; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -disable-peephole -mtriple=i386-apple-darwin -mattr=+sse2 -show-mc-encoding | FileCheck %s --check-prefixes=CHECK,X86,SSE,X86-SSE
; RUN: llc < %s -disable-peephole -mtriple=i386-apple-darwin -mattr=+avx -show-mc-encoding | FileCheck %s --check-prefixes=CHECK,X86,AVX,X86-AVX,AVX1,X86-AVX1
; RUN: llc < %s -disable-peephole -mtriple=i386-apple-darwin -mattr=+avx512f,+avx512bw,+avx512dq,+avx512vl -show-mc-encoding | FileCheck %s --check-prefixes=CHECK,X86,AVX,X86-AVX,AVX512,X86-AVX512
; RUN: llc < %s -disable-peephole -mtriple=x86_64-apple-darwin -mattr=+sse2 -show-mc-encoding | FileCheck %s --check-prefixes=CHECK,X64,SSE,X64-SSE
; RUN: llc < %s -disable-peephole -mtriple=x86_64-apple-darwin -mattr=+avx -show-mc-encoding | FileCheck %s --check-prefixes=CHECK,X64,AVX,X64-AVX,AVX1,X64-AVX1
; RUN: llc < %s -disable-peephole -mtriple=x86_64-apple-darwin -mattr=+avx512f,+avx512bw,+avx512dq,+avx512vl -show-mc-encoding | FileCheck %s --check-prefixes=CHECK,X64,AVX,X64-AVX,AVX512,X64-AVX512


define <2 x double> @test_x86_sse2_sqrt_pd(<2 x double> %a0) {
; SSE-LABEL: test_x86_sse2_sqrt_pd:
; SSE:       ## %bb.0:
; SSE-NEXT:    sqrtpd %xmm0, %xmm0 ## encoding: [0x66,0x0f,0x51,0xc0]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_sqrt_pd:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vsqrtpd %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0x51,0xc0]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_sqrt_pd:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vsqrtpd %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0x51,0xc0]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x double> @llvm.x86.sse2.sqrt.pd(<2 x double> %a0) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}
declare <2 x double> @llvm.x86.sse2.sqrt.pd(<2 x double>) nounwind readnone


define <2 x double> @test_x86_sse2_sqrt_sd(<2 x double> %a0) {
; SSE-LABEL: test_x86_sse2_sqrt_sd:
; SSE:       ## %bb.0:
; SSE-NEXT:    sqrtsd %xmm0, %xmm0 ## encoding: [0xf2,0x0f,0x51,0xc0]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_sqrt_sd:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vsqrtsd %xmm0, %xmm0, %xmm0 ## encoding: [0xc5,0xfb,0x51,0xc0]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_sqrt_sd:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vsqrtsd %xmm0, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x51,0xc0]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x double> @llvm.x86.sse2.sqrt.sd(<2 x double> %a0) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}
declare <2 x double> @llvm.x86.sse2.sqrt.sd(<2 x double>) nounwind readnone


define <2 x double> @test_x86_sse2_sqrt_sd_vec_load(<2 x double>* %a0) {
; X86-SSE-LABEL: test_x86_sse2_sqrt_sd_vec_load:
; X86-SSE:       ## %bb.0:
; X86-SSE-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-SSE-NEXT:    movapd (%eax), %xmm0 ## encoding: [0x66,0x0f,0x28,0x00]
; X86-SSE-NEXT:    sqrtsd %xmm0, %xmm0 ## encoding: [0xf2,0x0f,0x51,0xc0]
; X86-SSE-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX1-LABEL: test_x86_sse2_sqrt_sd_vec_load:
; X86-AVX1:       ## %bb.0:
; X86-AVX1-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-AVX1-NEXT:    vmovapd (%eax), %xmm0 ## encoding: [0xc5,0xf9,0x28,0x00]
; X86-AVX1-NEXT:    vsqrtsd %xmm0, %xmm0, %xmm0 ## encoding: [0xc5,0xfb,0x51,0xc0]
; X86-AVX1-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX512-LABEL: test_x86_sse2_sqrt_sd_vec_load:
; X86-AVX512:       ## %bb.0:
; X86-AVX512-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-AVX512-NEXT:    vmovapd (%eax), %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0x28,0x00]
; X86-AVX512-NEXT:    vsqrtsd %xmm0, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x51,0xc0]
; X86-AVX512-NEXT:    retl ## encoding: [0xc3]
;
; X64-SSE-LABEL: test_x86_sse2_sqrt_sd_vec_load:
; X64-SSE:       ## %bb.0:
; X64-SSE-NEXT:    movapd (%rdi), %xmm0 ## encoding: [0x66,0x0f,0x28,0x07]
; X64-SSE-NEXT:    sqrtsd %xmm0, %xmm0 ## encoding: [0xf2,0x0f,0x51,0xc0]
; X64-SSE-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX1-LABEL: test_x86_sse2_sqrt_sd_vec_load:
; X64-AVX1:       ## %bb.0:
; X64-AVX1-NEXT:    vmovapd (%rdi), %xmm0 ## encoding: [0xc5,0xf9,0x28,0x07]
; X64-AVX1-NEXT:    vsqrtsd %xmm0, %xmm0, %xmm0 ## encoding: [0xc5,0xfb,0x51,0xc0]
; X64-AVX1-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX512-LABEL: test_x86_sse2_sqrt_sd_vec_load:
; X64-AVX512:       ## %bb.0:
; X64-AVX512-NEXT:    vmovapd (%rdi), %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0x28,0x07]
; X64-AVX512-NEXT:    vsqrtsd %xmm0, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x51,0xc0]
; X64-AVX512-NEXT:    retq ## encoding: [0xc3]
  %a1 = load <2 x double>, <2 x double>* %a0, align 16
  %res = call <2 x double> @llvm.x86.sse2.sqrt.sd(<2 x double> %a1) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}


