax, %xmm0
+; CHECK-NEXT:    movq %rax, %xmm0
 ; CHECK-NEXT:    retq
   ret <2 x i32> <i32 1, i32 0>
 }
 
 define double @t4() nounwind {
 ; CHECK-LABEL: t4:
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    movl $1, %eax
 ; CHECK-NEXT:    movd %eax, %xmm0
 ; CHECK-NEXT:    retq
   ret double bitcast (<2 x i32> <i32 1, i32 0> to double)
 }
 

llvm/trunk/test/CodeGen/X86/sad_variations.ll

 }
 
 define i64 @sad8_64bit_icmp_sext_slt(i8* nocapture readonly %cur, i8* nocapture readonly %ref, i64 %stride) local_unnamed_addr #4 {
 ; SSE2-LABEL: sad8_64bit_icmp_sext_slt:
 ; SSE2:       # BB#0: # %entry
 ; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
 ; SSE2-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
 ; SSE2-NEXT:    psadbw %xmm0, %xmm1
-; SSE2-NEXT:    movd %xmm1, %rax
+; SSE2-NEXT:    movq %xmm1, %rax
 ; SSE2-NEXT:    retq
 ;
 ; AVX2-LABEL: sad8_64bit_icmp_sext_slt:
 ; AVX2:       # BB#0: # %entry
 ; AVX2-NEXT:    vmovq {{.*#+}} xmm0 = mem[0],zero
 ; AVX2-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
 ; AVX2-NEXT:    vpsadbw %xmm0, %xmm1, %xmm0
 ; AVX2-NEXT:    vmovq %xmm0, %rax

 }
 
 define i64 @sad8_64bit_icmp_zext_slt(i8* nocapture readonly %cur, i8* nocapture readonly %ref, i64 %stride) local_unnamed_addr #4 {
 ; SSE2-LABEL: sad8_64bit_icmp_zext_slt:
 ; SSE2:       # BB#0: # %entry
 ; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
 ; SSE2-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
 ; SSE2-NEXT:    psadbw %xmm0, %xmm1
-; SSE2-NEXT:    movd %xmm1, %rax
+; SSE2-NEXT:    movq %xmm1, %rax
 ; SSE2-NEXT:    retq
 ;
 ; AVX2-LABEL: sad8_64bit_icmp_zext_slt:
 ; AVX2:       # BB#0: # %entry
 ; AVX2-NEXT:    vmovq {{.*#+}} xmm0 = mem[0],zero
 ; AVX2-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
 ; AVX2-NEXT:    vpsadbw %xmm0, %xmm1, %xmm0
 ; AVX2-NEXT:    vmovq %xmm0, %rax

 }
 
 define i64 @sad8_early_64bit_icmp_zext_slt(i8* nocapture readonly %cur, i8* nocapture readonly %ref, i64 %stride) local_unnamed_addr #4 {
 ; SSE2-LABEL: sad8_early_64bit_icmp_zext_slt:
 ; SSE2:       # BB#0: # %entry
 ; SSE2-NEXT:    movq {{.*#+}} xmm0 = mem[0],zero
 ; SSE2-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
 ; SSE2-NEXT:    psadbw %xmm0, %xmm1
-; SSE2-NEXT:    movd %xmm1, %rax
+; SSE2-NEXT:    movq %xmm1, %rax
 ; SSE2-NEXT:    retq
 ;
 ; AVX2-LABEL: sad8_early_64bit_icmp_zext_slt:
 ; AVX2:       # BB#0: # %entry
 ; AVX2-NEXT:    vmovq {{.*#+}} xmm0 = mem[0],zero
 ; AVX2-NEXT:    vmovq {{.*#+}} xmm1 = mem[0],zero
 ; AVX2-NEXT:    vpsadbw %xmm0, %xmm1, %xmm0
 ; AVX2-NEXT:    vmovq %xmm0, %rax

llvm/trunk/test/CodeGen/X86/scalar-int-to-fp.ll

 ; SSE2_32-NEXT:    movlpd %xmm1, (%esp)
 ; SSE2_32-NEXT:    fldl (%esp)
 ; SSE2_32-NEXT:    movl %ebp, %esp
 ; SSE2_32-NEXT:    popl %ebp
 ; SSE2_32-NEXT:    retl
 ;
 ; SSE2_64-LABEL: u64_to_d:
 ; SSE2_64:       # BB#0:
-; SSE2_64-NEXT:    movd %rdi, %xmm1
+; SSE2_64-NEXT:    movq %rdi, %xmm1
 ; SSE2_64-NEXT:    punpckldq {{.*#+}} xmm1 = xmm1[0],mem[0],xmm1[1],mem[1]
 ; SSE2_64-NEXT:    subpd {{.*}}(%rip), %xmm1
 ; SSE2_64-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[2,3,0,1]
 ; SSE2_64-NEXT:    addpd %xmm1, %xmm0
 ; SSE2_64-NEXT:    retq
 ;
 ; X87-LABEL: u64_to_d:
 ; X87:       # BB#0:

llvm/trunk/test/CodeGen/X86/setcc-wide-types.ll

; AVX2-NEXT:    retq
   %zext = zext i1 %cmp to i32
   ret i32 %zext
 }
 
 define i32 @ne_i256(<4 x i64> %x, <4 x i64> %y) {
 ; SSE2-LABEL: ne_i256:
 ; SSE2:       # BB#0:
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm4 = xmm0[2,3,0,1]
-; SSE2-NEXT:    movd %xmm4, %r8
+; SSE2-NEXT:    movq %xmm4, %r8
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm4 = xmm1[2,3,0,1]
-; SSE2-NEXT:    movd %xmm4, %r9
-; SSE2-NEXT:    movd %xmm0, %r10
-; SSE2-NEXT:    movd %xmm1, %rsi
+; SSE2-NEXT:    movq %xmm4, %r9
+; SSE2-NEXT:    movq %xmm0, %r10
+; SSE2-NEXT:    movq %xmm1, %rsi
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm2[2,3,0,1]
-; SSE2-NEXT:    movd %xmm0, %rdi
+; SSE2-NEXT:    movq %xmm0, %rdi
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm3[2,3,0,1]
-; SSE2-NEXT:    movd %xmm0, %rax
-; SSE2-NEXT:    movd %xmm2, %rcx
-; SSE2-NEXT:    movd %xmm3, %rdx
+; SSE2-NEXT:    movq %xmm0, %rax
+; SSE2-NEXT:    movq %xmm2, %rcx
+; SSE2-NEXT:    movq %xmm3, %rdx
 ; SSE2-NEXT:    xorq %rsi, %rdx
 ; SSE2-NEXT:    xorq %r10, %rcx
 ; SSE2-NEXT:    orq %rdx, %rcx
 ; SSE2-NEXT:    xorq %r9, %rax
 ; SSE2-NEXT:    xorq %r8, %rdi
 ; SSE2-NEXT:    orq %rax, %rdi
 ; SSE2-NEXT:    xorl %eax, %eax
 ; SSE2-NEXT:    orq %rcx, %rdi
; AVX2-NEXT:    retq
   %zext = zext i1 %cmp to i32
   ret i32 %zext
 }
 
 define i32 @eq_i256(<4 x i64> %x, <4 x i64> %y) {
 ; SSE2-LABEL: eq_i256:
 ; SSE2:       # BB#0:
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm4 = xmm0[2,3,0,1]
-; SSE2-NEXT:    movd %xmm4, %r8
+; SSE2-NEXT:    movq %xmm4, %r8
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm4 = xmm1[2,3,0,1]
-; SSE2-NEXT:    movd %xmm4, %r9
-; SSE2-NEXT:    movd %xmm0, %r10
-; SSE2-NEXT:    movd %xmm1, %rsi
+; SSE2-NEXT:    movq %xmm4, %r9
+; SSE2-NEXT:    movq %xmm0, %r10
+; SSE2-NEXT:    movq %xmm1, %rsi
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm2[2,3,0,1]
-; SSE2-NEXT:    movd %xmm0, %rdi
+; SSE2-NEXT:    movq %xmm0, %rdi
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm3[2,3,0,1]
-; SSE2-NEXT:    movd %xmm0, %rax
-; SSE2-NEXT:    movd %xmm2, %rcx
-; SSE2-NEXT:    movd %xmm3, %rdx
+; SSE2-NEXT:    movq %xmm0, %rax
+; SSE2-NEXT:    movq %xmm2, %rcx
+; SSE2-NEXT:    movq %xmm3, %rdx
 ; SSE2-NEXT:    xorq %rsi, %rdx
 ; SSE2-NEXT:    xorq %r10, %rcx
 ; SSE2-NEXT:    orq %rdx, %rcx
 ; SSE2-NEXT:    xorq %r9, %rax
 ; SSE2-NEXT:    xorq %r8, %rdi
 ; SSE2-NEXT:    orq %rax, %rdi
 ; SSE2-NEXT:    xorl %eax, %eax
 ; SSE2-NEXT:    orq %rcx, %rdi

llvm/trunk/test/CodeGen/X86/shrink_vmul.ll

 ; CHECK-LABEL: mul_2xi16_varconst3:
 ; CHECK:       # BB#0: # %entry
 ; CHECK-NEXT:    movq {{.*}}(%rip), %rax
 ; CHECK-NEXT:    movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
 ; CHECK-NEXT:    pxor %xmm1, %xmm1
 ; CHECK-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
 ; CHECK-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,1,3]
 ; CHECK-NEXT:    movl $65536, %ecx # imm = 0x10000
-; CHECK-NEXT:    movd %rcx, %xmm1
+; CHECK-NEXT:    movq %rcx, %xmm1
 ; CHECK-NEXT:    pslldq {{.*#+}} xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0,1,2,3,4,5,6,7]
 ; CHECK-NEXT:    movdqa %xmm0, %xmm2
 ; CHECK-NEXT:    pmuludq %xmm1, %xmm2
 ; CHECK-NEXT:    psrlq $32, %xmm0
 ; CHECK-NEXT:    pmuludq %xmm1, %xmm0
 ; CHECK-NEXT:    psllq $32, %xmm0
 ; CHECK-NEXT:    paddq %xmm2, %xmm0
 ; CHECK-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]

 ; CHECK-LABEL: mul_2xi16_varconst4:
 ; CHECK:       # BB#0: # %entry
 ; CHECK-NEXT:    movq {{.*}}(%rip), %rax
 ; CHECK-NEXT:    movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
 ; CHECK-NEXT:    pshuflw {{.*#+}} xmm0 = xmm0[0,0,2,1,4,5,6,7]
 ; CHECK-NEXT:    psrad $16, %xmm0
 ; CHECK-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,1,3]
 ; CHECK-NEXT:    movl $32768, %ecx # imm = 0x8000
-; CHECK-NEXT:    movd %rcx, %xmm1
+; CHECK-NEXT:    movq %rcx, %xmm1
 ; CHECK-NEXT:    pslldq {{.*#+}} xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0,1,2,3,4,5,6,7]
 ; CHECK-NEXT:    movdqa %xmm0, %xmm2
 ; CHECK-NEXT:    pmuludq %xmm1, %xmm2
 ; CHECK-NEXT:    psrlq $32, %xmm0
 ; CHECK-NEXT:    pmuludq %xmm1, %xmm0
 ; CHECK-NEXT:    psllq $32, %xmm0
 ; CHECK-NEXT:    paddq %xmm2, %xmm0
 ; CHECK-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]

llvm/trunk/test/CodeGen/X86/sse2-intrinsics-fast-isel-x86_64.ll

; X64-NEXT:    retq
   %res = call i64 @llvm.x86.sse2.cvtsd2si64(<2 x double> %a0)
   ret i64 %res
 }
 declare i64 @llvm.x86.sse2.cvtsd2si64(<2 x double>) nounwind readnone
 
 define i64 @test_mm_cvtsi128_si64(<2 x i64> %a0) nounwind {
 ; X64-LABEL: test_mm_cvtsi128_si64:
 ; X64:       # BB#0:
-; X64-NEXT:    movd %xmm0, %rax
+; X64-NEXT:    movq %xmm0, %rax
 ; X64-NEXT:    retq
   %res = extractelement <2 x i64> %a0, i32 0
   ret i64 %res
 }
 
 define <2 x double> @test_mm_cvtsi64_sd(<2 x double> %a0, i64 %a1) nounwind {
 ; X64-LABEL: test_mm_cvtsi64_sd:
 ; X64:       # BB#0:
 ; X64-NEXT:    cvtsi2sdq %rdi, %xmm0
 ; X64-NEXT:    retq
   %res = call <2 x double> @llvm.x86.sse2.cvtsi642sd(<2 x double> %a0, i64 %a1)
   ret <2 x double> %res
 }
 declare <2 x double> @llvm.x86.sse2.cvtsi642sd(<2 x double>, i64) nounwind readnone
 
 define <2 x i64> @test_mm_cvtsi64_si128(i64 %a0) nounwind {
 ; X64-LABEL: test_mm_cvtsi64_si128:
 ; X64:       # BB#0:
-; X64-NEXT:    movd %rdi, %xmm0
+; X64-NEXT:    movq %rdi, %xmm0
 ; X64-NEXT:    retq
   %res0 = insertelement <2 x i64> undef, i64 %a0, i32 0
   %res1 = insertelement <2 x i64> %res0, i64 0, i32 1
   ret <2 x i64> %res1
 }
 
 define i64 @test_mm_cvttsd_si64(<2 x double> %a0) nounwind {
 ; X64-LABEL: test_mm_cvttsd_si64:

llvm/trunk/test/CodeGen/X86/sse2-intrinsics-fast-isel.ll

 ; X32-NEXT:    movd {{.*#+}} xmm1 = mem[0],zero,zero,zero
 ; X32-NEXT:    movd {{.*#+}} xmm2 = mem[0],zero,zero,zero
 ; X32-NEXT:    punpckldq {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
 ; X32-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: test_mm_set_epi64x:
 ; X64:       # BB#0:
-; X64-NEXT:    movd %rdi, %xmm1
-; X64-NEXT:    movd %rsi, %xmm0
+; X64-NEXT:    movq %rdi, %xmm1
+; X64-NEXT:    movq %rsi, %xmm0
 ; X64-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; X64-NEXT:    retq
   %res0  = insertelement <2 x i64> undef, i64 %a1, i32 0
   %res1  = insertelement <2 x i64> %res0, i64 %a0, i32 1
   ret <2 x i64> %res1
 }
 
 define <2 x double> @test_mm_set_pd(double %a0, double %a1) nounwind {

 ; X32-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,0,1,1]
 ; X32-NEXT:    movd {{.*#+}} xmm1 = mem[0],zero,zero,zero
 ; X32-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[0,0,1,1]
 ; X32-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: test_mm_set1_epi64x:
 ; X64:       # BB#0:
-; X64-NEXT:    movd %rdi, %xmm0
+; X64-NEXT:    movq %rdi, %xmm0
 ; X64-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]
 ; X64-NEXT:    retq
   %res0  = insertelement <2 x i64> undef, i64 %a0, i32 0
   %res1  = insertelement <2 x i64> %res0, i64 %a0, i32 1
   ret <2 x i64> %res1
 }
 
 define <2 x double> @test_mm_set1_pd(double %a0) nounwind {

 ; X32-NEXT:    movd {{.*#+}} xmm1 = mem[0],zero,zero,zero
 ; X32-NEXT:    movd {{.*#+}} xmm2 = mem[0],zero,zero,zero
 ; X32-NEXT:    punpckldq {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
 ; X32-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: test_mm_setr_epi64x:
 ; X64:       # BB#0:
-; X64-NEXT:    movd %rsi, %xmm1
-; X64-NEXT:    movd %rdi, %xmm0
+; X64-NEXT:    movq %rsi, %xmm1
+; X64-NEXT:    movq %rdi, %xmm0
 ; X64-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; X64-NEXT:    retq
   %res0  = insertelement <2 x i64> undef, i64 %a0, i32 0
   %res1  = insertelement <2 x i64> %res0, i64 %a1, i32 1
   ret <2 x i64> %res1
 }
 
 define <2 x double> @test_mm_setr_pd(double %a0, double %a1) nounwind {

 ; X32-LABEL: test_mm_storel_epi64:
 ; X32:       # BB#0:
 ; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
 ; X32-NEXT:    movlps %xmm0, (%eax)
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: test_mm_storel_epi64:
 ; X64:       # BB#0:
-; X64-NEXT:    movd %xmm0, %rax
+; X64-NEXT:    movq %xmm0, %rax
 ; X64-NEXT:    movq %rax, (%rdi)
 ; X64-NEXT:    retq
   %ext = extractelement <2 x i64> %a1, i32 0
   %bc = bitcast <2 x i64> *%a0 to i64*
   store i64 %ext, i64* %bc, align 8
   ret void
 }
 

llvm/trunk/test/CodeGen/X86/sse2-schedule.ll

; BTVER2-NEXT:    retq # sched: [4:1.00]
   %7 = extractelement <4 x i32> %5, i32 0
   store i32 %6, i32* %a2
   ret i32 %7
 }
 
