; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=i386-unknown -mattr=+mmx,+sse2,+ssse3 | FileCheck %s --check-prefix=X86
; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+mmx,+sse2,+ssse3 | FileCheck %s --check-prefixes=X64,ALIGN
; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+mmx,+sse2,+ssse3,sse-unaligned-mem | FileCheck %s --check-prefixes=X64,UNALIGN

; There are no MMX operations in @t1

define void  @t1(i32 %a, ptr %P) nounwind {
; X86-LABEL: t1:
; X86:       # %bb.0:
; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
; X86-NEXT:    movd {{.*#+}} xmm0 = mem[0],zero,zero,zero
; X86-NEXT:    pslld $12, %xmm0
; X86-NEXT:    psllq $32, %xmm0
; X86-NEXT:    movq %xmm0, (%eax)
; X86-NEXT:    retl
;
; X64-LABEL: t1:
; X64:       # %bb.0:
; X64-NEXT:    movd %edi, %xmm0
; X64-NEXT:    pslld $12, %xmm0
; X64-NEXT:    psllq $32, %xmm0
; 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 <1 x i64>
 store <1 x i64> %tmp23, ptr %P
 ret void
}

define <4 x float> @t2(ptr %P) nounwind {
; X86-LABEL: t2:
; X86:       # %bb.0:
; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
; X86-NEXT:    xorps %xmm0, %xmm0
; X86-NEXT:    xorps %xmm1, %xmm1
; X86-NEXT:    unpcklpd {{.*#+}} xmm1 = xmm1[0],mem[0]
; X86-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
; X86-NEXT:    retl
;
; X64-LABEL: t2:
; X64:       # %bb.0:
; X64-NEXT:    xorps %xmm0, %xmm0
; X64-NEXT:    xorps %xmm1, %xmm1
; X64-NEXT:    unpcklpd {{.*#+}} xmm1 = xmm1[0],mem[0]
; X64-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,1],xmm1[0,2]
; X64-NEXT:    retq
  %tmp1 = load <4 x float>, ptr %P
  %tmp2 = shufflevector <4 x float> %tmp1, <4 x float> zeroinitializer, <4 x i32> < i32 4, i32 4, i32 4, i32 0 >
  ret <4 x float> %tmp2
}

define <4 x float> @t3(ptr %P) nounwind {
; X86-LABEL: t3:
; X86:       # %bb.0:
; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
; X86-NEXT:    xorps %xmm0, %xmm0
; X86-NEXT:    movlps {{.*#+}} xmm0 = mem[0,1],xmm0[2,3]
; X86-NEXT:    retl
;
; X64-LABEL: t3:
; X64:       # %bb.0:
; X64-NEXT:    xorps %xmm0, %xmm0
; X64-NEXT:    movlps {{.*#+}} xmm0 = mem[0,1],xmm0[2,3]
; X64-NEXT:    retq
  %tmp1 = load <4 x float>, ptr %P
  %tmp2 = shufflevector <4 x float> %tmp1, <4 x float> zeroinitializer, <4 x i32> < i32 2, i32 3, i32 4, i32 4 >
  ret <4 x float> %tmp2
}

define <4 x float> @t4(ptr %P) nounwind {
; X86-LABEL: t4:
; X86:       # %bb.0:
; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
; X86-NEXT:    xorps %xmm1, %xmm1
; X86-NEXT:    xorps %xmm0, %xmm0
; X86-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,0],mem[3,0]
; X86-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
; X86-NEXT:    retl
;
; X64-LABEL: t4:
; X64:       # %bb.0:
; X64-NEXT:    xorps %xmm1, %xmm1
; X64-NEXT:    xorps %xmm0, %xmm0
; X64-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,0],mem[3,0]
; X64-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
; X64-NEXT:    retq
  %tmp1 = load <4 x float>, ptr %P
  %tmp2 = shufflevector <4 x float> zeroinitializer, <4 x float> %tmp1, <4 x i32> < i32 7, i32 0, i32 0, i32 0 >
  ret <4 x float> %tmp2
}