define <2 x i64> @test_x86_sse2_psll_dq_bs(<2 x i64> %a0) {
; SSE-LABEL: test_x86_sse2_psll_dq_bs:
; SSE:       ## %bb.0:
; SSE-NEXT:    pslldq $7, %xmm0 ## encoding: [0x66,0x0f,0x73,0xf8,0x07]
; SSE-NEXT:    ## xmm0 = zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4,5,6,7,8]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_psll_dq_bs:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vpslldq $7, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0x73,0xf8,0x07]
; AVX1-NEXT:    ## xmm0 = zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4,5,6,7,8]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_psll_dq_bs:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vpslldq $7, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0x73,0xf8,0x07]
; AVX512-NEXT:    ## xmm0 = zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4,5,6,7,8]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x i64> @llvm.x86.sse2.psll.dq.bs(<2 x i64> %a0, i32 7) ; <<2 x i64>> [#uses=1]
  ret <2 x i64> %res
}
declare <2 x i64> @llvm.x86.sse2.psll.dq.bs(<2 x i64>, i32) nounwind readnone


define <2 x i64> @test_x86_sse2_psrl_dq_bs(<2 x i64> %a0) {
; SSE-LABEL: test_x86_sse2_psrl_dq_bs:
; SSE:       ## %bb.0:
; SSE-NEXT:    psrldq $7, %xmm0 ## encoding: [0x66,0x0f,0x73,0xd8,0x07]
; SSE-NEXT:    ## xmm0 = xmm0[7,8,9,10,11,12,13,14,15],zero,zero,zero,zero,zero,zero,zero
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_psrl_dq_bs:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vpsrldq $7, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0x73,0xd8,0x07]
; AVX1-NEXT:    ## xmm0 = xmm0[7,8,9,10,11,12,13,14,15],zero,zero,zero,zero,zero,zero,zero
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_psrl_dq_bs:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vpsrldq $7, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0x73,0xd8,0x07]
; AVX512-NEXT:    ## xmm0 = xmm0[7,8,9,10,11,12,13,14,15],zero,zero,zero,zero,zero,zero,zero
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x i64> @llvm.x86.sse2.psrl.dq.bs(<2 x i64> %a0, i32 7) ; <<2 x i64>> [#uses=1]
  ret <2 x i64> %res
}
declare <2 x i64> @llvm.x86.sse2.psrl.dq.bs(<2 x i64>, i32) nounwind readnone

define <2 x i64> @test_x86_sse2_psll_dq(<2 x i64> %a0) {
; SSE-LABEL: test_x86_sse2_psll_dq:
; SSE:       ## %bb.0:
; SSE-NEXT:    pslldq $1, %xmm0 ## encoding: [0x66,0x0f,0x73,0xf8,0x01]
; SSE-NEXT:    ## xmm0 = zero,xmm0[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_psll_dq:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vpslldq $1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0x73,0xf8,0x01]
; AVX1-NEXT:    ## xmm0 = zero,xmm0[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_psll_dq:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vpslldq $1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0x73,0xf8,0x01]
; AVX512-NEXT:    ## xmm0 = zero,xmm0[0,1,2,3,4,5,6,7,8,9,10,11,12,13,14]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x i64> @llvm.x86.sse2.psll.dq(<2 x i64> %a0, i32 8) ; <<2 x i64>> [#uses=1]
  ret <2 x i64> %res
}
declare <2 x i64> @llvm.x86.sse2.psll.dq(<2 x i64>, i32) nounwind readnone


define <2 x i64> @test_x86_sse2_psrl_dq(<2 x i64> %a0) {
; SSE-LABEL: test_x86_sse2_psrl_dq:
; SSE:       ## %bb.0:
; SSE-NEXT:    psrldq $1, %xmm0 ## encoding: [0x66,0x0f,0x73,0xd8,0x01]
; SSE-NEXT:    ## xmm0 = xmm0[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zero
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_psrl_dq:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vpsrldq $1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0x73,0xd8,0x01]
; AVX1-NEXT:    ## xmm0 = xmm0[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zero
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_psrl_dq:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vpsrldq $1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0x73,0xd8,0x01]
; AVX512-NEXT:    ## xmm0 = xmm0[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15],zero
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x i64> @llvm.x86.sse2.psrl.dq(<2 x i64> %a0, i32 8) ; <<2 x i64>> [#uses=1]
  ret <2 x i64> %res
}
declare <2 x i64> @llvm.x86.sse2.psrl.dq(<2 x i64>, i32) nounwind readnone


define <2 x double> @test_x86_sse2_cvtdq2pd(<4 x i32> %a0) {
; SSE-LABEL: test_x86_sse2_cvtdq2pd:
; SSE:       ## %bb.0:
; SSE-NEXT:    cvtdq2pd %xmm0, %xmm0 ## encoding: [0xf3,0x0f,0xe6,0xc0]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_cvtdq2pd:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vcvtdq2pd %xmm0, %xmm0 ## encoding: [0xc5,0xfa,0xe6,0xc0]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_cvtdq2pd:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vcvtdq2pd %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfa,0xe6,0xc0]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x double> @llvm.x86.sse2.cvtdq2pd(<4 x i32> %a0) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}
declare <2 x double> @llvm.x86.sse2.cvtdq2pd(<4 x i32>) nounwind readnone


define <2 x double> @test_x86_sse2_cvtps2pd(<4 x float> %a0) {
; SSE-LABEL: test_x86_sse2_cvtps2pd:
; SSE:       ## %bb.0:
; SSE-NEXT:    cvtps2pd %xmm0, %xmm0 ## encoding: [0x0f,0x5a,0xc0]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_cvtps2pd:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vcvtps2pd %xmm0, %xmm0 ## encoding: [0xc5,0xf8,0x5a,0xc0]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_cvtps2pd:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vcvtps2pd %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf8,0x5a,0xc0]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x double> @llvm.x86.sse2.cvtps2pd(<4 x float> %a0) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}
declare <2 x double> @llvm.x86.sse2.cvtps2pd(<4 x float>) nounwind readnone