 define i64 @test_movd_64(<2 x i64> %a0, i64 %a1, i64 *%a2) {
 ; GENERIC-LABEL: test_movd_64:
 ; GENERIC:       # BB#0:
-; GENERIC-NEXT:    movd %rdi, %xmm1
+; GENERIC-NEXT:    movq %rdi, %xmm1
 ; GENERIC-NEXT:    movq {{.*#+}} xmm2 = mem[0],zero
 ; GENERIC-NEXT:    paddq %xmm0, %xmm1
 ; GENERIC-NEXT:    paddq %xmm0, %xmm2
-; GENERIC-NEXT:    movd %xmm2, %rax
+; GENERIC-NEXT:    movq %xmm2, %rax
 ; GENERIC-NEXT:    movq %xmm1, (%rsi)
 ; GENERIC-NEXT:    retq
 ;
 ; ATOM-LABEL: test_movd_64:
 ; ATOM:       # BB#0:
 ; ATOM-NEXT:    movq {{.*#+}} xmm1 = mem[0],zero
-; ATOM-NEXT:    movd %rdi, %xmm2
+; ATOM-NEXT:    movq %rdi, %xmm2
 ; ATOM-NEXT:    paddq %xmm0, %xmm2
 ; ATOM-NEXT:    paddq %xmm0, %xmm1
 ; ATOM-NEXT:    movq %xmm2, (%rsi)
-; ATOM-NEXT:    movd %xmm1, %rax
+; ATOM-NEXT:    movq %xmm1, %rax
 ; ATOM-NEXT:    retq
 ;
 ; SLM-LABEL: test_movd_64:
 ; SLM:       # BB#0:
 ; SLM-NEXT:    movq {{.*#+}} xmm2 = mem[0],zero sched: [3:1.00]
-; SLM-NEXT:    movd %rdi, %xmm1 # sched: [1:0.50]
+; SLM-NEXT:    movq %rdi, %xmm1 # sched: [1:0.50]
 ; SLM-NEXT:    paddq %xmm0, %xmm1 # sched: [1:0.50]
 ; SLM-NEXT:    movq %xmm1, (%rsi) # sched: [1:1.00]
 ; SLM-NEXT:    paddq %xmm0, %xmm2 # sched: [1:0.50]
-; SLM-NEXT:    movd %xmm2, %rax # sched: [1:0.50]
+; SLM-NEXT:    movq %xmm2, %rax # sched: [1:0.50]
 ; SLM-NEXT:    retq # sched: [4:1.00]
 ;
 ; SANDY-LABEL: test_movd_64:
 ; SANDY:       # BB#0:
 ; SANDY-NEXT:    vmovq %rdi, %xmm1 # sched: [1:0.33]
 ; SANDY-NEXT:    vmovq {{.*#+}} xmm2 = mem[0],zero sched: [4:0.50]
 ; SANDY-NEXT:    vpaddq %xmm1, %xmm0, %xmm1 # sched: [1:0.50]
 ; SANDY-NEXT:    vpaddq %xmm2, %xmm0, %xmm0 # sched: [1:0.50]

llvm/trunk/test/CodeGen/X86/statepoint-vector.ll

; CHECK: addq $24, %rsp
   ret <2 x i8 addrspace(1)*> %obj.relocated
 }
 
 ; Can we lower the base, derived pairs if both are vectors?
 define <2 x i8 addrspace(1)*> @test2(<2 x i8 addrspace(1)*> %obj, i64 %offset) gc "statepoint-example" {
 entry:
 ; CHECK-LABEL: @test2
 ; CHECK: subq $40, %rsp
-; CHECK: movd %rdi, %xmm1
+; CHECK: movq %rdi, %xmm1
 ; CHECK: pshufd $68, %xmm1, %xmm1       # xmm1 = xmm1[0,1,0,1]
 ; CHECK: paddq  %xmm0, %xmm1
 ; CHECK: movdqa %xmm0, 16(%rsp)
 ; CHECK: movdqa %xmm1, (%rsp)
 ; CHECK: callq  do_safepoint
 ; CHECK: movaps (%rsp), %xmm0
 ; CHECK: addq $40, %rsp
   %derived = getelementptr i8, <2 x i8 addrspace(1)*> %obj, i64 %offset

llvm/trunk/test/CodeGen/X86/vec_fneg.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=i686-unknown-unknown -mattr=+sse  | FileCheck %s --check-prefix=X32-SSE --check-prefix=X32-SSE1
 ; RUN: llc < %s -mtriple=i686-unknown-unknown -mattr=+sse2 | FileCheck %s --check-prefix=X32-SSE --check-prefix=X32-SSE2
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=-sse2 | FileCheck %s --check-prefix=X64-SSE --check-prefix=X64-SSE1
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse2 | FileCheck %s --check-prefix=X64-SSE --check-prefix=X64-SSE2
 
 ; FNEG is defined as subtraction from -0.0.
 
 ; This test verifies that we use an xor with a constant to flip the sign bits; no subtraction needed.
 define <4 x float> @t1(<4 x float> %Q) nounwind {
 ; X32-SSE-LABEL: t1:
 ; X32-SSE:       # BB#0:
-; X32-SSE-NEXT:    xorps .LCPI0_0, %xmm0
+; X32-SSE-NEXT:    xorps {{\.LCPI.*}}, %xmm0
 ; X32-SSE-NEXT:    retl
 ;
 ; X64-SSE-LABEL: t1:
 ; X64-SSE:       # BB#0:
 ; X64-SSE-NEXT:    xorps {{.*}}(%rip), %xmm0
 ; X64-SSE-NEXT:    retq
   %tmp = fsub <4 x float> < float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00 >, %Q
   ret <4 x float> %tmp

 ; X64-SSE1-NEXT:    movq %rax, -{{[0-9]+}}(%rsp)
 ; X64-SSE1-NEXT:    movaps -{{[0-9]+}}(%rsp), %xmm0
 ; X64-SSE1-NEXT:    retq
 ;
 ; X64-SSE2-LABEL: fneg_bitcast:
 ; X64-SSE2:       # BB#0:
 ; X64-SSE2-NEXT:    movabsq $-9223372034707292160, %rax # imm = 0x8000000080000000
 ; X64-SSE2-NEXT:    xorq %rdi, %rax
-; X64-SSE2-NEXT:    movd %rax, %xmm0
+; X64-SSE2-NEXT:    movq %rax, %xmm0
 ; X64-SSE2-NEXT:    retq
   %bitcast = bitcast i64 %i to <2 x float>
   %fneg = fsub <2 x float> <float -0.0, float -0.0>, %bitcast
   ret <2 x float> %fneg
 }

llvm/trunk/test/CodeGen/X86/vec_fp_to_int.ll

 ;
 ; Double to Signed Integer
 ;
 
 define <2 x i64> @fptosi_2f64_to_2i64(<2 x double> %a) {
 ; SSE-LABEL: fptosi_2f64_to_2i64:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    cvttsd2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm1
+; SSE-NEXT:    movq %rax, %xmm1
 ; SSE-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; SSE-NEXT:    cvttsd2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE-NEXT:    movdqa %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: fptosi_2f64_to_2i64:
 ; VEX:       # BB#0:
 ; VEX-NEXT:    vcvttsd2si %xmm0, %rax
 ; VEX-NEXT:    vmovq %rax, %xmm1
; AVX-NEXT:    retq
   %cvt = fptosi <4 x double> %ext to <4 x i32>
   ret <4 x i32> %cvt
 }
 
 define <4 x i64> @fptosi_4f64_to_4i64(<4 x double> %a) {
 ; SSE-LABEL: fptosi_4f64_to_4i64:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    cvttsd2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm2
+; SSE-NEXT:    movq %rax, %xmm2
 ; SSE-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; SSE-NEXT:    cvttsd2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm0[0]
 ; SSE-NEXT:    cvttsd2si %xmm1, %rax
-; SSE-NEXT:    movd %rax, %xmm3
+; SSE-NEXT:    movq %rax, %xmm3
 ; SSE-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
 ; SSE-NEXT:    cvttsd2si %xmm1, %rax
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm3 = xmm3[0],xmm0[0]
 ; SSE-NEXT:    movdqa %xmm2, %xmm0
 ; SSE-NEXT:    movdqa %xmm3, %xmm1
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: fptosi_4f64_to_4i64:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1

 ; SSE-NEXT:    movapd %xmm0, %xmm1
 ; SSE-NEXT:    subsd %xmm2, %xmm1
 ; SSE-NEXT:    cvttsd2si %xmm1, %rax
 ; SSE-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttsd2si %xmm0, %rdx
 ; SSE-NEXT:    ucomisd %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rdx
-; SSE-NEXT:    movd %rdx, %xmm1
+; SSE-NEXT:    movq %rdx, %xmm1
 ; SSE-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; SSE-NEXT:    movaps %xmm0, %xmm3
 ; SSE-NEXT:    subsd %xmm2, %xmm3
 ; SSE-NEXT:    cvttsd2si %xmm3, %rax
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttsd2si %xmm0, %rcx
 ; SSE-NEXT:    ucomisd %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rcx
-; SSE-NEXT:    movd %rcx, %xmm0
+; SSE-NEXT:    movq %rcx, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE-NEXT:    movdqa %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: fptoui_2f64_to_2i64:
 ; VEX:       # BB#0:
 ; VEX-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
 ; VEX-NEXT:    vsubsd %xmm1, %xmm0, %xmm2

 ; SSE-NEXT:    movapd %xmm0, %xmm1
 ; SSE-NEXT:    subsd %xmm2, %xmm1
 ; SSE-NEXT:    cvttsd2si %xmm1, %rax
 ; SSE-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttsd2si %xmm0, %rdx
 ; SSE-NEXT:    ucomisd %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rdx
-; SSE-NEXT:    movd %rdx, %xmm1
+; SSE-NEXT:    movq %rdx, %xmm1
 ; SSE-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; SSE-NEXT:    movaps %xmm0, %xmm3
 ; SSE-NEXT:    subsd %xmm2, %xmm3
 ; SSE-NEXT:    cvttsd2si %xmm3, %rax
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttsd2si %xmm0, %rcx
 ; SSE-NEXT:    ucomisd %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rcx
-; SSE-NEXT:    movd %rcx, %xmm0
+; SSE-NEXT:    movq %rcx, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE-NEXT:    pxor %xmm0, %xmm0
 ; SSE-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,2],xmm0[2,3]
 ; SSE-NEXT:    movaps %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: fptoui_2f64_to_4i32:
 ; VEX:       # BB#0:

 ; SSE-NEXT:    movapd %xmm0, %xmm2
 ; SSE-NEXT:    subsd %xmm1, %xmm2
 ; SSE-NEXT:    cvttsd2si %xmm2, %rax
 ; SSE-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttsd2si %xmm0, %rdx
 ; SSE-NEXT:    ucomisd %xmm1, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rdx
-; SSE-NEXT:    movd %rdx, %xmm2
+; SSE-NEXT:    movq %rdx, %xmm2
 ; SSE-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; SSE-NEXT:    movaps %xmm0, %xmm3
 ; SSE-NEXT:    subsd %xmm1, %xmm3
 ; SSE-NEXT:    cvttsd2si %xmm3, %rax
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttsd2si %xmm0, %rcx
 ; SSE-NEXT:    ucomisd %xmm1, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rcx
-; SSE-NEXT:    movd %rcx, %xmm0
+; SSE-NEXT:    movq %rcx, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm0[0]
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm2[0,2,2,3]
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: fptoui_2f64_to_2i32:
 ; VEX:       # BB#0:
 ; VEX-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
 ; VEX-NEXT:    vsubsd %xmm1, %xmm0, %xmm2

 ; SSE-NEXT:    movapd %xmm0, %xmm1
 ; SSE-NEXT:    subsd %xmm2, %xmm1
 ; SSE-NEXT:    cvttsd2si %xmm1, %rax
 ; SSE-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttsd2si %xmm0, %rdx
 ; SSE-NEXT:    ucomisd %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rdx
-; SSE-NEXT:    movd %rdx, %xmm1
+; SSE-NEXT:    movq %rdx, %xmm1
 ; SSE-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; SSE-NEXT:    movaps %xmm0, %xmm3
 ; SSE-NEXT:    subsd %xmm2, %xmm3
 ; SSE-NEXT:    cvttsd2si %xmm3, %rax
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttsd2si %xmm0, %rdx
 ; SSE-NEXT:    ucomisd %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rdx
-; SSE-NEXT:    movd %rdx, %xmm0
+; SSE-NEXT:    movq %rdx, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE-NEXT:    cvttsd2si %xmm0, %rax
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    ucomisd %xmm2, %xmm0
 ; SSE-NEXT:    cmovbq %rax, %rcx
-; SSE-NEXT:    movd %rcx, %xmm0
+; SSE-NEXT:    movq %rcx, %xmm0
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]
 ; SSE-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,2],xmm0[0,2]
 ; SSE-NEXT:    movaps %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: fptoui_4f64_to_2i32:
 ; VEX:       # BB#0:
 ; VEX-NEXT:    vpermilpd {{.*#+}} xmm1 = xmm0[1,0]

 ; SSE-NEXT:    movsd {{.*#+}} xmm3 = mem[0],zero
 ; SSE-NEXT:    subsd %xmm3, %xmm0
 ; SSE-NEXT:    cvttsd2si %xmm0, %rcx
 ; SSE-NEXT:    movabsq $-9223372036854775808, %rax # imm = 0x8000000000000000
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttsd2si %xmm2, %rdx
 ; SSE-NEXT:    ucomisd %xmm3, %xmm2
 ; SSE-NEXT:    cmovaeq %rcx, %rdx
-; SSE-NEXT:    movd %rdx, %xmm0
+; SSE-NEXT:    movq %rdx, %xmm0
 ; SSE-NEXT:    movhlps {{.*#+}} xmm2 = xmm2[1,1]
 ; SSE-NEXT:    movaps %xmm2, %xmm4
 ; SSE-NEXT:    subsd %xmm3, %xmm4
 ; SSE-NEXT:    cvttsd2si %xmm4, %rcx
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttsd2si %xmm2, %rdx
 ; SSE-NEXT:    ucomisd %xmm3, %xmm2
 ; SSE-NEXT:    cmovaeq %rcx, %rdx
-; SSE-NEXT:    movd %rdx, %xmm2
+; SSE-NEXT:    movq %rdx, %xmm2
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
 ; SSE-NEXT:    movapd %xmm1, %xmm2
 ; SSE-NEXT:    subsd %xmm3, %xmm2
 ; SSE-NEXT:    cvttsd2si %xmm2, %rcx
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttsd2si %xmm1, %rdx
 ; SSE-NEXT:    ucomisd %xmm3, %xmm1
 ; SSE-NEXT:    cmovaeq %rcx, %rdx
-; SSE-NEXT:    movd %rdx, %xmm2
+; SSE-NEXT:    movq %rdx, %xmm2
 ; SSE-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
 ; SSE-NEXT:    movaps %xmm1, %xmm4
 ; SSE-NEXT:    subsd %xmm3, %xmm4
 ; SSE-NEXT:    cvttsd2si %xmm4, %rcx
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttsd2si %xmm1, %rax
 ; SSE-NEXT:    ucomisd %xmm3, %xmm1
 ; SSE-NEXT:    cmovaeq %rcx, %rax
-; SSE-NEXT:    movd %rax, %xmm1
+; SSE-NEXT:    movq %rax, %xmm1
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
 ; SSE-NEXT:    movdqa %xmm2, %xmm1
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: fptoui_4f64_to_4i64:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
 ; AVX1-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero

 ; SSE-NEXT:    movapd %xmm1, %xmm3
 ; SSE-NEXT:    subsd %xmm2, %xmm3
 ; SSE-NEXT:    cvttsd2si %xmm3, %rcx
 ; SSE-NEXT:    movabsq $-9223372036854775808, %rax # imm = 0x8000000000000000
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttsd2si %xmm1, %rdx
 ; SSE-NEXT:    ucomisd %xmm2, %xmm1
 ; SSE-NEXT:    cmovaeq %rcx, %rdx
-; SSE-NEXT:    movd %rdx, %xmm3
+; SSE-NEXT:    movq %rdx, %xmm3
 ; SSE-NEXT:    movhlps {{.*#+}} xmm1 = xmm1[1,1]
 ; SSE-NEXT:    movaps %xmm1, %xmm4
 ; SSE-NEXT:    subsd %xmm2, %xmm4
 ; SSE-NEXT:    cvttsd2si %xmm4, %rcx
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttsd2si %xmm1, %rdx
 ; SSE-NEXT:    ucomisd %xmm2, %xmm1
 ; SSE-NEXT:    cmovaeq %rcx, %rdx
-; SSE-NEXT:    movd %rdx, %xmm1
+; SSE-NEXT:    movq %rdx, %xmm1
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm3 = xmm3[0],xmm1[0]
 ; SSE-NEXT:    movapd %xmm0, %xmm1
 ; SSE-NEXT:    subsd %xmm2, %xmm1
 ; SSE-NEXT:    cvttsd2si %xmm1, %rcx
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttsd2si %xmm0, %rdx
 ; SSE-NEXT:    ucomisd %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rcx, %rdx
-; SSE-NEXT:    movd %rdx, %xmm1
+; SSE-NEXT:    movq %rdx, %xmm1
 ; SSE-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; SSE-NEXT:    movaps %xmm0, %xmm4
 ; SSE-NEXT:    subsd %xmm2, %xmm4
 ; SSE-NEXT:    cvttsd2si %xmm4, %rcx
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttsd2si %xmm0, %rax
 ; SSE-NEXT:    ucomisd %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rcx, %rax
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,2],xmm3[0,2]
 ; SSE-NEXT:    movaps %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: fptoui_4f64_to_4i32:
 ; VEX:       # BB#0:
 ; VEX-NEXT:    vpermilpd {{.*#+}} xmm1 = xmm0[1,0]
; AVX-NEXT:    retq
   %cvt = fptosi <4 x float> %a to <4 x i32>
   ret <4 x i32> %cvt
 }
 
 define <2 x i64> @fptosi_2f32_to_2i64(<4 x float> %a) {
 ; SSE-LABEL: fptosi_2f32_to_2i64:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    cvttss2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm1
+; SSE-NEXT:    movq %rax, %xmm1
 ; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,2,3]
 ; SSE-NEXT:    cvttss2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE-NEXT:    movdqa %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: fptosi_2f32_to_2i64:
 ; VEX:       # BB#0:
 ; VEX-NEXT:    vcvttss2si %xmm0, %rax
 ; VEX-NEXT:    vmovq %rax, %xmm1
; AVX512VLDQ-NEXT:    retq
   %cvt = fptosi <2 x float> %shuf to <2 x i64>
   ret <2 x i64> %cvt
 }
 
 define <2 x i64> @fptosi_4f32_to_2i64(<4 x float> %a) {
 ; SSE-LABEL: fptosi_4f32_to_2i64:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    cvttss2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm1
+; SSE-NEXT:    movq %rax, %xmm1
 ; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,2,3]
 ; SSE-NEXT:    cvttss2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE-NEXT:    movdqa %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: fptosi_4f32_to_2i64:
 ; VEX:       # BB#0:
 ; VEX-NEXT:    vmovshdup {{.*#+}} xmm1 = xmm0[1,1,3,3]
 ; VEX-NEXT:    vcvttss2si %xmm1, %rax
; AVX-NEXT:    retq
   %cvt = fptosi <8 x float> %a to <8 x i32>
   ret <8 x i32> %cvt
 }
 
 define <4 x i64> @fptosi_4f32_to_4i64(<8 x float> %a) {
 ; SSE-LABEL: fptosi_4f32_to_4i64:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    cvttss2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm2
+; SSE-NEXT:    movq %rax, %xmm2
 ; SSE-NEXT:    movaps %xmm0, %xmm1
 ; SSE-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,1,2,3]
 ; SSE-NEXT:    cvttss2si %xmm1, %rax
-; SSE-NEXT:    movd %rax, %xmm1
+; SSE-NEXT:    movq %rax, %xmm1
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
 ; SSE-NEXT:    movaps %xmm0, %xmm1
 ; SSE-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,1,2,3]
 ; SSE-NEXT:    cvttss2si %xmm1, %rax
-; SSE-NEXT:    movd %rax, %xmm3
+; SSE-NEXT:    movq %rax, %xmm3
 ; SSE-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; SSE-NEXT:    cvttss2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm1
+; SSE-NEXT:    movq %rax, %xmm1
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
 ; SSE-NEXT:    movdqa %xmm2, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: fptosi_4f32_to_4i64:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vpermilps {{.*#+}} xmm1 = xmm0[3,1,2,3]
 ; AVX1-NEXT:    vcvttss2si %xmm1, %rax
; AVX512VLDQ-NEXT:    retq
   %cvt = fptosi <4 x float> %shuf to <4 x i64>
   ret <4 x i64> %cvt
 }
 
 define <4 x i64> @fptosi_8f32_to_4i64(<8 x float> %a) {
 ; SSE-LABEL: fptosi_8f32_to_4i64:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    cvttss2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm2
+; SSE-NEXT:    movq %rax, %xmm2
 ; SSE-NEXT:    movaps %xmm0, %xmm1
 ; SSE-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,1,2,3]
 ; SSE-NEXT:    cvttss2si %xmm1, %rax
-; SSE-NEXT:    movd %rax, %xmm1
+; SSE-NEXT:    movq %rax, %xmm1
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
 ; SSE-NEXT:    movaps %xmm0, %xmm1
 ; SSE-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,1,2,3]
 ; SSE-NEXT:    cvttss2si %xmm1, %rax
-; SSE-NEXT:    movd %rax, %xmm3
+; SSE-NEXT:    movq %rax, %xmm3
 ; SSE-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; SSE-NEXT:    cvttss2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm1
+; SSE-NEXT:    movq %rax, %xmm1
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
 ; SSE-NEXT:    movdqa %xmm2, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: fptosi_8f32_to_4i64:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vpermilps {{.*#+}} xmm1 = xmm0[3,1,2,3]
 ; AVX1-NEXT:    vcvttss2si %xmm1, %rax

 ; SSE-NEXT:    movaps %xmm0, %xmm1
 ; SSE-NEXT:    subss %xmm2, %xmm1
 ; SSE-NEXT:    cvttss2si %xmm1, %rax
 ; SSE-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttss2si %xmm0, %rdx
 ; SSE-NEXT:    ucomiss %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rdx
-; SSE-NEXT:    movd %rdx, %xmm1
+; SSE-NEXT:    movq %rdx, %xmm1
 ; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,2,3]
 ; SSE-NEXT:    movaps %xmm0, %xmm3
 ; SSE-NEXT:    subss %xmm2, %xmm3
 ; SSE-NEXT:    cvttss2si %xmm3, %rax
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttss2si %xmm0, %rcx
 ; SSE-NEXT:    ucomiss %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rcx
-; SSE-NEXT:    movd %rcx, %xmm0
+; SSE-NEXT:    movq %rcx, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE-NEXT:    movdqa %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: fptoui_2f32_to_2i32:
 ; VEX:       # BB#0:
 ; VEX-NEXT:    vmovss {{.*#+}} xmm1 = mem[0],zero,zero,zero
 ; VEX-NEXT:    vsubss %xmm1, %xmm0, %xmm2

 ; SSE-NEXT:    movaps %xmm0, %xmm1
 ; SSE-NEXT:    subss %xmm2, %xmm1
 ; SSE-NEXT:    cvttss2si %xmm1, %rax
 ; SSE-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttss2si %xmm0, %rdx
 ; SSE-NEXT:    ucomiss %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rdx
-; SSE-NEXT:    movd %rdx, %xmm1
+; SSE-NEXT:    movq %rdx, %xmm1
 ; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,2,3]
 ; SSE-NEXT:    movaps %xmm0, %xmm3
 ; SSE-NEXT:    subss %xmm2, %xmm3
 ; SSE-NEXT:    cvttss2si %xmm3, %rax
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttss2si %xmm0, %rcx
 ; SSE-NEXT:    ucomiss %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rcx
-; SSE-NEXT:    movd %rcx, %xmm0
+; SSE-NEXT:    movq %rcx, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE-NEXT:    movdqa %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: fptoui_2f32_to_2i64:
 ; VEX:       # BB#0:
 ; VEX-NEXT:    vmovss {{.*#+}} xmm1 = mem[0],zero,zero,zero
 ; VEX-NEXT:    vsubss %xmm1, %xmm0, %xmm2

 ; SSE-NEXT:    movaps %xmm0, %xmm1
 ; SSE-NEXT:    subss %xmm2, %xmm1
 ; SSE-NEXT:    cvttss2si %xmm1, %rax
 ; SSE-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttss2si %xmm0, %rdx
 ; SSE-NEXT:    ucomiss %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rdx
-; SSE-NEXT:    movd %rdx, %xmm1
+; SSE-NEXT:    movq %rdx, %xmm1
 ; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,2,3]
 ; SSE-NEXT:    movaps %xmm0, %xmm3
 ; SSE-NEXT:    subss %xmm2, %xmm3
 ; SSE-NEXT:    cvttss2si %xmm3, %rax
 ; SSE-NEXT:    xorq %rcx, %rax
 ; SSE-NEXT:    cvttss2si %xmm0, %rcx
 ; SSE-NEXT:    ucomiss %xmm2, %xmm0
 ; SSE-NEXT:    cmovaeq %rax, %rcx
-; SSE-NEXT:    movd %rcx, %xmm0
+; SSE-NEXT:    movq %rcx, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE-NEXT:    movdqa %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: fptoui_4f32_to_2i64:
 ; VEX:       # BB#0:
 ; VEX-NEXT:    vmovshdup {{.*#+}} xmm1 = xmm0[1,1,3,3]
 ; VEX-NEXT:    vmovss {{.*#+}} xmm2 = mem[0],zero,zero,zero

 ; SSE-NEXT:    movaps %xmm0, %xmm2
 ; SSE-NEXT:    subss %xmm1, %xmm2
 ; SSE-NEXT:    cvttss2si %xmm2, %rcx
 ; SSE-NEXT:    movabsq $-9223372036854775808, %rax # imm = 0x8000000000000000
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttss2si %xmm0, %rdx
 ; SSE-NEXT:    ucomiss %xmm1, %xmm0
 ; SSE-NEXT:    cmovaeq %rcx, %rdx
-; SSE-NEXT:    movd %rdx, %xmm2
+; SSE-NEXT:    movq %rdx, %xmm2
 ; SSE-NEXT:    movaps %xmm0, %xmm3
 ; SSE-NEXT:    shufps {{.*#+}} xmm3 = xmm3[1,1,2,3]
 ; SSE-NEXT:    movaps %xmm3, %xmm4
 ; SSE-NEXT:    subss %xmm1, %xmm4
 ; SSE-NEXT:    cvttss2si %xmm4, %rcx
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttss2si %xmm3, %rdx
 ; SSE-NEXT:    ucomiss %xmm1, %xmm3
 ; SSE-NEXT:    cmovaeq %rcx, %rdx
-; SSE-NEXT:    movd %rdx, %xmm3
+; SSE-NEXT:    movq %rdx, %xmm3
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm3[0]
 ; SSE-NEXT:    movaps %xmm0, %xmm3
 ; SSE-NEXT:    shufps {{.*#+}} xmm3 = xmm3[3,1,2,3]
 ; SSE-NEXT:    movaps %xmm3, %xmm4
 ; SSE-NEXT:    subss %xmm1, %xmm4
 ; SSE-NEXT:    cvttss2si %xmm4, %rcx
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttss2si %xmm3, %rdx
 ; SSE-NEXT:    ucomiss %xmm1, %xmm3
 ; SSE-NEXT:    cmovaeq %rcx, %rdx
-; SSE-NEXT:    movd %rdx, %xmm3
+; SSE-NEXT:    movq %rdx, %xmm3
 ; SSE-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; SSE-NEXT:    movaps %xmm0, %xmm4
 ; SSE-NEXT:    subss %xmm1, %xmm4
 ; SSE-NEXT:    cvttss2si %xmm4, %rcx
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttss2si %xmm0, %rax
 ; SSE-NEXT:    ucomiss %xmm1, %xmm0
 ; SSE-NEXT:    cmovaeq %rcx, %rax
-; SSE-NEXT:    movd %rax, %xmm1
+; SSE-NEXT:    movq %rax, %xmm1
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
 ; SSE-NEXT:    movdqa %xmm2, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: fptoui_4f32_to_4i64:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vpermilps {{.*#+}} xmm2 = xmm0[3,1,2,3]
 ; AVX1-NEXT:    vmovss {{.*#+}} xmm1 = mem[0],zero,zero,zero

 ; SSE-NEXT:    movaps %xmm0, %xmm2
 ; SSE-NEXT:    subss %xmm1, %xmm2
 ; SSE-NEXT:    cvttss2si %xmm2, %rcx
 ; SSE-NEXT:    movabsq $-9223372036854775808, %rax # imm = 0x8000000000000000
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttss2si %xmm0, %rdx
 ; SSE-NEXT:    ucomiss %xmm1, %xmm0
 ; SSE-NEXT:    cmovaeq %rcx, %rdx
-; SSE-NEXT:    movd %rdx, %xmm2
+; SSE-NEXT:    movq %rdx, %xmm2
 ; SSE-NEXT:    movaps %xmm0, %xmm3
 ; SSE-NEXT:    shufps {{.*#+}} xmm3 = xmm3[1,1,2,3]
 ; SSE-NEXT:    movaps %xmm3, %xmm4
 ; SSE-NEXT:    subss %xmm1, %xmm4
 ; SSE-NEXT:    cvttss2si %xmm4, %rcx
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttss2si %xmm3, %rdx
 ; SSE-NEXT:    ucomiss %xmm1, %xmm3
 ; SSE-NEXT:    cmovaeq %rcx, %rdx
-; SSE-NEXT:    movd %rdx, %xmm3
+; SSE-NEXT:    movq %rdx, %xmm3
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm3[0]
 ; SSE-NEXT:    movaps %xmm0, %xmm3
 ; SSE-NEXT:    shufps {{.*#+}} xmm3 = xmm3[3,1,2,3]
 ; SSE-NEXT:    movaps %xmm3, %xmm4
 ; SSE-NEXT:    subss %xmm1, %xmm4
 ; SSE-NEXT:    cvttss2si %xmm4, %rcx
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttss2si %xmm3, %rdx
 ; SSE-NEXT:    ucomiss %xmm1, %xmm3
 ; SSE-NEXT:    cmovaeq %rcx, %rdx
-; SSE-NEXT:    movd %rdx, %xmm3
+; SSE-NEXT:    movq %rdx, %xmm3
 ; SSE-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; SSE-NEXT:    movaps %xmm0, %xmm4
 ; SSE-NEXT:    subss %xmm1, %xmm4
 ; SSE-NEXT:    cvttss2si %xmm4, %rcx
 ; SSE-NEXT:    xorq %rax, %rcx
 ; SSE-NEXT:    cvttss2si %xmm0, %rax
 ; SSE-NEXT:    ucomiss %xmm1, %xmm0
 ; SSE-NEXT:    cmovaeq %rcx, %rax
-; SSE-NEXT:    movd %rax, %xmm1
+; SSE-NEXT:    movq %rax, %xmm1
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
 ; SSE-NEXT:    movdqa %xmm2, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: fptoui_8f32_to_4i64:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vpermilps {{.*#+}} xmm2 = xmm0[3,1,2,3]
 ; AVX1-NEXT:    vmovss {{.*#+}} xmm1 = mem[0],zero,zero,zero

 ; SSE-NEXT:    callq __gnu_h2f_ieee
 ; SSE-NEXT:    movss %xmm0, (%rsp) # 4-byte Spill
 ; SSE-NEXT:    movss {{[0-9]+}}(%rsp), %xmm0 # 4-byte Reload
 ; SSE-NEXT:    # xmm0 = mem[0],zero,zero,zero
 ; SSE-NEXT:    callq __gnu_f2h_ieee
 ; SSE-NEXT:    movzwl %ax, %edi
 ; SSE-NEXT:    callq __gnu_h2f_ieee
 ; SSE-NEXT:    cvttss2si %xmm0, %rax
-; SSE-NEXT:    movd %rax, %xmm1
+; SSE-NEXT:    movq %rax, %xmm1
 ; SSE-NEXT:    cvttss2si (%rsp), %rax # 4-byte Folded Reload
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE-NEXT:    pxor %xmm1, %xmm1
 ; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[2,3]
 ; SSE-NEXT:    popq %rax
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: fptosi_2f16_to_4i32:
 ; VEX:       # BB#0:

 ; SSE-NEXT:    pushq %r14
 ; SSE-NEXT:    pushq %rbx
 ; SSE-NEXT:    subq $24, %rsp
 ; SSE-NEXT:    movq %rsi, %r14
 ; SSE-NEXT:    movq %rdi, %rbx
 ; SSE-NEXT:    movq %rdx, %rdi
 ; SSE-NEXT:    movq %rcx, %rsi
 ; SSE-NEXT:    callq __fixtfdi
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    movdqa %xmm0, (%rsp) # 16-byte Spill
 ; SSE-NEXT:    movq %rbx, %rdi
 ; SSE-NEXT:    movq %r14, %rsi
 ; SSE-NEXT:    callq __fixtfdi
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    punpcklqdq (%rsp), %xmm0 # 16-byte Folded Reload
 ; SSE-NEXT:    # xmm0 = xmm0[0],mem[0]
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[2,3]
 ; SSE-NEXT:    addq $24, %rsp
 ; SSE-NEXT:    popq %rbx
 ; SSE-NEXT:    popq %r14
 ; SSE-NEXT:    retq

llvm/trunk/test/CodeGen/X86/vec_insert-3.ll

 ; X32-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
 ; X32-NEXT:    movss {{.*#+}} xmm1 = mem[0],zero,zero,zero
 ; X32-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,0],xmm0[2,0]
 ; X32-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[2,0]
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: t1:
 ; X64:       # BB#0:
-; X64-NEXT:    movd %rdi, %xmm1
+; X64-NEXT:    movq %rdi, %xmm1
 ; X64-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; X64-NEXT:    retq
   %tmp1 = insertelement <2 x i64> %tmp, i64 %s, i32 1
   ret <2 x i64> %tmp1
 }

llvm/trunk/test/CodeGen/X86/vec_insert-5.ll

 ; X32-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[1,0,1,1]
 ; X32-NEXT:    movq %xmm0, (%eax)
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: t1:
 ; X64:       # BB#0:
 ; X64-NEXT:    # kill: %EDI<def> %EDI<kill> %RDI<def>
 ; X64-NEXT:    shll $12, %edi
-; X64-NEXT:    movd %rdi, %xmm0
+; X64-NEXT:    movq %rdi, %xmm0
 ; X64-NEXT:    pslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4,5,6,7]
 ; X64-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
 ; X64-NEXT:    movq %xmm0, (%rsi)
 ; X64-NEXT:    retq
  %tmp12 = shl i32 %a, 12
  %tmp21 = insertelement <2 x i32> undef, i32 %tmp12, i32 1
  %tmp22 = insertelement <2 x i32> %tmp21, i32 0, i32 0
  %tmp23 = bitcast <2 x i32> %tmp22 to x86_mmx

llvm/trunk/test/CodeGen/X86/vec_insert-mmx.ll

 ; X32-NEXT:    movq %xmm0, (%esp)
 ; X32-NEXT:    movq (%esp), %mm0
 ; X32-NEXT:    addl $12, %esp
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: t0:
 ; X64:       ## BB#0:
 ; X64-NEXT:    ## kill: %EDI<def> %EDI<kill> %RDI<def>
-; X64-NEXT:    movd %rdi, %xmm0
+; X64-NEXT:    movq %rdi, %xmm0
 ; X64-NEXT:    pslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4,5,6,7]
 ; X64-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
 ; X64-NEXT:    retq
   %tmp3 = insertelement <2 x i32> < i32 0, i32 undef >, i32 %A, i32 1
   %tmp4 = bitcast <2 x i32> %tmp3 to x86_mmx
   ret x86_mmx %tmp4
 }
 

llvm/trunk/test/CodeGen/X86/vec_int_to_fp.ll

 
 ;
 ; Signed Integer to Double
 ;
 
 define <2 x double> @sitofp_2i64_to_2f64(<2 x i64> %a) {
 ; SSE-LABEL: sitofp_2i64_to_2f64:
 ; SSE:       # BB#0:
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    cvtsi2sdq %rax, %xmm1
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2sdq %rax, %xmm0
 ; SSE-NEXT:    unpcklpd {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE-NEXT:    movapd %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: sitofp_2i64_to_2f64:
 ; VEX:       # BB#0:
; AVX512-NEXT:    retq
   %cvt = sitofp <16 x i8> %a to <16 x double>
   %shuf = shufflevector <16 x double> %cvt, <16 x double> undef, <2 x i32> <i32 0, i32 1>
   ret <2 x double> %shuf
 }
 
 define <4 x double> @sitofp_4i64_to_4f64(<4 x i64> %a) {
 ; SSE-LABEL: sitofp_4i64_to_4f64:
 ; SSE:       # BB#0:
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    cvtsi2sdq %rax, %xmm2
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2sdq %rax, %xmm0
 ; SSE-NEXT:    unpcklpd {{.*#+}} xmm2 = xmm2[0],xmm0[0]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    cvtsi2sdq %rax, %xmm3
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[2,3,0,1]
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2sdq %rax, %xmm0
 ; SSE-NEXT:    unpcklpd {{.*#+}} xmm3 = xmm3[0],xmm0[0]
 ; SSE-NEXT:    movapd %xmm2, %xmm0
 ; SSE-NEXT:    movapd %xmm3, %xmm1
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: sitofp_4i64_to_4f64:

 
 ;
 ; Signed Integer to Float
 ;
 
 define <4 x float> @sitofp_2i64_to_4f32(<2 x i64> %a) {
 ; SSE-LABEL: sitofp_2i64_to_4f32:
 ; SSE:       # BB#0:
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
 ; SSE-NEXT:    movaps %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: sitofp_2i64_to_4f32:
 ; VEX:       # BB#0:
; AVX512VLDQ-NEXT:    retq
   %ext = shufflevector <2 x float> %cvt, <2 x float> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
   ret <4 x float> %ext
 }
 
 define <4 x float> @sitofp_2i64_to_4f32_zero(<2 x i64> %a) {
 ; SSE-LABEL: sitofp_2i64_to_4f32_zero:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
 ; SSE-NEXT:    movq {{.*#+}} xmm0 = xmm0[0],zero
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: sitofp_2i64_to_4f32_zero:
 ; VEX:       # BB#0:
; AVX512VLDQ-NEXT:    retq
   %ext = shufflevector <2 x float> %cvt, <2 x float> zeroinitializer, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
   ret <4 x float> %ext
 }
 
 define <4 x float> @sitofp_4i64_to_4f32_undef(<2 x i64> %a) {
 ; SSE-LABEL: sitofp_4i64_to_4f32_undef:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm2
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
 ; SSE-NEXT:    movaps %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: sitofp_4i64_to_4f32_undef:
; AVX512-NEXT:    retq
   %cvt = sitofp <16 x i8> %a to <16 x float>
   %shuf = shufflevector <16 x float> %cvt, <16 x float> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
   ret <4 x float> %shuf
 }
 
 define <4 x float> @sitofp_4i64_to_4f32(<4 x i64> %a) {
 ; SSE-LABEL: sitofp_4i64_to_4f32:
 ; SSE:       # BB#0:
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm3
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm2
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1]
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1]
 ; SSE-NEXT:    movaps %xmm2, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: sitofp_4i64_to_4f32:

 ;
 ; Unsigned Integer to Float
 ;
 
 define <4 x float> @uitofp_2i64_to_4f32(<2 x i64> %a) {
 ; SSE-LABEL: uitofp_2i64_to_4f32:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movdqa %xmm0, %xmm1
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB39_1
 ; SSE-NEXT:  # BB#2:
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    jmp .LBB39_3
 ; SSE-NEXT:  .LBB39_1:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    addss %xmm0, %xmm0
 ; SSE-NEXT:  .LBB39_3:
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB39_4
 ; SSE-NEXT:  # BB#5:
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
 ; SSE-NEXT:    retq
 ; SSE-NEXT:  .LBB39_4:
; AVX512VLDQ-NEXT:    retq
   %ext = shufflevector <2 x float> %cvt, <2 x float> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
   ret <4 x float> %ext
 }
 
 define <4 x float> @uitofp_2i64_to_2f32(<2 x i64> %a) {
 ; SSE-LABEL: uitofp_2i64_to_2f32:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB40_1
 ; SSE-NEXT:  # BB#2:
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    jmp .LBB40_3
 ; SSE-NEXT:  .LBB40_1:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    addss %xmm1, %xmm1
 ; SSE-NEXT:  .LBB40_3:
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB40_4
 ; SSE-NEXT:  # BB#5:
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    jmp .LBB40_6
 ; SSE-NEXT:  .LBB40_4:
 ; SSE-NEXT:    movq %rax, %rcx

 ; SSE-NEXT:    movdqa %xmm0, %xmm1
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    xorps %xmm2, %xmm2
 ; SSE-NEXT:    js .LBB41_2
 ; SSE-NEXT:  # BB#1:
 ; SSE-NEXT:    xorps %xmm2, %xmm2
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm2
 ; SSE-NEXT:  .LBB41_2:
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB41_3
 ; SSE-NEXT:  # BB#4:
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    jmp .LBB41_5
 ; SSE-NEXT:  .LBB41_3:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    addss %xmm0, %xmm0
 ; SSE-NEXT:  .LBB41_5:
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB41_6
 ; SSE-NEXT:  # BB#7:
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    jmp .LBB41_8
 ; SSE-NEXT:  .LBB41_6:
 ; SSE-NEXT:    movq %rax, %rcx
; AVX512-NEXT:    retq
   %cvt = uitofp <16 x i8> %a to <16 x float>
   %shuf = shufflevector <16 x float> %cvt, <16 x float> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
   ret <4 x float> %shuf
 }
 
 define <4 x float> @uitofp_4i64_to_4f32(<4 x i64> %a) {
 ; SSE-LABEL: uitofp_4i64_to_4f32:
 ; SSE:       # BB#0:
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB47_1
 ; SSE-NEXT:  # BB#2:
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm3
 ; SSE-NEXT:    jmp .LBB47_3
 ; SSE-NEXT:  .LBB47_1:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm3
 ; SSE-NEXT:    addss %xmm3, %xmm3
 ; SSE-NEXT:  .LBB47_3:
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB47_4
 ; SSE-NEXT:  # BB#5:
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm2
 ; SSE-NEXT:    jmp .LBB47_6
 ; SSE-NEXT:  .LBB47_4:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm2
 ; SSE-NEXT:    addss %xmm2, %xmm2
 ; SSE-NEXT:  .LBB47_6:
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB47_7
 ; SSE-NEXT:  # BB#8:
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    jmp .LBB47_9
 ; SSE-NEXT:  .LBB47_7:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    addss %xmm1, %xmm1
 ; SSE-NEXT:  .LBB47_9:
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1]
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB47_10
 ; SSE-NEXT:  # BB#11:
 ; SSE-NEXT:    xorps %xmm0, %xmm0
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    jmp .LBB47_12
 ; SSE-NEXT:  .LBB47_10:
 ; SSE-NEXT:    movq %rax, %rcx

 ;
 ; Load Signed Integer to Double
 ;
 
 define <2 x double> @sitofp_load_2i64_to_2f64(<2 x i64> *%a) {
 ; SSE-LABEL: sitofp_load_2i64_to_2f64:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movdqa (%rdi), %xmm1
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    cvtsi2sdq %rax, %xmm0
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2sdq %rax, %xmm1
 ; SSE-NEXT:    unpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE-NEXT:    retq
 ;
 ; VEX-LABEL: sitofp_load_2i64_to_2f64:
 ; VEX:       # BB#0:
 ; VEX-NEXT:    vmovdqa (%rdi), %xmm0
; AVX-NEXT:    retq
   ret <2 x double> %cvt
 }
 
 define <4 x double> @sitofp_load_4i64_to_4f64(<4 x i64> *%a) {
 ; SSE-LABEL: sitofp_load_4i64_to_4f64:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movdqa (%rdi), %xmm1
 ; SSE-NEXT:    movdqa 16(%rdi), %xmm2
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    cvtsi2sdq %rax, %xmm0
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2sdq %rax, %xmm1
 ; SSE-NEXT:    unpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; SSE-NEXT:    movd %xmm2, %rax
+; SSE-NEXT:    movq %xmm2, %rax
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2sdq %rax, %xmm1
 ; SSE-NEXT:    pshufd {{.*#+}} xmm2 = xmm2[2,3,0,1]
-; SSE-NEXT:    movd %xmm2, %rax
+; SSE-NEXT:    movq %xmm2, %rax
 ; SSE-NEXT:    xorps %xmm2, %xmm2
 ; SSE-NEXT:    cvtsi2sdq %rax, %xmm2
 ; SSE-NEXT:    unpcklpd {{.*#+}} xmm1 = xmm1[0],xmm2[0]
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: sitofp_load_4i64_to_4f64:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vmovdqa (%rdi), %ymm0

 ; Load Signed Integer to Float
 ;
 
 define <4 x float> @sitofp_load_4i64_to_4f32(<4 x i64> *%a) {
 ; SSE-LABEL: sitofp_load_4i64_to_4f32:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movdqa (%rdi), %xmm1
 ; SSE-NEXT:    movdqa 16(%rdi), %xmm2
-; SSE-NEXT:    movd %xmm2, %rax
+; SSE-NEXT:    movq %xmm2, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm3
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1]
 ; SSE-NEXT:    pshufd {{.*#+}} xmm2 = xmm2[2,3,0,1]
-; SSE-NEXT:    movd %xmm2, %rax
+; SSE-NEXT:    movq %xmm2, %rax
 ; SSE-NEXT:    xorps %xmm2, %xmm2
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm2
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: sitofp_load_4i64_to_4f32:
 ; AVX1:       # BB#0:

 
 define <8 x float> @sitofp_load_8i64_to_8f32(<8 x i64> *%a) {
 ; SSE-LABEL: sitofp_load_8i64_to_8f32:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movdqa (%rdi), %xmm1
 ; SSE-NEXT:    movdqa 16(%rdi), %xmm2
 ; SSE-NEXT:    movdqa 32(%rdi), %xmm3
 ; SSE-NEXT:    movdqa 48(%rdi), %xmm4
-; SSE-NEXT:    movd %xmm2, %rax
+; SSE-NEXT:    movq %xmm2, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm5
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm5[0],xmm0[1],xmm5[1]
 ; SSE-NEXT:    pshufd {{.*#+}} xmm2 = xmm2[2,3,0,1]
-; SSE-NEXT:    movd %xmm2, %rax
+; SSE-NEXT:    movq %xmm2, %rax
 ; SSE-NEXT:    xorps %xmm2, %xmm2
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm2
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1]
-; SSE-NEXT:    movd %xmm4, %rax
+; SSE-NEXT:    movq %xmm4, %rax
 ; SSE-NEXT:    xorps %xmm2, %xmm2
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm2
-; SSE-NEXT:    movd %xmm3, %rax
+; SSE-NEXT:    movq %xmm3, %rax
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
 ; SSE-NEXT:    pshufd {{.*#+}} xmm2 = xmm4[2,3,0,1]
-; SSE-NEXT:    movd %xmm2, %rax
+; SSE-NEXT:    movq %xmm2, %rax
 ; SSE-NEXT:    xorps %xmm2, %xmm2
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm2
 ; SSE-NEXT:    pshufd {{.*#+}} xmm3 = xmm3[2,3,0,1]
-; SSE-NEXT:    movd %xmm3, %rax
+; SSE-NEXT:    movq %xmm3, %rax
 ; SSE-NEXT:    xorps %xmm3, %xmm3
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm3
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm3 = xmm3[0],xmm2[0],xmm3[1],xmm2[1]
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1]
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: sitofp_load_8i64_to_8f32:
 ; AVX1:       # BB#0:

 ; Load Unsigned Integer to Float
 ;
 
 define <4 x float> @uitofp_load_4i64_to_4f32(<4 x i64> *%a) {
 ; SSE-LABEL: uitofp_load_4i64_to_4f32:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movdqa (%rdi), %xmm1
 ; SSE-NEXT:    movdqa 16(%rdi), %xmm3
-; SSE-NEXT:    movd %xmm3, %rax
+; SSE-NEXT:    movq %xmm3, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB76_1
 ; SSE-NEXT:  # BB#2:
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm2
 ; SSE-NEXT:    jmp .LBB76_3
 ; SSE-NEXT:  .LBB76_1:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm2
 ; SSE-NEXT:    addss %xmm2, %xmm2
 ; SSE-NEXT:  .LBB76_3:
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB76_4
 ; SSE-NEXT:  # BB#5:
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    jmp .LBB76_6
 ; SSE-NEXT:  .LBB76_4:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    addss %xmm0, %xmm0
 ; SSE-NEXT:  .LBB76_6:
 ; SSE-NEXT:    pshufd {{.*#+}} xmm3 = xmm3[2,3,0,1]
-; SSE-NEXT:    movd %xmm3, %rax
+; SSE-NEXT:    movq %xmm3, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB76_7
 ; SSE-NEXT:  # BB#8:
 ; SSE-NEXT:    xorps %xmm3, %xmm3
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm3
 ; SSE-NEXT:    jmp .LBB76_9
 ; SSE-NEXT:  .LBB76_7:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    xorps %xmm3, %xmm3
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm3
 ; SSE-NEXT:    addss %xmm3, %xmm3
 ; SSE-NEXT:  .LBB76_9:
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB76_10
 ; SSE-NEXT:  # BB#11:
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    jmp .LBB76_12
 ; SSE-NEXT:  .LBB76_10:
 ; SSE-NEXT:    movq %rax, %rcx

 
 define <8 x float> @uitofp_load_8i64_to_8f32(<8 x i64> *%a) {
 ; SSE-LABEL: uitofp_load_8i64_to_8f32:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movdqa (%rdi), %xmm1
 ; SSE-NEXT:    movdqa 16(%rdi), %xmm5
 ; SSE-NEXT:    movdqa 32(%rdi), %xmm2
 ; SSE-NEXT:    movdqa 48(%rdi), %xmm3
-; SSE-NEXT:    movd %xmm5, %rax
+; SSE-NEXT:    movq %xmm5, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB80_1
 ; SSE-NEXT:  # BB#2:
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm4
 ; SSE-NEXT:    jmp .LBB80_3
 ; SSE-NEXT:  .LBB80_1:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm4
 ; SSE-NEXT:    addss %xmm4, %xmm4
 ; SSE-NEXT:  .LBB80_3:
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB80_4
 ; SSE-NEXT:  # BB#5:
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    jmp .LBB80_6
 ; SSE-NEXT:  .LBB80_4:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm0
 ; SSE-NEXT:    addss %xmm0, %xmm0
 ; SSE-NEXT:  .LBB80_6:
 ; SSE-NEXT:    pshufd {{.*#+}} xmm5 = xmm5[2,3,0,1]
-; SSE-NEXT:    movd %xmm5, %rax
+; SSE-NEXT:    movq %xmm5, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB80_7
 ; SSE-NEXT:  # BB#8:
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm6
 ; SSE-NEXT:    jmp .LBB80_9
 ; SSE-NEXT:  .LBB80_7:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm6
 ; SSE-NEXT:    addss %xmm6, %xmm6
 ; SSE-NEXT:  .LBB80_9:
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB80_10
 ; SSE-NEXT:  # BB#11:
 ; SSE-NEXT:    xorps %xmm5, %xmm5
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm5
 ; SSE-NEXT:    jmp .LBB80_12
 ; SSE-NEXT:  .LBB80_10:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    xorps %xmm5, %xmm5
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm5
 ; SSE-NEXT:    addss %xmm5, %xmm5
 ; SSE-NEXT:  .LBB80_12:
-; SSE-NEXT:    movd %xmm3, %rax
+; SSE-NEXT:    movq %xmm3, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB80_13
 ; SSE-NEXT:  # BB#14:
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm7
 ; SSE-NEXT:    jmp .LBB80_15
 ; SSE-NEXT:  .LBB80_13:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm7
 ; SSE-NEXT:    addss %xmm7, %xmm7
 ; SSE-NEXT:  .LBB80_15:
-; SSE-NEXT:    movd %xmm2, %rax
+; SSE-NEXT:    movq %xmm2, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB80_16
 ; SSE-NEXT:  # BB#17:
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    jmp .LBB80_18
 ; SSE-NEXT:  .LBB80_16:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    xorps %xmm1, %xmm1
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm1
 ; SSE-NEXT:    addss %xmm1, %xmm1
 ; SSE-NEXT:  .LBB80_18:
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm4[0],xmm0[1],xmm4[1]
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm5 = xmm5[0],xmm6[0],xmm5[1],xmm6[1]
 ; SSE-NEXT:    pshufd {{.*#+}} xmm3 = xmm3[2,3,0,1]
-; SSE-NEXT:    movd %xmm3, %rax
+; SSE-NEXT:    movq %xmm3, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB80_19
 ; SSE-NEXT:  # BB#20:
 ; SSE-NEXT:    xorps %xmm3, %xmm3
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm3
 ; SSE-NEXT:    jmp .LBB80_21
 ; SSE-NEXT:  .LBB80_19:
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shrq %rcx
 ; SSE-NEXT:    andl $1, %eax
 ; SSE-NEXT:    orq %rcx, %rax
 ; SSE-NEXT:    xorps %xmm3, %xmm3
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm3
 ; SSE-NEXT:    addss %xmm3, %xmm3
 ; SSE-NEXT:  .LBB80_21:
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm0 = xmm0[0],xmm5[0],xmm0[1],xmm5[1]
 ; SSE-NEXT:    unpcklps {{.*#+}} xmm1 = xmm1[0],xmm7[0],xmm1[1],xmm7[1]
 ; SSE-NEXT:    pshufd {{.*#+}} xmm2 = xmm2[2,3,0,1]
-; SSE-NEXT:    movd %xmm2, %rax
+; SSE-NEXT:    movq %xmm2, %rax
 ; SSE-NEXT:    testq %rax, %rax
 ; SSE-NEXT:    js .LBB80_22
 ; SSE-NEXT:  # BB#23:
 ; SSE-NEXT:    xorps %xmm2, %xmm2
 ; SSE-NEXT:    cvtsi2ssq %rax, %xmm2
 ; SSE-NEXT:    jmp .LBB80_24
 ; SSE-NEXT:  .LBB80_22:
 ; SSE-NEXT:    movq %rax, %rcx