define <4 x float> @t4_under_aligned(ptr %P) nounwind {
; X86-LABEL: t4_under_aligned:
; X86:       # %bb.0:
; X86-NEXT:    movl {{[0-9]+}}(%esp), %eax
; X86-NEXT:    movups (%eax), %xmm0
; X86-NEXT:    xorps %xmm1, %xmm1
; X86-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[1,0]
; X86-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[2,3]
; X86-NEXT:    retl
;
; ALIGN-LABEL: t4_under_aligned:
; ALIGN:       # %bb.0:
; ALIGN-NEXT:    movups (%rdi), %xmm0
; ALIGN-NEXT:    xorps %xmm1, %xmm1
; ALIGN-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,0],xmm1[1,0]
; ALIGN-NEXT:    shufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[2,3]
; ALIGN-NEXT:    retq
;
; UNALIGN-LABEL: t4_under_aligned:
; UNALIGN:       # %bb.0:
; UNALIGN-NEXT:    xorps %xmm1, %xmm1
; UNALIGN-NEXT:    xorps %xmm0, %xmm0
; UNALIGN-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,0],mem[3,0]
; UNALIGN-NEXT:    shufps {{.*#+}} xmm0 = xmm0[2,0],xmm1[2,3]
; UNALIGN-NEXT:    retq
  %tmp1 = load <4 x float>, ptr %P, align 4
  %tmp2 = shufflevector <4 x float> zeroinitializer, <4 x float> %tmp1, <4 x i32> < i32 7, i32 0, i32 0, i32 0 >
  ret <4 x float> %tmp2
}

define <16 x i8> @t5(<16 x i8> %x) nounwind {
; X86-LABEL: t5:
; X86:       # %bb.0:
; X86-NEXT:    psrlw $8, %xmm0
; X86-NEXT:    retl
;
; X64-LABEL: t5:
; X64:       # %bb.0:
; X64-NEXT:    psrlw $8, %xmm0
; X64-NEXT:    retq
  %s = shufflevector <16 x i8> %x, <16 x i8> zeroinitializer, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 17>
  ret <16 x i8> %s
}

define <16 x i8> @t6(<16 x i8> %x) nounwind {
; X86-LABEL: t6:
; X86:       # %bb.0:
; X86-NEXT:    psrlw $8, %xmm0
; X86-NEXT:    retl
;
; X64-LABEL: t6:
; X64:       # %bb.0:
; X64-NEXT:    psrlw $8, %xmm0
; X64-NEXT:    retq
  %s = shufflevector <16 x i8> %x, <16 x i8> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
  ret <16 x i8> %s
}

define <16 x i8> @t7(<16 x i8> %x) nounwind {
; X86-LABEL: t7:
; X86:       # %bb.0:
; X86-NEXT:    pslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2]
; X86-NEXT:    retl
;
; X64-LABEL: t7:
; X64:       # %bb.0:
; X64-NEXT:    pslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2]
; X64-NEXT:    retq
  %s = shufflevector <16 x i8> %x, <16 x i8> undef, <16 x i32> <i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 1, i32 2>
  ret <16 x i8> %s
}

define <16 x i8> @t8(<16 x i8> %x) nounwind {
; X86-LABEL: t8:
; X86:       # %bb.0:
; X86-NEXT:    psrldq {{.*#+}} xmm0 = xmm0[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zero
; X86-NEXT:    retl
;
; X64-LABEL: t8:
; X64:       # %bb.0:
; X64-NEXT:    psrldq {{.*#+}} xmm0 = xmm0[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zero
; X64-NEXT:    retq
  %s = shufflevector <16 x i8> %x, <16 x i8> zeroinitializer, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 8, i32 9, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 17>
  ret <16 x i8> %s
}

define <16 x i8> @t9(<16 x i8> %x) nounwind {
; X86-LABEL: t9:
; X86:       # %bb.0:
; X86-NEXT:    psrldq {{.*#+}} xmm0 = xmm0[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zero
; X86-NEXT:    retl
;
; X64-LABEL: t9:
; X64:       # %bb.0:
; X64-NEXT:    psrldq {{.*#+}} xmm0 = xmm0[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zero
; X64-NEXT:    retq
  %s = shufflevector <16 x i8> %x, <16 x i8> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 7, i32 8, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 14, i32 undef, i32 undef>
  ret <16 x i8> %s
}