define void @test_x86_sse2_storel_dq(i8* %a0, <4 x i32> %a1) {
; X86-SSE-LABEL: test_x86_sse2_storel_dq:
; X86-SSE:       ## %bb.0:
; X86-SSE-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-SSE-NEXT:    movlps %xmm0, (%eax) ## encoding: [0x0f,0x13,0x00]
; X86-SSE-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX1-LABEL: test_x86_sse2_storel_dq:
; X86-AVX1:       ## %bb.0:
; X86-AVX1-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-AVX1-NEXT:    vmovlps %xmm0, (%eax) ## encoding: [0xc5,0xf8,0x13,0x00]
; X86-AVX1-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX512-LABEL: test_x86_sse2_storel_dq:
; X86-AVX512:       ## %bb.0:
; X86-AVX512-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-AVX512-NEXT:    vmovlps %xmm0, (%eax) ## EVEX TO VEX Compression encoding: [0xc5,0xf8,0x13,0x00]
; X86-AVX512-NEXT:    retl ## encoding: [0xc3]
;
; X64-SSE-LABEL: test_x86_sse2_storel_dq:
; X64-SSE:       ## %bb.0:
; X64-SSE-NEXT:    movlps %xmm0, (%rdi) ## encoding: [0x0f,0x13,0x07]
; X64-SSE-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX1-LABEL: test_x86_sse2_storel_dq:
; X64-AVX1:       ## %bb.0:
; X64-AVX1-NEXT:    vmovlps %xmm0, (%rdi) ## encoding: [0xc5,0xf8,0x13,0x07]
; X64-AVX1-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX512-LABEL: test_x86_sse2_storel_dq:
; X64-AVX512:       ## %bb.0:
; X64-AVX512-NEXT:    vmovlps %xmm0, (%rdi) ## EVEX TO VEX Compression encoding: [0xc5,0xf8,0x13,0x07]
; X64-AVX512-NEXT:    retq ## encoding: [0xc3]
  call void @llvm.x86.sse2.storel.dq(i8* %a0, <4 x i32> %a1)
  ret void
}
declare void @llvm.x86.sse2.storel.dq(i8*, <4 x i32>) nounwind


define void @test_x86_sse2_storeu_dq(i8* %a0, <16 x i8> %a1) {
  ; add operation forces the execution domain.
; X86-SSE-LABEL: test_x86_sse2_storeu_dq:
; X86-SSE:       ## %bb.0:
; X86-SSE-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-SSE-NEXT:    pcmpeqd %xmm1, %xmm1 ## encoding: [0x66,0x0f,0x76,0xc9]
; X86-SSE-NEXT:    psubb %xmm1, %xmm0 ## encoding: [0x66,0x0f,0xf8,0xc1]
; X86-SSE-NEXT:    movdqu %xmm0, (%eax) ## encoding: [0xf3,0x0f,0x7f,0x00]
; X86-SSE-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX1-LABEL: test_x86_sse2_storeu_dq:
; X86-AVX1:       ## %bb.0:
; X86-AVX1-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-AVX1-NEXT:    vpcmpeqd %xmm1, %xmm1, %xmm1 ## encoding: [0xc5,0xf1,0x76,0xc9]
; X86-AVX1-NEXT:    vpsubb %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0xf8,0xc1]
; X86-AVX1-NEXT:    vmovdqu %xmm0, (%eax) ## encoding: [0xc5,0xfa,0x7f,0x00]
; X86-AVX1-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX512-LABEL: test_x86_sse2_storeu_dq:
; X86-AVX512:       ## %bb.0:
; X86-AVX512-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-AVX512-NEXT:    vpcmpeqd %xmm1, %xmm1, %xmm1 ## encoding: [0xc5,0xf1,0x76,0xc9]
; X86-AVX512-NEXT:    vpsubb %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0xf8,0xc1]
; X86-AVX512-NEXT:    vmovdqu %xmm0, (%eax) ## EVEX TO VEX Compression encoding: [0xc5,0xfa,0x7f,0x00]
; X86-AVX512-NEXT:    retl ## encoding: [0xc3]
;
; X64-SSE-LABEL: test_x86_sse2_storeu_dq:
; X64-SSE:       ## %bb.0:
; X64-SSE-NEXT:    pcmpeqd %xmm1, %xmm1 ## encoding: [0x66,0x0f,0x76,0xc9]
; X64-SSE-NEXT:    psubb %xmm1, %xmm0 ## encoding: [0x66,0x0f,0xf8,0xc1]
; X64-SSE-NEXT:    movdqu %xmm0, (%rdi) ## encoding: [0xf3,0x0f,0x7f,0x07]
; X64-SSE-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX1-LABEL: test_x86_sse2_storeu_dq:
; X64-AVX1:       ## %bb.0:
; X64-AVX1-NEXT:    vpcmpeqd %xmm1, %xmm1, %xmm1 ## encoding: [0xc5,0xf1,0x76,0xc9]
; X64-AVX1-NEXT:    vpsubb %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0xf8,0xc1]
; X64-AVX1-NEXT:    vmovdqu %xmm0, (%rdi) ## encoding: [0xc5,0xfa,0x7f,0x07]
; X64-AVX1-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX512-LABEL: test_x86_sse2_storeu_dq:
; X64-AVX512:       ## %bb.0:
; X64-AVX512-NEXT:    vpcmpeqd %xmm1, %xmm1, %xmm1 ## encoding: [0xc5,0xf1,0x76,0xc9]
; X64-AVX512-NEXT:    vpsubb %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0xf8,0xc1]
; X64-AVX512-NEXT:    vmovdqu %xmm0, (%rdi) ## EVEX TO VEX Compression encoding: [0xc5,0xfa,0x7f,0x07]
; X64-AVX512-NEXT:    retq ## encoding: [0xc3]
  %a2 = add <16 x i8> %a1, <i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1, i8 1>
  call void @llvm.x86.sse2.storeu.dq(i8* %a0, <16 x i8> %a2)
  ret void
}
declare void @llvm.x86.sse2.storeu.dq(i8*, <16 x i8>) nounwind