llvm/trunk/test/CodeGen/X86/vec_set-8.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+sse4.2 | FileCheck %s
 
 define <2 x i64> @test(i64 %i) nounwind  {
 ; CHECK-LABEL: test:
 ; CHECK:       # BB#0:
-; CHECK-NEXT:    movd %rdi, %xmm0
+; CHECK-NEXT:    movq %rdi, %xmm0
 ; CHECK-NEXT:    retq
   %tmp10 = insertelement <2 x i64> undef, i64 %i, i32 0
   %tmp11 = insertelement <2 x i64> %tmp10, i64 0, i32 1
   ret <2 x i64> %tmp11
 }

llvm/trunk/test/CodeGen/X86/vec_set-C.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=i386-linux-gnu -mattr=+sse2,-avx | FileCheck %s --check-prefix=X32
 ; RUN: llc < %s -mtriple=x86_64-pc-linux -mattr=+sse2,-avx | FileCheck %s --check-prefix=X64
 
 define <2 x i64> @t1(i64 %x) nounwind  {
 ; X32-LABEL: t1:
 ; X32:       # BB#0:
 ; X32-NEXT:    movsd {{.*#+}} xmm0 = mem[0],zero
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: t1:
 ; X64:       # BB#0:
-; X64-NEXT:    movd %rdi, %xmm0
+; X64-NEXT:    movq %rdi, %xmm0
 ; X64-NEXT:    retq
   %tmp8 = insertelement <2 x i64> zeroinitializer, i64 %x, i32 0
   ret <2 x i64> %tmp8
 }

llvm/trunk/test/CodeGen/X86/vec_shift7.ll

 ; X32-NEXT:    movsd {{.*#+}} xmm1 = xmm0[0],xmm1[1]
 ; X32-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[1,1,2,3]
 ; X32-NEXT:    movd %xmm1, %edx
 ; X32-NEXT:    movd %xmm0, %eax
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: test1:
 ; X64:       # BB#0: # %entry
-; X64-NEXT:    movd %xmm0, %rax
+; X64-NEXT:    movq %xmm0, %rax
 ; X64-NEXT:    retq
 entry:
  %c = shl <2 x i64> %a, <i64 0, i64 2>
  %d = extractelement <2 x i64> %c, i32 0
  ret i64 %d
 }

llvm/trunk/test/CodeGen/X86/vector-compare-all_of.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+sse4.2 | FileCheck %s --check-prefix=SSE
 ; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+avx | FileCheck %s --check-prefix=AVX --check-prefix=AVX1
 ; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+avx2 | FileCheck %s --check-prefix=AVX --check-prefix=AVX2
 ; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+avx512f,+avx512bw,+avx512vl | FileCheck %s --check-prefix=AVX512
 
 define i64 @test_v2f64_sext(<2 x double> %a0, <2 x double> %a1) {
 ; SSE-LABEL: test_v2f64_sext:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    cmpltpd %xmm0, %xmm1
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[2,3,0,1]
 ; SSE-NEXT:    pand %xmm1, %xmm0
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: test_v2f64_sext:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vcmpltpd %xmm0, %xmm1, %xmm0
 ; AVX-NEXT:    vpermilpd {{.*#+}} xmm1 = xmm0[1,0]
 ; AVX-NEXT:    vandpd %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovq %xmm0, %rax

 define i64 @test_v4f64_sext(<4 x double> %a0, <4 x double> %a1) {
 ; SSE-LABEL: test_v4f64_sext:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    cmpltpd %xmm1, %xmm3
 ; SSE-NEXT:    cmpltpd %xmm0, %xmm2
 ; SSE-NEXT:    andpd %xmm3, %xmm2
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm2[2,3,0,1]
 ; SSE-NEXT:    pand %xmm2, %xmm0
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: test_v4f64_sext:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vcmpltpd %ymm0, %ymm1, %ymm0
 ; AVX-NEXT:    vmovmskpd %ymm0, %eax
 ; AVX-NEXT:    xorl %ecx, %ecx
 ; AVX-NEXT:    cmpl $15, %eax

 }
 
 define i64 @test_v2i64_sext(<2 x i64> %a0, <2 x i64> %a1) {
 ; SSE-LABEL: test_v2i64_sext:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    pcmpgtq %xmm1, %xmm0
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
 ; SSE-NEXT:    pand %xmm0, %xmm1
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: test_v2i64_sext:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vpcmpgtq %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
 ; AVX-NEXT:    vpand %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovq %xmm0, %rax

 define i64 @test_v4i64_sext(<4 x i64> %a0, <4 x i64> %a1) {
 ; SSE-LABEL: test_v4i64_sext:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    pcmpgtq %xmm3, %xmm1
 ; SSE-NEXT:    pcmpgtq %xmm2, %xmm0
 ; SSE-NEXT:    pand %xmm1, %xmm0
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
 ; SSE-NEXT:    pand %xmm0, %xmm1
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: test_v4i64_sext:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm2
 ; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm3
 ; AVX1-NEXT:    vpcmpgtq %xmm2, %xmm3, %xmm2
 ; AVX1-NEXT:    vpcmpgtq %xmm1, %xmm0, %xmm0

llvm/trunk/test/CodeGen/X86/vector-compare-any_of.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+sse4.2 | FileCheck %s --check-prefix=SSE
 ; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+avx | FileCheck %s --check-prefix=AVX --check-prefix=AVX1
 ; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+avx2 | FileCheck %s --check-prefix=AVX --check-prefix=AVX2
 ; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+avx512f,+avx512bw,+avx512vl | FileCheck %s --check-prefix=AVX512
 
 define i64 @test_v2f64_sext(<2 x double> %a0, <2 x double> %a1) {
 ; SSE-LABEL: test_v2f64_sext:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    cmpltpd %xmm0, %xmm1
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[2,3,0,1]
 ; SSE-NEXT:    por %xmm1, %xmm0
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: test_v2f64_sext:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vcmpltpd %xmm0, %xmm1, %xmm0
 ; AVX-NEXT:    vpermilpd {{.*#+}} xmm1 = xmm0[1,0]
 ; AVX-NEXT:    vorpd %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovq %xmm0, %rax

 define i64 @test_v4f64_sext(<4 x double> %a0, <4 x double> %a1) {
 ; SSE-LABEL: test_v4f64_sext:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    cmpltpd %xmm1, %xmm3
 ; SSE-NEXT:    cmpltpd %xmm0, %xmm2
 ; SSE-NEXT:    orpd %xmm3, %xmm2
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm2[2,3,0,1]
 ; SSE-NEXT:    por %xmm2, %xmm0
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: test_v4f64_sext:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vcmpltpd %ymm0, %ymm1, %ymm0
 ; AVX-NEXT:    vmovmskpd %ymm0, %eax
 ; AVX-NEXT:    negl %eax
 ; AVX-NEXT:    sbbq %rax, %rax

 }
 
 define i64 @test_v2i64_sext(<2 x i64> %a0, <2 x i64> %a1) {
 ; SSE-LABEL: test_v2i64_sext:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    pcmpgtq %xmm1, %xmm0
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
 ; SSE-NEXT:    por %xmm0, %xmm1
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: test_v2i64_sext:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vpcmpgtq %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
 ; AVX-NEXT:    vpor %xmm1, %xmm0, %xmm0
 ; AVX-NEXT:    vmovq %xmm0, %rax

 define i64 @test_v4i64_sext(<4 x i64> %a0, <4 x i64> %a1) {
 ; SSE-LABEL: test_v4i64_sext:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    pcmpgtq %xmm3, %xmm1
 ; SSE-NEXT:    pcmpgtq %xmm2, %xmm0
 ; SSE-NEXT:    por %xmm1, %xmm0
 ; SSE-NEXT:    pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1]
 ; SSE-NEXT:    por %xmm0, %xmm1
-; SSE-NEXT:    movd %xmm1, %rax
+; SSE-NEXT:    movq %xmm1, %rax
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: test_v4i64_sext:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm2
 ; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm3
 ; AVX1-NEXT:    vpcmpgtq %xmm2, %xmm3, %xmm2
 ; AVX1-NEXT:    vpcmpgtq %xmm1, %xmm0, %xmm0

llvm/trunk/test/CodeGen/X86/vector-idiv-sdiv-128.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse2 | FileCheck %s --check-prefix=SSE --check-prefix=SSE2
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse4.1 | FileCheck %s --check-prefix=SSE --check-prefix=SSE41
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=AVX --check-prefix=AVX1
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=AVX --check-prefix=AVX2
 
 ;
 ; sdiv by 7
 ;
 
 define <2 x i64> @test_div7_2i64(<2 x i64> %a) nounwind {
 ; SSE2-LABEL: test_div7_2i64:
 ; SSE2:       # BB#0:
-; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    movq %xmm0, %rax
 ; SSE2-NEXT:    movabsq $5270498306774157605, %rcx # imm = 0x4924924924924925
 ; SSE2-NEXT:    imulq %rcx
 ; SSE2-NEXT:    movq %rdx, %rax
 ; SSE2-NEXT:    shrq $63, %rax
 ; SSE2-NEXT:    sarq %rdx
 ; SSE2-NEXT:    addq %rax, %rdx
-; SSE2-NEXT:    movd %rdx, %xmm1
+; SSE2-NEXT:    movq %rdx, %xmm1
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    movq %xmm0, %rax
 ; SSE2-NEXT:    imulq %rcx
 ; SSE2-NEXT:    movq %rdx, %rax
 ; SSE2-NEXT:    shrq $63, %rax
 ; SSE2-NEXT:    sarq %rdx
 ; SSE2-NEXT:    addq %rax, %rdx
-; SSE2-NEXT:    movd %rdx, %xmm0
+; SSE2-NEXT:    movq %rdx, %xmm0
 ; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE2-NEXT:    movdqa %xmm1, %xmm0
 ; SSE2-NEXT:    retq
 ;
 ; SSE41-LABEL: test_div7_2i64:
 ; SSE41:       # BB#0:
 ; SSE41-NEXT:    pextrq $1, %xmm0, %rax
 ; SSE41-NEXT:    movabsq $5270498306774157605, %rcx # imm = 0x4924924924924925
 ; SSE41-NEXT:    imulq %rcx
 ; SSE41-NEXT:    movq %rdx, %rax
 ; SSE41-NEXT:    shrq $63, %rax
 ; SSE41-NEXT:    sarq %rdx
 ; SSE41-NEXT:    addq %rax, %rdx
-; SSE41-NEXT:    movd %rdx, %xmm1
-; SSE41-NEXT:    movd %xmm0, %rax
+; SSE41-NEXT:    movq %rdx, %xmm1
+; SSE41-NEXT:    movq %xmm0, %rax
 ; SSE41-NEXT:    imulq %rcx
 ; SSE41-NEXT:    movq %rdx, %rax
 ; SSE41-NEXT:    shrq $63, %rax
 ; SSE41-NEXT:    sarq %rdx
 ; SSE41-NEXT:    addq %rax, %rdx
-; SSE41-NEXT:    movd %rdx, %xmm0
+; SSE41-NEXT:    movq %rdx, %xmm0
 ; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: test_div7_2i64:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vpextrq $1, %xmm0, %rax
 ; AVX-NEXT:    movabsq $5270498306774157605, %rcx # imm = 0x4924924924924925
 ; AVX-NEXT:    imulq %rcx

 
 ;
 ; srem by 7
 ;
 
 define <2 x i64> @test_rem7_2i64(<2 x i64> %a) nounwind {
 ; SSE2-LABEL: test_rem7_2i64:
 ; SSE2:       # BB#0:
-; SSE2-NEXT:    movd %xmm0, %rcx
+; SSE2-NEXT:    movq %xmm0, %rcx
 ; SSE2-NEXT:    movabsq $5270498306774157605, %rsi # imm = 0x4924924924924925
 ; SSE2-NEXT:    movq %rcx, %rax
 ; SSE2-NEXT:    imulq %rsi
 ; SSE2-NEXT:    movq %rdx, %rax
 ; SSE2-NEXT:    shrq $63, %rax
 ; SSE2-NEXT:    sarq %rdx
 ; SSE2-NEXT:    addq %rax, %rdx
 ; SSE2-NEXT:    leaq (,%rdx,8), %rax
 ; SSE2-NEXT:    subq %rdx, %rax
 ; SSE2-NEXT:    subq %rax, %rcx
-; SSE2-NEXT:    movd %rcx, %xmm1
+; SSE2-NEXT:    movq %rcx, %xmm1
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE2-NEXT:    movd %xmm0, %rcx
+; SSE2-NEXT:    movq %xmm0, %rcx
 ; SSE2-NEXT:    movq %rcx, %rax
 ; SSE2-NEXT:    imulq %rsi
 ; SSE2-NEXT:    movq %rdx, %rax
 ; SSE2-NEXT:    shrq $63, %rax
 ; SSE2-NEXT:    sarq %rdx
 ; SSE2-NEXT:    addq %rax, %rdx
 ; SSE2-NEXT:    leaq (,%rdx,8), %rax
 ; SSE2-NEXT:    subq %rdx, %rax
 ; SSE2-NEXT:    subq %rax, %rcx
-; SSE2-NEXT:    movd %rcx, %xmm0
+; SSE2-NEXT:    movq %rcx, %xmm0
 ; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE2-NEXT:    movdqa %xmm1, %xmm0
 ; SSE2-NEXT:    retq
 ;
 ; SSE41-LABEL: test_rem7_2i64:
 ; SSE41:       # BB#0:
 ; SSE41-NEXT:    pextrq $1, %xmm0, %rcx
 ; SSE41-NEXT:    movabsq $5270498306774157605, %rsi # imm = 0x4924924924924925
 ; SSE41-NEXT:    movq %rcx, %rax
 ; SSE41-NEXT:    imulq %rsi
 ; SSE41-NEXT:    movq %rdx, %rax
 ; SSE41-NEXT:    shrq $63, %rax
 ; SSE41-NEXT:    sarq %rdx
 ; SSE41-NEXT:    addq %rax, %rdx
 ; SSE41-NEXT:    leaq (,%rdx,8), %rax
 ; SSE41-NEXT:    subq %rdx, %rax
 ; SSE41-NEXT:    subq %rax, %rcx
-; SSE41-NEXT:    movd %rcx, %xmm1
-; SSE41-NEXT:    movd %xmm0, %rcx
+; SSE41-NEXT:    movq %rcx, %xmm1
+; SSE41-NEXT:    movq %xmm0, %rcx
 ; SSE41-NEXT:    movq %rcx, %rax
 ; SSE41-NEXT:    imulq %rsi
 ; SSE41-NEXT:    movq %rdx, %rax
 ; SSE41-NEXT:    shrq $63, %rax
 ; SSE41-NEXT:    sarq %rdx
 ; SSE41-NEXT:    addq %rax, %rdx
 ; SSE41-NEXT:    leaq (,%rdx,8), %rax
 ; SSE41-NEXT:    subq %rdx, %rax
 ; SSE41-NEXT:    subq %rax, %rcx
-; SSE41-NEXT:    movd %rcx, %xmm0
+; SSE41-NEXT:    movq %rcx, %xmm0
 ; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: test_rem7_2i64:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vpextrq $1, %xmm0, %rcx
 ; AVX-NEXT:    movabsq $5270498306774157605, %rsi # imm = 0x4924924924924925
 ; AVX-NEXT:    movq %rcx, %rax

llvm/trunk/test/CodeGen/X86/vector-idiv-udiv-128.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse2 | FileCheck %s --check-prefix=SSE --check-prefix=SSE2
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse4.1 | FileCheck %s --check-prefix=SSE --check-prefix=SSE41
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=AVX --check-prefix=AVX1
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=AVX --check-prefix=AVX2
 