define void @test_x86_sse2_storeu_pd(i8* %a0, <2 x double> %a1) {
  ; fadd operation forces the execution domain.
; X86-SSE-LABEL: test_x86_sse2_storeu_pd:
; X86-SSE:       ## %bb.0:
; X86-SSE-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-SSE-NEXT:    xorpd %xmm1, %xmm1 ## encoding: [0x66,0x0f,0x57,0xc9]
; X86-SSE-NEXT:    movhpd LCPI11_0, %xmm1 ## encoding: [0x66,0x0f,0x16,0x0d,A,A,A,A]
; X86-SSE-NEXT:    ## fixup A - offset: 4, value: LCPI11_0, kind: FK_Data_4
; X86-SSE-NEXT:    ## xmm1 = xmm1[0],mem[0]
; X86-SSE-NEXT:    addpd %xmm0, %xmm1 ## encoding: [0x66,0x0f,0x58,0xc8]
; X86-SSE-NEXT:    movupd %xmm1, (%eax) ## encoding: [0x66,0x0f,0x11,0x08]
; X86-SSE-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX1-LABEL: test_x86_sse2_storeu_pd:
; X86-AVX1:       ## %bb.0:
; X86-AVX1-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-AVX1-NEXT:    vxorpd %xmm1, %xmm1, %xmm1 ## encoding: [0xc5,0xf1,0x57,0xc9]
; X86-AVX1-NEXT:    vmovhpd LCPI11_0, %xmm1, %xmm1 ## encoding: [0xc5,0xf1,0x16,0x0d,A,A,A,A]
; X86-AVX1-NEXT:    ## fixup A - offset: 4, value: LCPI11_0, kind: FK_Data_4
; X86-AVX1-NEXT:    ## xmm1 = xmm1[0],mem[0]
; X86-AVX1-NEXT:    vaddpd %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0x58,0xc1]
; X86-AVX1-NEXT:    vmovupd %xmm0, (%eax) ## encoding: [0xc5,0xf9,0x11,0x00]
; X86-AVX1-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX512-LABEL: test_x86_sse2_storeu_pd:
; X86-AVX512:       ## %bb.0:
; X86-AVX512-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-AVX512-NEXT:    vmovsd LCPI11_0, %xmm1 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x10,0x0d,A,A,A,A]
; X86-AVX512-NEXT:    ## fixup A - offset: 4, value: LCPI11_0, kind: FK_Data_4
; X86-AVX512-NEXT:    ## xmm1 = mem[0],zero
; X86-AVX512-NEXT:    vpslldq $8, %xmm1, %xmm1 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0x73,0xf9,0x08]
; X86-AVX512-NEXT:    ## xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0,1,2,3,4,5,6,7]
; X86-AVX512-NEXT:    vaddpd %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0x58,0xc1]
; X86-AVX512-NEXT:    vmovupd %xmm0, (%eax) ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0x11,0x00]
; X86-AVX512-NEXT:    retl ## encoding: [0xc3]
;
; X64-SSE-LABEL: test_x86_sse2_storeu_pd:
; X64-SSE:       ## %bb.0:
; X64-SSE-NEXT:    xorpd %xmm1, %xmm1 ## encoding: [0x66,0x0f,0x57,0xc9]
; X64-SSE-NEXT:    movhpd {{.*}}(%rip), %xmm1 ## encoding: [0x66,0x0f,0x16,0x0d,A,A,A,A]
; X64-SSE-NEXT:    ## fixup A - offset: 4, value: LCPI11_0-4, kind: reloc_riprel_4byte
; X64-SSE-NEXT:    ## xmm1 = xmm1[0],mem[0]
; X64-SSE-NEXT:    addpd %xmm0, %xmm1 ## encoding: [0x66,0x0f,0x58,0xc8]
; X64-SSE-NEXT:    movupd %xmm1, (%rdi) ## encoding: [0x66,0x0f,0x11,0x0f]
; X64-SSE-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX1-LABEL: test_x86_sse2_storeu_pd:
; X64-AVX1:       ## %bb.0:
; X64-AVX1-NEXT:    vxorpd %xmm1, %xmm1, %xmm1 ## encoding: [0xc5,0xf1,0x57,0xc9]
; X64-AVX1-NEXT:    vmovhpd {{.*}}(%rip), %xmm1, %xmm1 ## encoding: [0xc5,0xf1,0x16,0x0d,A,A,A,A]
; X64-AVX1-NEXT:    ## fixup A - offset: 4, value: LCPI11_0-4, kind: reloc_riprel_4byte
; X64-AVX1-NEXT:    ## xmm1 = xmm1[0],mem[0]
; X64-AVX1-NEXT:    vaddpd %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0x58,0xc1]
; X64-AVX1-NEXT:    vmovupd %xmm0, (%rdi) ## encoding: [0xc5,0xf9,0x11,0x07]
; X64-AVX1-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX512-LABEL: test_x86_sse2_storeu_pd:
; X64-AVX512:       ## %bb.0:
; X64-AVX512-NEXT:    vmovsd {{.*}}(%rip), %xmm1 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x10,0x0d,A,A,A,A]
; X64-AVX512-NEXT:    ## fixup A - offset: 4, value: LCPI11_0-4, kind: reloc_riprel_4byte
; X64-AVX512-NEXT:    ## xmm1 = mem[0],zero
; X64-AVX512-NEXT:    vpslldq $8, %xmm1, %xmm1 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0x73,0xf9,0x08]
; X64-AVX512-NEXT:    ## xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0,1,2,3,4,5,6,7]
; X64-AVX512-NEXT:    vaddpd %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0x58,0xc1]
; X64-AVX512-NEXT:    vmovupd %xmm0, (%rdi) ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0x11,0x07]
; X64-AVX512-NEXT:    retq ## encoding: [0xc3]
  %a2 = fadd <2 x double> %a1, <double 0x0, double 0x4200000000000000>
  call void @llvm.x86.sse2.storeu.pd(i8* %a0, <2 x double> %a2)
  ret void
}
declare void @llvm.x86.sse2.storeu.pd(i8*, <2 x double>) nounwind

define <4 x i32> @test_x86_sse2_pshuf_d(<4 x i32> %a) {
; SSE-LABEL: test_x86_sse2_pshuf_d:
; SSE:       ## %bb.0: ## %entry
; SSE-NEXT:    pshufd $27, %xmm0, %xmm0 ## encoding: [0x66,0x0f,0x70,0xc0,0x1b]
; SSE-NEXT:    ## xmm0 = xmm0[3,2,1,0]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_pshuf_d:
; AVX1:       ## %bb.0: ## %entry
; AVX1-NEXT:    vpermilps $27, %xmm0, %xmm0 ## encoding: [0xc4,0xe3,0x79,0x04,0xc0,0x1b]
; AVX1-NEXT:    ## xmm0 = xmm0[3,2,1,0]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_pshuf_d:
; AVX512:       ## %bb.0: ## %entry
; AVX512-NEXT:    vpermilps $27, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc4,0xe3,0x79,0x04,0xc0,0x1b]
; AVX512-NEXT:    ## xmm0 = xmm0[3,2,1,0]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
entry:
  %res = call <4 x i32> @llvm.x86.sse2.pshuf.d(<4 x i32> %a, i8 27) nounwind readnone
  ret <4 x i32> %res
}
declare <4 x i32> @llvm.x86.sse2.pshuf.d(<4 x i32>, i8) nounwind readnone

define <8 x i16> @test_x86_sse2_pshufl_w(<8 x i16> %a) {
; SSE-LABEL: test_x86_sse2_pshufl_w:
; SSE:       ## %bb.0: ## %entry
; SSE-NEXT:    pshuflw $27, %xmm0, %xmm0 ## encoding: [0xf2,0x0f,0x70,0xc0,0x1b]
; SSE-NEXT:    ## xmm0 = xmm0[3,2,1,0,4,5,6,7]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_pshufl_w:
; AVX1:       ## %bb.0: ## %entry
; AVX1-NEXT:    vpshuflw $27, %xmm0, %xmm0 ## encoding: [0xc5,0xfb,0x70,0xc0,0x1b]
; AVX1-NEXT:    ## xmm0 = xmm0[3,2,1,0,4,5,6,7]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_pshufl_w:
; AVX512:       ## %bb.0: ## %entry
; AVX512-NEXT:    vpshuflw $27, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x70,0xc0,0x1b]
; AVX512-NEXT:    ## xmm0 = xmm0[3,2,1,0,4,5,6,7]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
entry:
  %res = call <8 x i16> @llvm.x86.sse2.pshufl.w(<8 x i16> %a, i8 27) nounwind readnone
  ret <8 x i16> %res
}
declare <8 x i16> @llvm.x86.sse2.pshufl.w(<8 x i16>, i8) nounwind readnone

define <8 x i16> @test_x86_sse2_pshufh_w(<8 x i16> %a) {
; SSE-LABEL: test_x86_sse2_pshufh_w:
; SSE:       ## %bb.0: ## %entry
; SSE-NEXT:    pshufhw $27, %xmm0, %xmm0 ## encoding: [0xf3,0x0f,0x70,0xc0,0x1b]
; SSE-NEXT:    ## xmm0 = xmm0[0,1,2,3,7,6,5,4]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_pshufh_w:
; AVX1:       ## %bb.0: ## %entry
; AVX1-NEXT:    vpshufhw $27, %xmm0, %xmm0 ## encoding: [0xc5,0xfa,0x70,0xc0,0x1b]
; AVX1-NEXT:    ## xmm0 = xmm0[0,1,2,3,7,6,5,4]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_pshufh_w:
; AVX512:       ## %bb.0: ## %entry
; AVX512-NEXT:    vpshufhw $27, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfa,0x70,0xc0,0x1b]
; AVX512-NEXT:    ## xmm0 = xmm0[0,1,2,3,7,6,5,4]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
entry:
  %res = call <8 x i16> @llvm.x86.sse2.pshufh.w(<8 x i16> %a, i8 27) nounwind readnone
  ret <8 x i16> %res
}
declare <8 x i16> @llvm.x86.sse2.pshufh.w(<8 x i16>, i8) nounwind readnone

define <16 x i8> @max_epu8(<16 x i8> %a0, <16 x i8> %a1) {
; SSE-LABEL: max_epu8:
; SSE:       ## %bb.0:
; SSE-NEXT:    pmaxub %xmm1, %xmm0 ## encoding: [0x66,0x0f,0xde,0xc1]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: max_epu8:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vpmaxub %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0xde,0xc1]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: max_epu8:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vpmaxub %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0xde,0xc1]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <16 x i8> @llvm.x86.sse2.pmaxu.b(<16 x i8> %a0, <16 x i8> %a1)
  ret <16 x i8> %res
}
declare <16 x i8> @llvm.x86.sse2.pmaxu.b(<16 x i8>, <16 x i8>) nounwind readnone

define <16 x i8> @min_epu8(<16 x i8> %a0, <16 x i8> %a1) {
; SSE-LABEL: min_epu8:
; SSE:       ## %bb.0:
; SSE-NEXT:    pminub %xmm1, %xmm0 ## encoding: [0x66,0x0f,0xda,0xc1]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: min_epu8:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vpminub %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0xda,0xc1]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: min_epu8:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vpminub %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0xda,0xc1]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <16 x i8> @llvm.x86.sse2.pminu.b(<16 x i8> %a0, <16 x i8> %a1)
  ret <16 x i8> %res
}
declare <16 x i8> @llvm.x86.sse2.pminu.b(<16 x i8>, <16 x i8>) nounwind readnone

define <8 x i16> @max_epi16(<8 x i16> %a0, <8 x i16> %a1) {
; SSE-LABEL: max_epi16:
; SSE:       ## %bb.0:
; SSE-NEXT:    pmaxsw %xmm1, %xmm0 ## encoding: [0x66,0x0f,0xee,0xc1]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: max_epi16:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vpmaxsw %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0xee,0xc1]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: max_epi16:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vpmaxsw %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0xee,0xc1]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <8 x i16> @llvm.x86.sse2.pmaxs.w(<8 x i16> %a0, <8 x i16> %a1)
  ret <8 x i16> %res
}
declare <8 x i16> @llvm.x86.sse2.pmaxs.w(<8 x i16>, <8 x i16>) nounwind readnone

define <8 x i16> @min_epi16(<8 x i16> %a0, <8 x i16> %a1) {
; SSE-LABEL: min_epi16:
; SSE:       ## %bb.0:
; SSE-NEXT:    pminsw %xmm1, %xmm0 ## encoding: [0x66,0x0f,0xea,0xc1]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: min_epi16:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vpminsw %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0xea,0xc1]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: min_epi16:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vpminsw %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0xea,0xc1]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <8 x i16> @llvm.x86.sse2.pmins.w(<8 x i16> %a0, <8 x i16> %a1)
  ret <8 x i16> %res
}
declare <8 x i16> @llvm.x86.sse2.pmins.w(<8 x i16>, <8 x i16>) nounwind readnone