 ;
 ; udiv by 7
 ;
 
 define <2 x i64> @test_div7_2i64(<2 x i64> %a) nounwind {
 ; SSE2-LABEL: test_div7_2i64:
 ; SSE2:       # BB#0:
-; SSE2-NEXT:    movd %xmm0, %rcx
+; SSE2-NEXT:    movq %xmm0, %rcx
 ; SSE2-NEXT:    movabsq $2635249153387078803, %rsi # imm = 0x2492492492492493
 ; SSE2-NEXT:    movq %rcx, %rax
 ; SSE2-NEXT:    mulq %rsi
 ; SSE2-NEXT:    subq %rdx, %rcx
 ; SSE2-NEXT:    shrq %rcx
 ; SSE2-NEXT:    addq %rdx, %rcx
 ; SSE2-NEXT:    shrq $2, %rcx
-; SSE2-NEXT:    movd %rcx, %xmm1
+; SSE2-NEXT:    movq %rcx, %xmm1
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE2-NEXT:    movd %xmm0, %rcx
+; SSE2-NEXT:    movq %xmm0, %rcx
 ; SSE2-NEXT:    movq %rcx, %rax
 ; SSE2-NEXT:    mulq %rsi
 ; SSE2-NEXT:    subq %rdx, %rcx
 ; SSE2-NEXT:    shrq %rcx
 ; SSE2-NEXT:    addq %rdx, %rcx
 ; SSE2-NEXT:    shrq $2, %rcx
-; SSE2-NEXT:    movd %rcx, %xmm0
+; SSE2-NEXT:    movq %rcx, %xmm0
 ; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE2-NEXT:    movdqa %xmm1, %xmm0
 ; SSE2-NEXT:    retq
 ;
 ; SSE41-LABEL: test_div7_2i64:
 ; SSE41:       # BB#0:
 ; SSE41-NEXT:    pextrq $1, %xmm0, %rcx
 ; SSE41-NEXT:    movabsq $2635249153387078803, %rsi # imm = 0x2492492492492493
 ; SSE41-NEXT:    movq %rcx, %rax
 ; SSE41-NEXT:    mulq %rsi
 ; SSE41-NEXT:    subq %rdx, %rcx
 ; SSE41-NEXT:    shrq %rcx
 ; SSE41-NEXT:    addq %rdx, %rcx
 ; SSE41-NEXT:    shrq $2, %rcx
-; SSE41-NEXT:    movd %rcx, %xmm1
-; SSE41-NEXT:    movd %xmm0, %rcx
+; SSE41-NEXT:    movq %rcx, %xmm1
+; SSE41-NEXT:    movq %xmm0, %rcx
 ; SSE41-NEXT:    movq %rcx, %rax
 ; SSE41-NEXT:    mulq %rsi
 ; SSE41-NEXT:    subq %rdx, %rcx
 ; SSE41-NEXT:    shrq %rcx
 ; SSE41-NEXT:    addq %rdx, %rcx
 ; SSE41-NEXT:    shrq $2, %rcx
-; SSE41-NEXT:    movd %rcx, %xmm0
+; SSE41-NEXT:    movq %rcx, %xmm0
 ; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: test_div7_2i64:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vpextrq $1, %xmm0, %rcx
 ; AVX-NEXT:    movabsq $2635249153387078803, %rsi # imm = 0x2492492492492493
 ; AVX-NEXT:    movq %rcx, %rax

 
 ;
 ; urem by 7
 ;
 
 define <2 x i64> @test_rem7_2i64(<2 x i64> %a) nounwind {
 ; SSE2-LABEL: test_rem7_2i64:
 ; SSE2:       # BB#0:
-; SSE2-NEXT:    movd %xmm0, %rcx
+; SSE2-NEXT:    movq %xmm0, %rcx
 ; SSE2-NEXT:    movabsq $2635249153387078803, %rsi # imm = 0x2492492492492493
 ; SSE2-NEXT:    movq %rcx, %rax
 ; SSE2-NEXT:    mulq %rsi
 ; SSE2-NEXT:    movq %rcx, %rax
 ; SSE2-NEXT:    subq %rdx, %rax
 ; SSE2-NEXT:    shrq %rax
 ; SSE2-NEXT:    addq %rdx, %rax
 ; SSE2-NEXT:    shrq $2, %rax
 ; SSE2-NEXT:    leaq (,%rax,8), %rdx
 ; SSE2-NEXT:    subq %rax, %rdx
 ; SSE2-NEXT:    subq %rdx, %rcx
-; SSE2-NEXT:    movd %rcx, %xmm1
+; SSE2-NEXT:    movq %rcx, %xmm1
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[2,3,0,1]
-; SSE2-NEXT:    movd %xmm0, %rcx
+; SSE2-NEXT:    movq %xmm0, %rcx
 ; SSE2-NEXT:    movq %rcx, %rax
 ; SSE2-NEXT:    mulq %rsi
 ; SSE2-NEXT:    movq %rcx, %rax
 ; SSE2-NEXT:    subq %rdx, %rax
 ; SSE2-NEXT:    shrq %rax
 ; SSE2-NEXT:    addq %rdx, %rax
 ; SSE2-NEXT:    shrq $2, %rax
 ; SSE2-NEXT:    leaq (,%rax,8), %rdx
 ; SSE2-NEXT:    subq %rax, %rdx
 ; SSE2-NEXT:    subq %rdx, %rcx
-; SSE2-NEXT:    movd %rcx, %xmm0
+; SSE2-NEXT:    movq %rcx, %xmm0
 ; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
 ; SSE2-NEXT:    movdqa %xmm1, %xmm0
 ; SSE2-NEXT:    retq
 ;
 ; SSE41-LABEL: test_rem7_2i64:
 ; SSE41:       # BB#0:
 ; SSE41-NEXT:    pextrq $1, %xmm0, %rcx
 ; SSE41-NEXT:    movabsq $2635249153387078803, %rsi # imm = 0x2492492492492493
 ; SSE41-NEXT:    movq %rcx, %rax
 ; SSE41-NEXT:    mulq %rsi
 ; SSE41-NEXT:    movq %rcx, %rax
 ; SSE41-NEXT:    subq %rdx, %rax
 ; SSE41-NEXT:    shrq %rax
 ; SSE41-NEXT:    addq %rdx, %rax
 ; SSE41-NEXT:    shrq $2, %rax
 ; SSE41-NEXT:    leaq (,%rax,8), %rdx
 ; SSE41-NEXT:    subq %rax, %rdx
 ; SSE41-NEXT:    subq %rdx, %rcx
-; SSE41-NEXT:    movd %rcx, %xmm1
-; SSE41-NEXT:    movd %xmm0, %rcx
+; SSE41-NEXT:    movq %rcx, %xmm1
+; SSE41-NEXT:    movq %xmm0, %rcx
 ; SSE41-NEXT:    movq %rcx, %rax
 ; SSE41-NEXT:    mulq %rsi
 ; SSE41-NEXT:    movq %rcx, %rax
 ; SSE41-NEXT:    subq %rdx, %rax
 ; SSE41-NEXT:    shrq %rax
 ; SSE41-NEXT:    addq %rdx, %rax
 ; SSE41-NEXT:    shrq $2, %rax
 ; SSE41-NEXT:    leaq (,%rax,8), %rdx
 ; SSE41-NEXT:    subq %rax, %rdx
 ; SSE41-NEXT:    subq %rdx, %rcx
-; SSE41-NEXT:    movd %rcx, %xmm0
+; SSE41-NEXT:    movq %rcx, %xmm0
 ; SSE41-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: test_rem7_2i64:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vpextrq $1, %xmm0, %rcx
 ; AVX-NEXT:    movabsq $2635249153387078803, %rsi # imm = 0x2492492492492493
 ; AVX-NEXT:    movq %rcx, %rax

llvm/trunk/test/CodeGen/X86/vector-lzcnt-128.ll

; X32-SSE-NEXT:    retl
   %out = call <16 x i8> @llvm.ctlz.v16i8(<16 x i8> %in, i1 -1)
   ret <16 x i8> %out
 }
 
 define <2 x i64> @foldv2i64() nounwind {
 ; SSE-LABEL: foldv2i64:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movl $55, %eax
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: foldv2i64:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    movl $55, %eax
 ; AVX-NEXT:    vmovq %rax, %xmm0
 ; AVX-NEXT:    retq
 ;
; X32-SSE-NEXT:    retl
   %out = call <2 x i64> @llvm.ctlz.v2i64(<2 x i64> <i64 256, i64 -1>, i1 0)
   ret <2 x i64> %out
 }
 
 define <2 x i64> @foldv2i64u() nounwind {
 ; SSE-LABEL: foldv2i64u:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movl $55, %eax
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: foldv2i64u:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    movl $55, %eax
 ; AVX-NEXT:    vmovq %rax, %xmm0
 ; AVX-NEXT:    retq
 ;

llvm/trunk/test/CodeGen/X86/vector-pcmp.ll

; AVX-NEXT:    retq
   %not = xor <2 x i64> %sign, <i64 -1, i64 -1>
   ret <2 x i64> %not
 }
 
 define <1 x i128> @test_strange_type(<1 x i128> %x) {
 ; SSE2-LABEL: test_strange_type:
 ; SSE2:       # BB#0:
 ; SSE2-NEXT:    sarq $63, %rsi
-; SSE2-NEXT:    movd %rsi, %xmm0
+; SSE2-NEXT:    movq %rsi, %xmm0
 ; SSE2-NEXT:    notq %rsi
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]
 ; SSE2-NEXT:    pcmpeqd %xmm1, %xmm1
 ; SSE2-NEXT:    pxor %xmm0, %xmm1
-; SSE2-NEXT:    movd %xmm1, %rax
+; SSE2-NEXT:    movq %xmm1, %rax
 ; SSE2-NEXT:    movq %rsi, %rdx
 ; SSE2-NEXT:    retq
 ;
 ; SSE42-LABEL: test_strange_type:
 ; SSE42:       # BB#0:
 ; SSE42-NEXT:    sarq $63, %rsi
-; SSE42-NEXT:    movd %rsi, %xmm0
+; SSE42-NEXT:    movq %rsi, %xmm0
 ; SSE42-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]
 ; SSE42-NEXT:    pcmpeqd %xmm1, %xmm1
 ; SSE42-NEXT:    pxor %xmm0, %xmm1
-; SSE42-NEXT:    movd %xmm1, %rax
+; SSE42-NEXT:    movq %xmm1, %rax
 ; SSE42-NEXT:    pextrq $1, %xmm1, %rdx
 ; SSE42-NEXT:    retq
 ;
 ; AVX1-LABEL: test_strange_type:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    sarq $63, %rsi
 ; AVX1-NEXT:    vmovq %rsi, %xmm0
 ; AVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,1,0,1]

llvm/trunk/test/CodeGen/X86/vector-sext.ll

 
 define <2 x i64> @load_sext_2i1_to_2i64(<2 x i1> *%ptr) {
 ; SSE-LABEL: load_sext_2i1_to_2i64:
 ; SSE:       # BB#0: # %entry
 ; SSE-NEXT:    movzbl (%rdi), %eax
 ; SSE-NEXT:    movq %rax, %rcx
 ; SSE-NEXT:    shlq $62, %rcx
 ; SSE-NEXT:    sarq $63, %rcx
-; SSE-NEXT:    movd %rcx, %xmm1
+; SSE-NEXT:    movq %rcx, %xmm1
 ; SSE-NEXT:    shlq $63, %rax
 ; SSE-NEXT:    sarq $63, %rax
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: load_sext_2i1_to_2i64:
 ; AVX1:       # BB#0: # %entry
 ; AVX1-NEXT:    movzbl (%rdi), %eax
 ; AVX1-NEXT:    movq %rax, %rcx
 ; AVX1-NEXT:    shlq $62, %rcx

  %Y = sext <4 x i1> %X to <4 x i64>
  ret <4 x i64> %Y
 }
 
 define <4 x i64> @load_sext_4i8_to_4i64(<4 x i8> *%ptr) {
 ; SSE2-LABEL: load_sext_4i8_to_4i64:
 ; SSE2:       # BB#0: # %entry
 ; SSE2-NEXT:    movsbq 1(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    movq %rax, %xmm1
 ; SSE2-NEXT:    movsbq (%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm0
+; SSE2-NEXT:    movq %rax, %xmm0
 ; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE2-NEXT:    movsbq 3(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm2
+; SSE2-NEXT:    movq %rax, %xmm2
 ; SSE2-NEXT:    movsbq 2(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    movq %rax, %xmm1
 ; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
 ; SSE2-NEXT:    retq
 ;
 ; SSSE3-LABEL: load_sext_4i8_to_4i64:
 ; SSSE3:       # BB#0: # %entry
 ; SSSE3-NEXT:    movsbq 1(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    movq %rax, %xmm1
 ; SSSE3-NEXT:    movsbq (%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm0
+; SSSE3-NEXT:    movq %rax, %xmm0
 ; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSSE3-NEXT:    movsbq 3(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm2
+; SSSE3-NEXT:    movq %rax, %xmm2
 ; SSSE3-NEXT:    movsbq 2(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    movq %rax, %xmm1
 ; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
 ; SSSE3-NEXT:    retq
 ;
 ; SSE41-LABEL: load_sext_4i8_to_4i64:
 ; SSE41:       # BB#0: # %entry
 ; SSE41-NEXT:    pmovsxbq (%rdi), %xmm0
 ; SSE41-NEXT:    pmovsxbq 2(%rdi), %xmm1
 ; SSE41-NEXT:    retq

  %Y = sext <8 x i8> %X to <8 x i16>
  ret <8 x i16> %Y
 }
 
 define <8 x i64> @load_sext_8i8_to_8i64(<8 x i8> *%ptr) {
 ; SSE2-LABEL: load_sext_8i8_to_8i64:
 ; SSE2:       # BB#0: # %entry
 ; SSE2-NEXT:    movsbq 1(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    movq %rax, %xmm1
 ; SSE2-NEXT:    movsbq (%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm0
+; SSE2-NEXT:    movq %rax, %xmm0
 ; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE2-NEXT:    movsbq 3(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm2
+; SSE2-NEXT:    movq %rax, %xmm2
 ; SSE2-NEXT:    movsbq 2(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    movq %rax, %xmm1
 ; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
 ; SSE2-NEXT:    movsbq 5(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm3
+; SSE2-NEXT:    movq %rax, %xmm3
 ; SSE2-NEXT:    movsbq 4(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm2
+; SSE2-NEXT:    movq %rax, %xmm2
 ; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm3[0]
 ; SSE2-NEXT:    movsbq 7(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm4
+; SSE2-NEXT:    movq %rax, %xmm4
 ; SSE2-NEXT:    movsbq 6(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm3
+; SSE2-NEXT:    movq %rax, %xmm3
 ; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm3 = xmm3[0],xmm4[0]
 ; SSE2-NEXT:    retq
 ;
 ; SSSE3-LABEL: load_sext_8i8_to_8i64:
 ; SSSE3:       # BB#0: # %entry
 ; SSSE3-NEXT:    movsbq 1(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    movq %rax, %xmm1
 ; SSSE3-NEXT:    movsbq (%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm0
+; SSSE3-NEXT:    movq %rax, %xmm0
 ; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSSE3-NEXT:    movsbq 3(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm2
+; SSSE3-NEXT:    movq %rax, %xmm2
 ; SSSE3-NEXT:    movsbq 2(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    movq %rax, %xmm1
 ; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
 ; SSSE3-NEXT:    movsbq 5(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm3
+; SSSE3-NEXT:    movq %rax, %xmm3
 ; SSSE3-NEXT:    movsbq 4(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm2
+; SSSE3-NEXT:    movq %rax, %xmm2
 ; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm3[0]
 ; SSSE3-NEXT:    movsbq 7(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm4
+; SSSE3-NEXT:    movq %rax, %xmm4
 ; SSSE3-NEXT:    movsbq 6(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm3
+; SSSE3-NEXT:    movq %rax, %xmm3
 ; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm3 = xmm3[0],xmm4[0]
 ; SSSE3-NEXT:    retq
 ;
 ; SSE41-LABEL: load_sext_8i8_to_8i64:
 ; SSE41:       # BB#0: # %entry
 ; SSE41-NEXT:    pmovsxbq (%rdi), %xmm0
 ; SSE41-NEXT:    pmovsxbq 2(%rdi), %xmm1
 ; SSE41-NEXT:    pmovsxbq 4(%rdi), %xmm2