define <2 x double> @test_x86_sse2_add_sd(<2 x double> %a0, <2 x double> %a1) {
; SSE-LABEL: test_x86_sse2_add_sd:
; SSE:       ## %bb.0:
; SSE-NEXT:    addsd %xmm1, %xmm0 ## encoding: [0xf2,0x0f,0x58,0xc1]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_add_sd:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vaddsd %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xfb,0x58,0xc1]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_add_sd:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vaddsd %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x58,0xc1]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x double> @llvm.x86.sse2.add.sd(<2 x double> %a0, <2 x double> %a1) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}
declare <2 x double> @llvm.x86.sse2.add.sd(<2 x double>, <2 x double>) nounwind readnone


define <2 x double> @test_x86_sse2_sub_sd(<2 x double> %a0, <2 x double> %a1) {
; SSE-LABEL: test_x86_sse2_sub_sd:
; SSE:       ## %bb.0:
; SSE-NEXT:    subsd %xmm1, %xmm0 ## encoding: [0xf2,0x0f,0x5c,0xc1]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_sub_sd:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vsubsd %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xfb,0x5c,0xc1]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_sub_sd:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vsubsd %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x5c,0xc1]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x double> @llvm.x86.sse2.sub.sd(<2 x double> %a0, <2 x double> %a1) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}
declare <2 x double> @llvm.x86.sse2.sub.sd(<2 x double>, <2 x double>) nounwind readnone


define <2 x double> @test_x86_sse2_mul_sd(<2 x double> %a0, <2 x double> %a1) {
; SSE-LABEL: test_x86_sse2_mul_sd:
; SSE:       ## %bb.0:
; SSE-NEXT:    mulsd %xmm1, %xmm0 ## encoding: [0xf2,0x0f,0x59,0xc1]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_mul_sd:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vmulsd %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xfb,0x59,0xc1]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_mul_sd:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vmulsd %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x59,0xc1]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x double> @llvm.x86.sse2.mul.sd(<2 x double> %a0, <2 x double> %a1) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}
declare <2 x double> @llvm.x86.sse2.mul.sd(<2 x double>, <2 x double>) nounwind readnone


define <2 x double> @test_x86_sse2_div_sd(<2 x double> %a0, <2 x double> %a1) {
; SSE-LABEL: test_x86_sse2_div_sd:
; SSE:       ## %bb.0:
; SSE-NEXT:    divsd %xmm1, %xmm0 ## encoding: [0xf2,0x0f,0x5e,0xc1]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_div_sd:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vdivsd %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xfb,0x5e,0xc1]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_div_sd:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vdivsd %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x5e,0xc1]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x double> @llvm.x86.sse2.div.sd(<2 x double> %a0, <2 x double> %a1) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}
declare <2 x double> @llvm.x86.sse2.div.sd(<2 x double>, <2 x double>) nounwind readnone

define <16 x i8> @mm_avg_epu8(<16 x i8> %a0, <16 x i8> %a1) {
; SSE-LABEL: mm_avg_epu8:
; SSE:       ## %bb.0:
; SSE-NEXT:    pavgb %xmm1, %xmm0 ## encoding: [0x66,0x0f,0xe0,0xc1]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: mm_avg_epu8:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vpavgb %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0xe0,0xc1]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: mm_avg_epu8:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vpavgb %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0xe0,0xc1]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <16 x i8> @llvm.x86.sse2.pavg.b(<16 x i8> %a0, <16 x i8> %a1) ; <<16 x i8>> [#uses=1]
  ret <16 x i8> %res
}
declare <16 x i8> @llvm.x86.sse2.pavg.b(<16 x i8>, <16 x i8>) nounwind readnone

define <8 x i16> @mm_avg_epu16(<8 x i16> %a0, <8 x i16> %a1) {
; SSE-LABEL: mm_avg_epu16:
; SSE:       ## %bb.0:
; SSE-NEXT:    pavgw %xmm1, %xmm0 ## encoding: [0x66,0x0f,0xe3,0xc1]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: mm_avg_epu16:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vpavgw %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0xe3,0xc1]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: mm_avg_epu16:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vpavgw %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0xe3,0xc1]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <8 x i16> @llvm.x86.sse2.pavg.w(<8 x i16> %a0, <8 x i16> %a1) ; <<8 x i16>> [#uses=1]
  ret <8 x i16> %res
}
declare <8 x i16> @llvm.x86.sse2.pavg.w(<8 x i16>, <8 x i16>) nounwind readnone


define <2 x i64> @test_x86_sse2_pmulu_dq(<4 x i32> %a0, <4 x i32> %a1) {
; SSE-LABEL: test_x86_sse2_pmulu_dq:
; SSE:       ## %bb.0:
; SSE-NEXT:    pmuludq %xmm1, %xmm0 ## encoding: [0x66,0x0f,0xf4,0xc1]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_pmulu_dq:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vpmuludq %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xf9,0xf4,0xc1]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_pmulu_dq:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vpmuludq %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf9,0xf4,0xc1]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x i64> @llvm.x86.sse2.pmulu.dq(<4 x i32> %a0, <4 x i32> %a1) ; <<2 x i64>> [#uses=1]
  ret <2 x i64> %res
}
declare <2 x i64> @llvm.x86.sse2.pmulu.dq(<4 x i32>, <4 x i32>) nounwind readnone


define <2 x double> @test_x86_sse2_cvtsi2sd(<2 x double> %a0, i32 %a1) {
; X86-SSE-LABEL: test_x86_sse2_cvtsi2sd:
; X86-SSE:       ## %bb.0:
; X86-SSE-NEXT:    cvtsi2sdl {{[0-9]+}}(%esp), %xmm0 ## encoding: [0xf2,0x0f,0x2a,0x44,0x24,0x04]
; X86-SSE-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX1-LABEL: test_x86_sse2_cvtsi2sd:
; X86-AVX1:       ## %bb.0:
; X86-AVX1-NEXT:    vcvtsi2sdl {{[0-9]+}}(%esp), %xmm0, %xmm0 ## encoding: [0xc5,0xfb,0x2a,0x44,0x24,0x04]
; X86-AVX1-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX512-LABEL: test_x86_sse2_cvtsi2sd:
; X86-AVX512:       ## %bb.0:
; X86-AVX512-NEXT:    vcvtsi2sdl {{[0-9]+}}(%esp), %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x2a,0x44,0x24,0x04]
; X86-AVX512-NEXT:    retl ## encoding: [0xc3]
;
; X64-SSE-LABEL: test_x86_sse2_cvtsi2sd:
; X64-SSE:       ## %bb.0:
; X64-SSE-NEXT:    cvtsi2sdl %edi, %xmm0 ## encoding: [0xf2,0x0f,0x2a,0xc7]
; X64-SSE-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX1-LABEL: test_x86_sse2_cvtsi2sd:
; X64-AVX1:       ## %bb.0:
; X64-AVX1-NEXT:    vcvtsi2sdl %edi, %xmm0, %xmm0 ## encoding: [0xc5,0xfb,0x2a,0xc7]
; X64-AVX1-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX512-LABEL: test_x86_sse2_cvtsi2sd:
; X64-AVX512:       ## %bb.0:
; X64-AVX512-NEXT:    vcvtsi2sdl %edi, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x2a,0xc7]
; X64-AVX512-NEXT:    retq ## encoding: [0xc3]
  %res = call <2 x double> @llvm.x86.sse2.cvtsi2sd(<2 x double> %a0, i32 %a1) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}
declare <2 x double> @llvm.x86.sse2.cvtsi2sd(<2 x double>, i32) nounwind readnone


define <2 x double> @test_x86_sse2_cvtss2sd(<2 x double> %a0, <4 x float> %a1) {
; SSE-LABEL: test_x86_sse2_cvtss2sd:
; SSE:       ## %bb.0:
; SSE-NEXT:    cvtss2sd %xmm1, %xmm0 ## encoding: [0xf3,0x0f,0x5a,0xc1]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_cvtss2sd:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vcvtss2sd %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xfa,0x5a,0xc1]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_cvtss2sd:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vcvtss2sd %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfa,0x5a,0xc1]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <2 x double> @llvm.x86.sse2.cvtss2sd(<2 x double> %a0, <4 x float> %a1) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}
declare <2 x double> @llvm.x86.sse2.cvtss2sd(<2 x double>, <4 x float>) nounwind readnone


define <2 x double> @test_x86_sse2_cvtss2sd_load(<2 x double> %a0, <4 x float>* %p1) {
; X86-SSE-LABEL: test_x86_sse2_cvtss2sd_load:
; X86-SSE:       ## %bb.0:
; X86-SSE-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-SSE-NEXT:    movss (%eax), %xmm1 ## encoding: [0xf3,0x0f,0x10,0x08]
; X86-SSE-NEXT:    ## xmm1 = mem[0],zero,zero,zero
; X86-SSE-NEXT:    cvtss2sd %xmm1, %xmm1 ## encoding: [0xf3,0x0f,0x5a,0xc9]
; X86-SSE-NEXT:    movsd %xmm1, %xmm0 ## encoding: [0xf2,0x0f,0x10,0xc1]
; X86-SSE-NEXT:    ## xmm0 = xmm1[0],xmm0[1]
; X86-SSE-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX1-LABEL: test_x86_sse2_cvtss2sd_load:
; X86-AVX1:       ## %bb.0:
; X86-AVX1-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-AVX1-NEXT:    vmovss (%eax), %xmm1 ## encoding: [0xc5,0xfa,0x10,0x08]
; X86-AVX1-NEXT:    ## xmm1 = mem[0],zero,zero,zero
; X86-AVX1-NEXT:    vcvtss2sd %xmm1, %xmm1, %xmm1 ## encoding: [0xc5,0xf2,0x5a,0xc9]
; X86-AVX1-NEXT:    vblendps $3, %xmm1, %xmm0, %xmm0 ## encoding: [0xc4,0xe3,0x79,0x0c,0xc1,0x03]
; X86-AVX1-NEXT:    ## xmm0 = xmm1[0,1],xmm0[2,3]
; X86-AVX1-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX512-LABEL: test_x86_sse2_cvtss2sd_load:
; X86-AVX512:       ## %bb.0:
; X86-AVX512-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-AVX512-NEXT:    vmovss (%eax), %xmm1 ## EVEX TO VEX Compression encoding: [0xc5,0xfa,0x10,0x08]
; X86-AVX512-NEXT:    ## xmm1 = mem[0],zero,zero,zero
; X86-AVX512-NEXT:    vcvtss2sd %xmm1, %xmm1, %xmm1 ## EVEX TO VEX Compression encoding: [0xc5,0xf2,0x5a,0xc9]
; X86-AVX512-NEXT:    vblendps $3, %xmm1, %xmm0, %xmm0 ## encoding: [0xc4,0xe3,0x79,0x0c,0xc1,0x03]
; X86-AVX512-NEXT:    ## xmm0 = xmm1[0,1],xmm0[2,3]
; X86-AVX512-NEXT:    retl ## encoding: [0xc3]
;
; X64-SSE-LABEL: test_x86_sse2_cvtss2sd_load:
; X64-SSE:       ## %bb.0:
; X64-SSE-NEXT:    movss (%rdi), %xmm1 ## encoding: [0xf3,0x0f,0x10,0x0f]
; X64-SSE-NEXT:    ## xmm1 = mem[0],zero,zero,zero
; X64-SSE-NEXT:    cvtss2sd %xmm1, %xmm1 ## encoding: [0xf3,0x0f,0x5a,0xc9]
; X64-SSE-NEXT:    movsd %xmm1, %xmm0 ## encoding: [0xf2,0x0f,0x10,0xc1]
; X64-SSE-NEXT:    ## xmm0 = xmm1[0],xmm0[1]
; X64-SSE-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX1-LABEL: test_x86_sse2_cvtss2sd_load:
; X64-AVX1:       ## %bb.0:
; X64-AVX1-NEXT:    vmovss (%rdi), %xmm1 ## encoding: [0xc5,0xfa,0x10,0x0f]
; X64-AVX1-NEXT:    ## xmm1 = mem[0],zero,zero,zero
; X64-AVX1-NEXT:    vcvtss2sd %xmm1, %xmm1, %xmm1 ## encoding: [0xc5,0xf2,0x5a,0xc9]
; X64-AVX1-NEXT:    vblendps $3, %xmm1, %xmm0, %xmm0 ## encoding: [0xc4,0xe3,0x79,0x0c,0xc1,0x03]
; X64-AVX1-NEXT:    ## xmm0 = xmm1[0,1],xmm0[2,3]
; X64-AVX1-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX512-LABEL: test_x86_sse2_cvtss2sd_load:
; X64-AVX512:       ## %bb.0:
; X64-AVX512-NEXT:    vmovss (%rdi), %xmm1 ## EVEX TO VEX Compression encoding: [0xc5,0xfa,0x10,0x0f]
; X64-AVX512-NEXT:    ## xmm1 = mem[0],zero,zero,zero
; X64-AVX512-NEXT:    vcvtss2sd %xmm1, %xmm1, %xmm1 ## EVEX TO VEX Compression encoding: [0xc5,0xf2,0x5a,0xc9]
; X64-AVX512-NEXT:    vblendps $3, %xmm1, %xmm0, %xmm0 ## encoding: [0xc4,0xe3,0x79,0x0c,0xc1,0x03]
; X64-AVX512-NEXT:    ## xmm0 = xmm1[0,1],xmm0[2,3]
; X64-AVX512-NEXT:    retq ## encoding: [0xc3]
  %a1 = load <4 x float>, <4 x float>* %p1
  %res = call <2 x double> @llvm.x86.sse2.cvtss2sd(<2 x double> %a0, <4 x float> %a1) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}