  %Y = sext <4 x i16> %X to <4 x i32>
  ret <4 x i32> %Y
 }
 
 define <4 x i64> @load_sext_4i16_to_4i64(<4 x i16> *%ptr) {
 ; SSE2-LABEL: load_sext_4i16_to_4i64:
 ; SSE2:       # BB#0: # %entry
 ; SSE2-NEXT:    movswq 2(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    movq %rax, %xmm1
 ; SSE2-NEXT:    movswq (%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm0
+; SSE2-NEXT:    movq %rax, %xmm0
 ; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE2-NEXT:    movswq 6(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm2
+; SSE2-NEXT:    movq %rax, %xmm2
 ; SSE2-NEXT:    movswq 4(%rdi), %rax
-; SSE2-NEXT:    movd %rax, %xmm1
+; SSE2-NEXT:    movq %rax, %xmm1
 ; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
 ; SSE2-NEXT:    retq
 ;
 ; SSSE3-LABEL: load_sext_4i16_to_4i64:
 ; SSSE3:       # BB#0: # %entry
 ; SSSE3-NEXT:    movswq 2(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    movq %rax, %xmm1
 ; SSSE3-NEXT:    movswq (%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm0
+; SSSE3-NEXT:    movq %rax, %xmm0
 ; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSSE3-NEXT:    movswq 6(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm2
+; SSSE3-NEXT:    movq %rax, %xmm2
 ; SSSE3-NEXT:    movswq 4(%rdi), %rax
-; SSSE3-NEXT:    movd %rax, %xmm1
+; SSSE3-NEXT:    movq %rax, %xmm1
 ; SSSE3-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
 ; SSSE3-NEXT:    retq
 ;
 ; SSE41-LABEL: load_sext_4i16_to_4i64:
 ; SSE41:       # BB#0: # %entry
 ; SSE41-NEXT:    pmovsxwq (%rdi), %xmm0
 ; SSE41-NEXT:    pmovsxwq 4(%rdi), %xmm1
 ; SSE41-NEXT:    retq

llvm/trunk/test/CodeGen/X86/vector-shuffle-128-v2.ll

; AVX512VL-NEXT:    retq
   %bitcast64 = bitcast <4 x float> %shuffle32 to <2 x i64>
   %and = and <2 x i64> %bitcast64, <i64 -4294967296, i64 -1>
   ret <2 x i64> %and
 }
 
 define <2 x i64> @insert_reg_and_zero_v2i64(i64 %a) {
 ; SSE-LABEL: insert_reg_and_zero_v2i64:
 ; SSE:       # BB#0:
-; SSE-NEXT:    movd %rdi, %xmm0
+; SSE-NEXT:    movq %rdi, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: insert_reg_and_zero_v2i64:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vmovq %rdi, %xmm0
 ; AVX-NEXT:    retq
   %v = insertelement <2 x i64> undef, i64 %a, i32 0
   %shuffle = shufflevector <2 x i64> %v, <2 x i64> zeroinitializer, <2 x i32> <i32 0, i32 3>
; AVX-NEXT:    retq
   %v = insertelement <2 x double> undef, double %a, i32 0
   %shuffle = shufflevector <2 x double> %v, <2 x double> zeroinitializer, <2 x i32> <i32 0, i32 3>
   ret <2 x double> %shuffle
 }
 
 define <2 x i64> @insert_reg_lo_v2i64(i64 %a, <2 x i64> %b) {
 ; SSE2-LABEL: insert_reg_lo_v2i64:
 ; SSE2:       # BB#0:
-; SSE2-NEXT:    movd %rdi, %xmm1
+; SSE2-NEXT:    movq %rdi, %xmm1
 ; SSE2-NEXT:    movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
 ; SSE2-NEXT:    retq
 ;
 ; SSE3-LABEL: insert_reg_lo_v2i64:
 ; SSE3:       # BB#0:
-; SSE3-NEXT:    movd %rdi, %xmm1
+; SSE3-NEXT:    movq %rdi, %xmm1
 ; SSE3-NEXT:    movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
 ; SSE3-NEXT:    retq
 ;
 ; SSSE3-LABEL: insert_reg_lo_v2i64:
 ; SSSE3:       # BB#0:
-; SSSE3-NEXT:    movd %rdi, %xmm1
+; SSSE3-NEXT:    movq %rdi, %xmm1
 ; SSSE3-NEXT:    movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
 ; SSSE3-NEXT:    retq
 ;
 ; SSE41-LABEL: insert_reg_lo_v2i64:
 ; SSE41:       # BB#0:
-; SSE41-NEXT:    movd %rdi, %xmm1
+; SSE41-NEXT:    movq %rdi, %xmm1
 ; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
 ; SSE41-NEXT:    retq
 ;
 ; AVX1-LABEL: insert_reg_lo_v2i64:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vmovq %rdi, %xmm1
 ; AVX1-NEXT:    vpblendw {{.*#+}} xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
 ; AVX1-NEXT:    retq
; AVX512VL-NEXT:    retq
   %v = insertelement <2 x i64> undef, i64 %a, i32 0
   %shuffle = shufflevector <2 x i64> %v, <2 x i64> %b, <2 x i32> <i32 0, i32 3>
   ret <2 x i64> %shuffle
 }
 
 define <2 x i64> @insert_reg_hi_v2i64(i64 %a, <2 x i64> %b) {
 ; SSE-LABEL: insert_reg_hi_v2i64:
 ; SSE:       # BB#0:
-; SSE-NEXT:    movd %rdi, %xmm1
+; SSE-NEXT:    movq %rdi, %xmm1
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: insert_reg_hi_v2i64:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vmovq %rdi, %xmm1
 ; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; AVX-NEXT:    retq

llvm/trunk/test/CodeGen/X86/vector-shuffle-128-v4.ll

; AVX-NEXT:    retq
   %v = insertelement <4 x float> undef, float %a, i32 0
   %shuffle = shufflevector <4 x float> %v, <4 x float> zeroinitializer, <4 x i32> <i32 0, i32 5, i32 6, i32 7>
   ret <4 x float> %shuffle
 }
 
 define <4 x i32> @insert_reg_lo_v4i32(i64 %a, <4 x i32> %b) {
 ; SSE2-LABEL: insert_reg_lo_v4i32:
 ; SSE2:       # BB#0:
-; SSE2-NEXT:    movd %rdi, %xmm1
+; SSE2-NEXT:    movq %rdi, %xmm1
 ; SSE2-NEXT:    movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
 ; SSE2-NEXT:    retq
 ;
 ; SSE3-LABEL: insert_reg_lo_v4i32:
 ; SSE3:       # BB#0:
-; SSE3-NEXT:    movd %rdi, %xmm1
+; SSE3-NEXT:    movq %rdi, %xmm1
 ; SSE3-NEXT:    movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
 ; SSE3-NEXT:    retq
 ;
 ; SSSE3-LABEL: insert_reg_lo_v4i32:
 ; SSSE3:       # BB#0:
-; SSSE3-NEXT:    movd %rdi, %xmm1
+; SSSE3-NEXT:    movq %rdi, %xmm1
 ; SSSE3-NEXT:    movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
 ; SSSE3-NEXT:    retq
 ;
 ; SSE41-LABEL: insert_reg_lo_v4i32:
 ; SSE41:       # BB#0:
-; SSE41-NEXT:    movd %rdi, %xmm1
+; SSE41-NEXT:    movq %rdi, %xmm1
 ; SSE41-NEXT:    pblendw {{.*#+}} xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
 ; SSE41-NEXT:    retq
 ;
 ; AVX1-LABEL: insert_reg_lo_v4i32:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vmovq %rdi, %xmm1
 ; AVX1-NEXT:    vpblendw {{.*#+}} xmm0 = xmm1[0,1,2,3],xmm0[4,5,6,7]
 ; AVX1-NEXT:    retq
; AVX512VL-NEXT:    retq
   %v = shufflevector <2 x i32> %a, <2 x i32> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
   %shuffle = shufflevector <4 x i32> %v, <4 x i32> %b, <4 x i32> <i32 0, i32 1, i32 6, i32 7>
   ret <4 x i32> %shuffle
 }
 
 define <4 x i32> @insert_reg_hi_v4i32(i64 %a, <4 x i32> %b) {
 ; SSE-LABEL: insert_reg_hi_v4i32:
 ; SSE:       # BB#0:
-; SSE-NEXT:    movd %rdi, %xmm1
+; SSE-NEXT:    movq %rdi, %xmm1
 ; SSE-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: insert_reg_hi_v4i32:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    vmovq %rdi, %xmm1
 ; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
 ; AVX-NEXT:    retq

llvm/trunk/test/CodeGen/X86/vector-trunc-math.ll

 ;
 ; sub to constant
 ;
 
 define <4 x i32> @trunc_sub_const_v4i64_v4i32(<4 x i64> %a0) nounwind {
 ; SSE-LABEL: trunc_sub_const_v4i64_v4i32:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movl $1, %eax
-; SSE-NEXT:    movd %rax, %xmm2
+; SSE-NEXT:    movq %rax, %xmm2
 ; SSE-NEXT:    pslldq {{.*#+}} xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,xmm2[0,1,2,3,4,5,6,7]
 ; SSE-NEXT:    psubq %xmm2, %xmm0
 ; SSE-NEXT:    psubq {{.*}}(%rip), %xmm1
 ; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
 ; SSE-NEXT:    retq
 ;
 ; AVX1-LABEL: trunc_sub_const_v4i64_v4i32:
 ; AVX1:       # BB#0:
; AVX512-NEXT:    retq
   %2 = trunc <4 x i64> %1 to <4 x i32>
   ret <4 x i32> %2
 }
 
 define <8 x i16> @trunc_sub_const_v8i64_v8i16(<8 x i64> %a0) nounwind {
 ; SSE-LABEL: trunc_sub_const_v8i64_v8i16:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movl $1, %eax
-; SSE-NEXT:    movd %rax, %xmm4
+; SSE-NEXT:    movq %rax, %xmm4
 ; SSE-NEXT:    pslldq {{.*#+}} xmm4 = zero,zero,zero,zero,zero,zero,zero,zero,xmm4[0,1,2,3,4,5,6,7]
 ; SSE-NEXT:    psubq %xmm4, %xmm0
 ; SSE-NEXT:    psubq {{.*}}(%rip), %xmm1
 ; SSE-NEXT:    psubq {{.*}}(%rip), %xmm2
 ; SSE-NEXT:    psubq {{.*}}(%rip), %xmm3
 ; SSE-NEXT:    pshufd {{.*#+}} xmm3 = xmm3[0,2,2,3]
 ; SSE-NEXT:    pshuflw {{.*#+}} xmm3 = xmm3[0,1,0,2,4,5,6,7]
 ; SSE-NEXT:    pshufd {{.*#+}} xmm2 = xmm2[0,2,2,3]
; AVX512-NEXT:    retq
   %2 = trunc <8 x i32> %1 to <8 x i16>
   ret <8 x i16> %2
 }
 
 define <16 x i8> @trunc_sub_const_v16i64_v16i8(<16 x i64> %a0) nounwind {
 ; SSE-LABEL: trunc_sub_const_v16i64_v16i8:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movl $1, %eax
-; SSE-NEXT:    movd %rax, %xmm8
+; SSE-NEXT:    movq %rax, %xmm8
 ; SSE-NEXT:    pslldq {{.*#+}} xmm8 = zero,zero,zero,zero,zero,zero,zero,zero,xmm8[0,1,2,3,4,5,6,7]
 ; SSE-NEXT:    psubq %xmm8, %xmm0
 ; SSE-NEXT:    psubq {{.*}}(%rip), %xmm1
 ; SSE-NEXT:    psubq {{.*}}(%rip), %xmm2
 ; SSE-NEXT:    psubq {{.*}}(%rip), %xmm3
 ; SSE-NEXT:    psubq {{.*}}(%rip), %xmm4
 ; SSE-NEXT:    psubq {{.*}}(%rip), %xmm5
 ; SSE-NEXT:    psubq {{.*}}(%rip), %xmm6

 ; SSE-NEXT:    movdqa {{.*#+}} xmm2 = [2,3]
 ; SSE-NEXT:    movdqa %xmm1, %xmm3
 ; SSE-NEXT:    pmuludq %xmm2, %xmm3
 ; SSE-NEXT:    psrlq $32, %xmm1
 ; SSE-NEXT:    pmuludq %xmm2, %xmm1
 ; SSE-NEXT:    psllq $32, %xmm1
 ; SSE-NEXT:    paddq %xmm3, %xmm1
 ; SSE-NEXT:    movl $1, %eax
-; SSE-NEXT:    movd %rax, %xmm2
+; SSE-NEXT:    movq %rax, %xmm2
 ; SSE-NEXT:    pslldq {{.*#+}} xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,xmm2[0,1,2,3,4,5,6,7]
 ; SSE-NEXT:    movdqa %xmm0, %xmm3
 ; SSE-NEXT:    pmuludq %xmm2, %xmm3
 ; SSE-NEXT:    psrlq $32, %xmm0
 ; SSE-NEXT:    pmuludq %xmm2, %xmm0
 ; SSE-NEXT:    psllq $32, %xmm0
 ; SSE-NEXT:    paddq %xmm3, %xmm0
 ; SSE-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
; AVX512-NEXT:    retq
   %2 = trunc <8 x i32> %1 to <8 x i16>
   ret <8 x i16> %2
 }
 
 define <16 x i8> @trunc_mul_const_v16i64_v16i8(<16 x i64> %a0) nounwind {
 ; SSE-LABEL: trunc_mul_const_v16i64_v16i8:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movl $1, %eax
-; SSE-NEXT:    movd %rax, %xmm8
+; SSE-NEXT:    movq %rax, %xmm8
 ; SSE-NEXT:    pslldq {{.*#+}} xmm8 = zero,zero,zero,zero,zero,zero,zero,zero,xmm8[0,1,2,3,4,5,6,7]
 ; SSE-NEXT:    movdqa %xmm0, %xmm9
 ; SSE-NEXT:    pmuludq %xmm8, %xmm9
 ; SSE-NEXT:    psrlq $32, %xmm0
 ; SSE-NEXT:    pmuludq %xmm8, %xmm0
 ; SSE-NEXT:    psllq $32, %xmm0
 ; SSE-NEXT:    paddq %xmm9, %xmm0
 ; SSE-NEXT:    movdqa {{.*#+}} xmm8 = [2,3]

llvm/trunk/test/CodeGen/X86/vector-trunc.ll

entry:
   %2 = shufflevector <2 x i32> %0, <2 x i32> %1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
   ret <4 x i32> %2
 }
 
 define i64 @trunc2i64_i64(<2 x i64> %inval) {
 ; SSE-LABEL: trunc2i64_i64:
 ; SSE:       # BB#0: # %entry
 ; SSE-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; SSE-NEXT:    movd %xmm0, %rax
+; SSE-NEXT:    movq %xmm0, %rax
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: trunc2i64_i64:
 ; AVX:       # BB#0: # %entry
 ; AVX-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
 ; AVX-NEXT:    vmovq %xmm0, %rax
 ; AVX-NEXT:    retq
 ;

 
 ; PR15524 http://llvm.org/bugs/show_bug.cgi?id=15524
 define i64 @trunc4i32_i64(<4 x i32> %inval) {
 ; SSE2-LABEL: trunc4i32_i64:
 ; SSE2:       # BB#0: # %entry
 ; SSE2-NEXT:    pshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
 ; SSE2-NEXT:    pshufhw {{.*#+}} xmm0 = xmm0[0,1,2,3,4,6,6,7]
 ; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
-; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    movq %xmm0, %rax
 ; SSE2-NEXT:    retq
 ;
 ; SSSE3-LABEL: trunc4i32_i64:
 ; SSSE3:       # BB#0: # %entry
 ; SSSE3-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
-; SSSE3-NEXT:    movd %xmm0, %rax
+; SSSE3-NEXT:    movq %xmm0, %rax
 ; SSSE3-NEXT:    retq
 ;
 ; SSE41-LABEL: trunc4i32_i64:
 ; SSE41:       # BB#0: # %entry
 ; SSE41-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
-; SSE41-NEXT:    movd %xmm0, %rax
+; SSE41-NEXT:    movq %xmm0, %rax
 ; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: trunc4i32_i64:
 ; AVX:       # BB#0: # %entry
 ; AVX-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
 ; AVX-NEXT:    vmovq %xmm0, %rax
 ; AVX-NEXT:    retq
 ;