define <2 x double> @test_x86_sse2_cvtss2sd_load_optsize(<2 x double> %a0, <4 x float>* %p1) optsize {
; X86-SSE-LABEL: test_x86_sse2_cvtss2sd_load_optsize:
; X86-SSE:       ## %bb.0:
; X86-SSE-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-SSE-NEXT:    cvtss2sd (%eax), %xmm1 ## encoding: [0xf3,0x0f,0x5a,0x08]
; X86-SSE-NEXT:    movsd %xmm1, %xmm0 ## encoding: [0xf2,0x0f,0x10,0xc1]
; X86-SSE-NEXT:    ## xmm0 = xmm1[0],xmm0[1]
; X86-SSE-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX1-LABEL: test_x86_sse2_cvtss2sd_load_optsize:
; X86-AVX1:       ## %bb.0:
; X86-AVX1-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-AVX1-NEXT:    vcvtss2sd (%eax), %xmm1, %xmm1 ## encoding: [0xc5,0xf2,0x5a,0x08]
; X86-AVX1-NEXT:    vmovsd %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xfb,0x10,0xc1]
; X86-AVX1-NEXT:    ## xmm0 = xmm1[0],xmm0[1]
; X86-AVX1-NEXT:    retl ## encoding: [0xc3]
;
; X86-AVX512-LABEL: test_x86_sse2_cvtss2sd_load_optsize:
; X86-AVX512:       ## %bb.0:
; X86-AVX512-NEXT:    movl {{[0-9]+}}(%esp), %eax ## encoding: [0x8b,0x44,0x24,0x04]
; X86-AVX512-NEXT:    vcvtss2sd (%eax), %xmm1, %xmm1 ## EVEX TO VEX Compression encoding: [0xc5,0xf2,0x5a,0x08]
; X86-AVX512-NEXT:    vmovsd %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x10,0xc1]
; X86-AVX512-NEXT:    ## xmm0 = xmm1[0],xmm0[1]
; X86-AVX512-NEXT:    retl ## encoding: [0xc3]
;
; X64-SSE-LABEL: test_x86_sse2_cvtss2sd_load_optsize:
; X64-SSE:       ## %bb.0:
; X64-SSE-NEXT:    cvtss2sd (%rdi), %xmm1 ## encoding: [0xf3,0x0f,0x5a,0x0f]
; X64-SSE-NEXT:    movsd %xmm1, %xmm0 ## encoding: [0xf2,0x0f,0x10,0xc1]
; X64-SSE-NEXT:    ## xmm0 = xmm1[0],xmm0[1]
; X64-SSE-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX1-LABEL: test_x86_sse2_cvtss2sd_load_optsize:
; X64-AVX1:       ## %bb.0:
; X64-AVX1-NEXT:    vcvtss2sd (%rdi), %xmm1, %xmm1 ## encoding: [0xc5,0xf2,0x5a,0x0f]
; X64-AVX1-NEXT:    vmovsd %xmm1, %xmm0, %xmm0 ## encoding: [0xc5,0xfb,0x10,0xc1]
; X64-AVX1-NEXT:    ## xmm0 = xmm1[0],xmm0[1]
; X64-AVX1-NEXT:    retq ## encoding: [0xc3]
;
; X64-AVX512-LABEL: test_x86_sse2_cvtss2sd_load_optsize:
; X64-AVX512:       ## %bb.0:
; X64-AVX512-NEXT:    vcvtss2sd (%rdi), %xmm1, %xmm1 ## EVEX TO VEX Compression encoding: [0xc5,0xf2,0x5a,0x0f]
; X64-AVX512-NEXT:    vmovsd %xmm1, %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xfb,0x10,0xc1]
; X64-AVX512-NEXT:    ## xmm0 = xmm1[0],xmm0[1]
; X64-AVX512-NEXT:    retq ## encoding: [0xc3]
  %a1 = load <4 x float>, <4 x float>* %p1
  %res = call <2 x double> @llvm.x86.sse2.cvtss2sd(<2 x double> %a0, <4 x float> %a1) ; <<2 x double>> [#uses=1]
  ret <2 x double> %res
}


define <4 x float> @test_x86_sse2_cvtdq2ps(<4 x i32> %a0) {
; SSE-LABEL: test_x86_sse2_cvtdq2ps:
; SSE:       ## %bb.0:
; SSE-NEXT:    cvtdq2ps %xmm0, %xmm0 ## encoding: [0x0f,0x5b,0xc0]
; SSE-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX1-LABEL: test_x86_sse2_cvtdq2ps:
; AVX1:       ## %bb.0:
; AVX1-NEXT:    vcvtdq2ps %xmm0, %xmm0 ## encoding: [0xc5,0xf8,0x5b,0xc0]
; AVX1-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
;
; AVX512-LABEL: test_x86_sse2_cvtdq2ps:
; AVX512:       ## %bb.0:
; AVX512-NEXT:    vcvtdq2ps %xmm0, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf8,0x5b,0xc0]
; AVX512-NEXT:    ret{{[l|q]}} ## encoding: [0xc3]
  %res = call <4 x float> @llvm.x86.sse2.cvtdq2ps(<4 x i32> %a0) ; <<4 x float>> [#uses=1]
  ret <4 x float> %res
}
declare <4 x float> @llvm.x86.sse2.cvtdq2ps(<4 x i32>) nounwind readnone