 }
 
 ; PR15524 http://llvm.org/bugs/show_bug.cgi?id=15524
 define i64 @trunc8i16_i64(<8 x i16> %inval) {
 ; SSE2-LABEL: trunc8i16_i64:
 ; SSE2:       # BB#0: # %entry
 ; SSE2-NEXT:    pand {{.*}}(%rip), %xmm0
 ; SSE2-NEXT:    packuswb %xmm0, %xmm0
-; SSE2-NEXT:    movd %xmm0, %rax
+; SSE2-NEXT:    movq %xmm0, %rax
 ; SSE2-NEXT:    retq
 ;
 ; SSSE3-LABEL: trunc8i16_i64:
 ; SSSE3:       # BB#0: # %entry
 ; SSSE3-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u]
-; SSSE3-NEXT:    movd %xmm0, %rax
+; SSSE3-NEXT:    movq %xmm0, %rax
 ; SSSE3-NEXT:    retq
 ;
 ; SSE41-LABEL: trunc8i16_i64:
 ; SSE41:       # BB#0: # %entry
 ; SSE41-NEXT:    pshufb {{.*#+}} xmm0 = xmm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u]
-; SSE41-NEXT:    movd %xmm0, %rax
+; SSE41-NEXT:    movq %xmm0, %rax
 ; SSE41-NEXT:    retq
 ;
 ; AVX-LABEL: trunc8i16_i64:
 ; AVX:       # BB#0: # %entry
 ; AVX-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u]
 ; AVX-NEXT:    vmovq %xmm0, %rax
 ; AVX-NEXT:    retq
 ;

llvm/trunk/test/CodeGen/X86/vector-tzcnt-128.ll

; X32-SSE-NEXT:    retl
   %out = call <16 x i8> @llvm.cttz.v16i8(<16 x i8> %in, i1 -1)
   ret <16 x i8> %out
 }
 
 define <2 x i64> @foldv2i64() nounwind {
 ; SSE-LABEL: foldv2i64:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movl $8, %eax
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: foldv2i64:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    movl $8, %eax
 ; AVX-NEXT:    vmovq %rax, %xmm0
 ; AVX-NEXT:    retq
 ;
 ; X32-SSE-LABEL: foldv2i64:
 ; X32-SSE:       # BB#0:
 ; X32-SSE-NEXT:    movl $8, %eax
 ; X32-SSE-NEXT:    movd %eax, %xmm0
 ; X32-SSE-NEXT:    retl
   %out = call <2 x i64> @llvm.cttz.v2i64(<2 x i64> <i64 256, i64 -1>, i1 0)
   ret <2 x i64> %out
 }
 
 define <2 x i64> @foldv2i64u() nounwind {
 ; SSE-LABEL: foldv2i64u:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movl $8, %eax
-; SSE-NEXT:    movd %rax, %xmm0
+; SSE-NEXT:    movq %rax, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: foldv2i64u:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    movl $8, %eax
 ; AVX-NEXT:    vmovq %rax, %xmm0
 ; AVX-NEXT:    retq
 ;

llvm/trunk/test/CodeGen/X86/vmovq.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=sse4.1 | FileCheck %s --check-prefix=SSE
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=avx    | FileCheck %s --check-prefix=AVX
 
 define <2 x i64> @PR25554(<2 x i64> %v0, <2 x i64> %v1) {
 ; SSE-LABEL: PR25554:
 ; SSE:       # BB#0:
 ; SSE-NEXT:    movl $1, %eax
-; SSE-NEXT:    movd %rax, %xmm1
+; SSE-NEXT:    movq %rax, %xmm1
 ; SSE-NEXT:    por %xmm1, %xmm0
 ; SSE-NEXT:    pslldq {{.*#+}} xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0,1,2,3,4,5,6,7]
 ; SSE-NEXT:    paddq %xmm1, %xmm0
 ; SSE-NEXT:    retq
 ;
 ; AVX-LABEL: PR25554:
 ; AVX:       # BB#0:
 ; AVX-NEXT:    movl $1, %eax

llvm/trunk/test/CodeGen/X86/vshift-1.ll

 ; X32-NEXT:    pshufd {{.*#+}} xmm2 = xmm2[0,0,1,1]
 ; X32-NEXT:    punpckldq {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
 ; X32-NEXT:    psllq %xmm2, %xmm0
 ; X32-NEXT:    movdqa %xmm0, (%eax)
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: shift1b:
 ; X64:       # BB#0: # %entry
-; X64-NEXT:    movd %rsi, %xmm1
+; X64-NEXT:    movq %rsi, %xmm1
 ; X64-NEXT:    psllq %xmm1, %xmm0
 ; X64-NEXT:    movdqa %xmm0, (%rdi)
 ; X64-NEXT:    retq
 entry:
   %0 = insertelement <2 x i64> undef, i64 %amt, i32 0
   %1 = insertelement <2 x i64> %0, i64 %amt, i32 1
   %shl = shl <2 x i64> %val, %1
   store <2 x i64> %shl, <2 x i64>* %dst

llvm/trunk/test/CodeGen/X86/vshift-2.ll

 ; X32-NEXT:    pshufd {{.*#+}} xmm2 = xmm2[0,0,1,1]
 ; X32-NEXT:    punpckldq {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
 ; X32-NEXT:    psrlq %xmm2, %xmm0
 ; X32-NEXT:    movdqa %xmm0, (%eax)
 ; X32-NEXT:    retl
 ;
 ; X64-LABEL: shift1b:
 ; X64:       # BB#0: # %entry
-; X64-NEXT:    movd %rsi, %xmm1
+; X64-NEXT:    movq %rsi, %xmm1
 ; X64-NEXT:    psrlq %xmm1, %xmm0
 ; X64-NEXT:    movdqa %xmm0, (%rdi)
 ; X64-NEXT:    retq
 entry:
   %0 = insertelement <2 x i64> undef, i64 %amt, i32 0
   %1 = insertelement <2 x i64> %0, i64 %amt, i32 1
   %lshr = lshr <2 x i64> %val, %1
   store <2 x i64> %lshr, <2 x i64>* %dst

llvm/trunk/test/CodeGen/X86/vsplit-and.ll

; CHECK-NEXT:    retq
   %t2 = sext <2 x i1> %t1 to <2 x i64>
   store <2 x i64> %t2, <2 x i64>* %dst
   ret void
 }
 
 define void @t2(<3 x i64>* %dst, <3 x i64> %src1, <3 x i64> %src2) nounwind readonly {
 ; CHECK-LABEL: t2:
 ; CHECK:       # BB#0:
-; CHECK-NEXT:    movd %r9, %xmm1
-; CHECK-NEXT:    movd %r8, %xmm0
+; CHECK-NEXT:    movq %r9, %xmm1
+; CHECK-NEXT:    movq %r8, %xmm0
 ; CHECK-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
-; CHECK-NEXT:    movd %rdx, %xmm2
-; CHECK-NEXT:    movd %rsi, %xmm1
+; CHECK-NEXT:    movq %rdx, %xmm2
+; CHECK-NEXT:    movq %rsi, %xmm1
 ; CHECK-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
-; CHECK-NEXT:    movd %rcx, %xmm2
+; CHECK-NEXT:    movq %rcx, %xmm2
 ; CHECK-NEXT:    movq {{.*#+}} xmm3 = mem[0],zero
 ; CHECK-NEXT:    pxor %xmm4, %xmm4
 ; CHECK-NEXT:    pcmpeqq %xmm4, %xmm2
 ; CHECK-NEXT:    pcmpeqd %xmm5, %xmm5
 ; CHECK-NEXT:    pxor %xmm5, %xmm2
 ; CHECK-NEXT:    pcmpeqq %xmm4, %xmm1
 ; CHECK-NEXT:    pxor %xmm5, %xmm1
 ; CHECK-NEXT:    shufps {{.*#+}} xmm1 = xmm1[0,2],xmm2[0,2]

llvm/trunk/test/CodeGen/X86/widen_cast-5.ll

 ; X86-NEXT:    pmovzxdq {{.*#+}} xmm0 = mem[0],zero,mem[1],zero
 ; X86-NEXT:    pxor LCPI0_0, %xmm0
 ; X86-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
 ; X86-NEXT:    movq %xmm0, (%eax)
 ; X86-NEXT:    retl
 ;
 ; X64-LABEL: convert:
 ; X64:       ## BB#0: ## %entry
-; X64-NEXT:    movd %rsi, %xmm0
+; X64-NEXT:    movq %rsi, %xmm0
 ; X64-NEXT:    pmovzxdq {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero
 ; X64-NEXT:    pxor {{.*}}(%rip), %xmm0
 ; X64-NEXT:    pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
 ; X64-NEXT:    movq %xmm0, (%rdi)
 ; X64-NEXT:    retq
 entry:
   %conv = bitcast i64 %src to <2 x i32>
   %xor = xor <2 x i32> %conv, < i32 255, i32 32767 >
   store <2 x i32> %xor, <2 x i32>* %dst.addr
   ret void
 }

llvm/trunk/test/CodeGen/X86/widen_conv-3.ll

 ; X86-SSE42-NEXT:    extractps $1, %xmm0, 4(%eax)
 ; X86-SSE42-NEXT:    movss %xmm0, (%eax)
 ; X86-SSE42-NEXT:    popl %eax
 ; X86-SSE42-NEXT:    retl
 ;
 ; X64-SSE2-LABEL: convert_v3i8_to_v3f32:
 ; X64-SSE2:       # BB#0: # %entry
 ; X64-SSE2-NEXT:    movzwl (%rsi), %eax
-; X64-SSE2-NEXT:    movd %rax, %xmm0
+; X64-SSE2-NEXT:    movq %rax, %xmm0
 ; X64-SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
 ; X64-SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
 ; X64-SSE2-NEXT:    movdqa %xmm0, -{{[0-9]+}}(%rsp)
 ; X64-SSE2-NEXT:    movl -{{[0-9]+}}(%rsp), %eax
 ; X64-SSE2-NEXT:    shll $8, %eax
 ; X64-SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %ecx
 ; X64-SSE2-NEXT:    orl %eax, %ecx
 ; X64-SSE2-NEXT:    movd %ecx, %xmm0
 ; X64-SSE2-NEXT:    movzbl 2(%rsi), %eax
 ; X64-SSE2-NEXT:    pinsrw $1, %eax, %xmm0
 ; X64-SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
 ; X64-SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
 ; X64-SSE2-NEXT:    psrad $24, %xmm0
 ; X64-SSE2-NEXT:    cvtdq2ps %xmm0, %xmm0
 ; X64-SSE2-NEXT:    movlps %xmm0, (%rdi)
 ; X64-SSE2-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; X64-SSE2-NEXT:    movss %xmm0, 8(%rdi)
 ; X64-SSE2-NEXT:    retq
 ;
 ; X64-SSE42-LABEL: convert_v3i8_to_v3f32:
 ; X64-SSE42:       # BB#0: # %entry
 ; X64-SSE42-NEXT:    movzbl 2(%rsi), %eax
 ; X64-SSE42-NEXT:    movzwl (%rsi), %ecx
-; X64-SSE42-NEXT:    movd %rcx, %xmm0
+; X64-SSE42-NEXT:    movq %rcx, %xmm0
 ; X64-SSE42-NEXT:    pmovzxbd {{.*#+}} xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero
 ; X64-SSE42-NEXT:    pinsrd $2, %eax, %xmm0
 ; X64-SSE42-NEXT:    pslld $24, %xmm0
 ; X64-SSE42-NEXT:    psrad $24, %xmm0
 ; X64-SSE42-NEXT:    cvtdq2ps %xmm0, %xmm0
 ; X64-SSE42-NEXT:    extractps $2, %xmm0, 8(%rdi)
 ; X64-SSE42-NEXT:    movlps %xmm0, (%rdi)
 ; X64-SSE42-NEXT:    retq
 entry:
   %load = load <3 x i8>, <3 x i8>* %src.addr, align 1
   %cvt = sitofp <3 x i8> %load to <3 x float>
   store <3 x float> %cvt, <3 x float>* %dst.addr, align 4
   ret void
 }

llvm/trunk/test/CodeGen/X86/widen_conv-4.ll

 ; X86-SSE42-NEXT:    extractps $1, %xmm0, 4(%eax)
 ; X86-SSE42-NEXT:    movss %xmm0, (%eax)
 ; X86-SSE42-NEXT:    popl %eax
 ; X86-SSE42-NEXT:    retl
 ;
 ; X64-SSE2-LABEL: convert_v3i8_to_v3f32:
 ; X64-SSE2:       # BB#0: # %entry
 ; X64-SSE2-NEXT:    movzwl (%rsi), %eax
-; X64-SSE2-NEXT:    movd %rax, %xmm0
+; X64-SSE2-NEXT:    movq %rax, %xmm0
 ; X64-SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
 ; X64-SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3]
 ; X64-SSE2-NEXT:    movdqa %xmm0, -{{[0-9]+}}(%rsp)
 ; X64-SSE2-NEXT:    movl -{{[0-9]+}}(%rsp), %eax
 ; X64-SSE2-NEXT:    shll $8, %eax
 ; X64-SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %ecx
 ; X64-SSE2-NEXT:    orl %eax, %ecx
 ; X64-SSE2-NEXT:    movd %ecx, %xmm0
 ; X64-SSE2-NEXT:    movzbl 2(%rsi), %eax
 ; X64-SSE2-NEXT:    pinsrw $1, %eax, %xmm0
 ; X64-SSE2-NEXT:    pxor %xmm1, %xmm1
 ; X64-SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
 ; X64-SSE2-NEXT:    punpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
 ; X64-SSE2-NEXT:    cvtdq2ps %xmm0, %xmm0
 ; X64-SSE2-NEXT:    movlps %xmm0, (%rdi)
 ; X64-SSE2-NEXT:    movhlps {{.*#+}} xmm0 = xmm0[1,1]
 ; X64-SSE2-NEXT:    movss %xmm0, 8(%rdi)
 ; X64-SSE2-NEXT:    retq
 ;
 ; X64-SSE42-LABEL: convert_v3i8_to_v3f32:
 ; X64-SSE42:       # BB#0: # %entry
 ; X64-SSE42-NEXT:    movzbl 2(%rsi), %eax
 ; X64-SSE42-NEXT:    movzwl (%rsi), %ecx
-; X64-SSE42-NEXT:    movd %rcx, %xmm0
+; X64-SSE42-NEXT:    movq %rcx, %xmm0
 ; X64-SSE42-NEXT:    pmovzxbd {{.*#+}} xmm0 = xmm0[0],zero,zero,zero,xmm0[1],zero,zero,zero,xmm0[2],zero,zero,zero,xmm0[3],zero,zero,zero
 ; X64-SSE42-NEXT:    pinsrd $2, %eax, %xmm0
 ; X64-SSE42-NEXT:    pand {{.*}}(%rip), %xmm0
 ; X64-SSE42-NEXT:    cvtdq2ps %xmm0, %xmm0
 ; X64-SSE42-NEXT:    extractps $2, %xmm0, 8(%rdi)
 ; X64-SSE42-NEXT:    movlps %xmm0, (%rdi)
 ; X64-SSE42-NEXT:    retq
 entry:
   %load = load <3 x i8>, <3 x i8>* %src.addr, align 1
   %cvt = uitofp <3 x i8> %load to <3 x float>
   store <3 x float> %cvt, <3 x float>* %dst.addr, align 4
   ret void
 }

llvm/trunk/test/MC/Disassembler/X86/x86-64.txt

 0x48 0x0f 0x6e 0xc0
 
 # CHECK: movd %mm0, (%rax)
 0x48 0x0f 0x7e 0x00
 
 # CHECK: movd %mm0, %rax
 0x48 0x0f 0x7e 0xc0
 
-# CHECK: movd (%rax), %xmm0
+# CHECK: movq (%rax), %xmm0
 0x66 0x48 0x0f 0x6e 0x00
 
-# CHECK: movd %rax, %xmm0
+# CHECK: movq %rax, %xmm0
 0x66 0x48 0x0f 0x6e 0xc0
 
-# CHECK: movd %xmm0, (%rax)
+# CHECK: movq %xmm0, (%rax)
 0x66 0x48 0x0f 0x7e 0x00
 
-# CHECK: movd %xmm0, %rax
+# CHECK: movq %xmm0, %rax
 0x66 0x48 0x0f 0x7e 0xc0
 
 # CHECK: pextrw $3, %xmm3, %ecx
 0x66 0x0f 0x3a 0x15 0xd9 0x03
 
 # CHECK: pextrw $3, %xmm3, (%rax)
 0x66 0x0f 0x3a 0x15 0x18 0x03
 

llvm/trunk/test/MC/X86/x86-64.s

 // CHECK: strw
 // CHECK: encoding: [0x66,0x0f,0x00,0xc8]
   str %ax
 
 // CHECK: strq
 // CHECK: encoding: [0x48,0x0f,0x00,0xc8]
   str %rax
 
-// CHECK: movd %rdi, %xmm0
+// CHECK: movq %rdi, %xmm0
 // CHECK: encoding: [0x66,0x48,0x0f,0x6e,0xc7]
   movq %rdi,%xmm0
 
-// CHECK: movd %rdi, %xmm0
-// CHECK: encoding: [0x66,0x48,0x0f,0x6e,0xc7]
-  movd %rdi,%xmm0
-
-// CHECK: movd  %xmm0, %rax
+// CHECK: movq  %xmm0, %rax
 // CHECK: encoding: [0x66,0x48,0x0f,0x7e,0xc0]
-        movd  %xmm0, %rax
+    movq  %xmm0, %rax
 
 // CHECK: movntil %eax, (%rdi)
 // CHECK: encoding: [0x0f,0xc3,0x07]
 // CHECK: movntil
 movntil %eax, (%rdi)
 movnti %eax, (%rdi)
 
 // CHECK: movntiq %rax, (%rdi)

 fadd %st(1)
 fmul %st(1)
 fsub %st(1)
 fsubr %st(1)
 fdiv %st(1)
 fdivr %st(1)
 
 // CHECK: movd %xmm0, %eax
-// CHECK: movd %xmm0, %rax
-// CHECK: movd %xmm0, %rax
+// CHECK: movq %xmm0, %rax
+// CHECK: movq %xmm0, %rax
 // CHECK: vmovd %xmm0, %eax
 // CHECK: vmovq %xmm0, %rax
 // CHECK: vmovq %xmm0, %rax
 movd %xmm0, %eax
-movd %xmm0, %rax
+movq %xmm0, %rax
 movq %xmm0, %rax
 vmovd %xmm0, %eax
 vmovd %xmm0, %rax
 vmovq %xmm0, %rax
 
 // CHECK: seto 3735928559(%r10,%r9,8)
 // CHECK:  encoding: [0x43,0x0f,0x90,0x84,0xca,0xef,0xbe,0xad,0xde]
   seto 0xdeadbeef(%r10,%r9,8)