[llvm] r323531 - [X86][SSE] Add tests for vector truncation with PACKUS style signed saturation
Simon Pilgrim via llvm-commits
llvm-commits at lists.llvm.org
Fri Jan 26 06:58:51 PST 2018
Author: rksimon
Date: Fri Jan 26 06:58:50 2018
New Revision: 323531
URL: http://llvm.org/viewvc/llvm-project?rev=323531&view=rev
Log:
[X86][SSE] Add tests for vector truncation with PACKUS style signed saturation
PACKUS - truncates signed value, saturating to [0,unsigned_max_trunc]
Added:
llvm/trunk/test/CodeGen/X86/vector-trunc-packus.ll
Added: llvm/trunk/test/CodeGen/X86/vector-trunc-packus.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/vector-trunc-packus.ll?rev=323531&view=auto
==============================================================================
--- llvm/trunk/test/CodeGen/X86/vector-trunc-packus.ll (added)
+++ llvm/trunk/test/CodeGen/X86/vector-trunc-packus.ll Fri Jan 26 06:58:50 2018
@@ -0,0 +1,3315 @@
+; 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=+ssse3 | FileCheck %s --check-prefix=SSE --check-prefix=SSSE3
+; 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-prefixes=AVX,AVX2,AVX2-SLOW
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2,+fast-variable-shuffle | FileCheck %s --check-prefixes=AVX,AVX2,AVX2-FAST
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512f | FileCheck %s --check-prefix=AVX512 --check-prefix=AVX512F
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512vl,+fast-variable-shuffle | FileCheck %s --check-prefixes=AVX512,AVX512VL
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512bw,+fast-variable-shuffle | FileCheck %s --check-prefixes=AVX512,AVX512BW
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512bw,+avx512vl,+fast-variable-shuffle | FileCheck %s --check-prefixes=AVX512,AVX512BWVL
+
+;
+; PACKUS saturation truncation to vXi32
+;
+
+define <4 x i32> @trunc_packus_v4i64_v4i32(<4 x i64> %a0) {
+; SSE2-LABEL: trunc_packus_v4i64_v4i32:
+; SSE2: # %bb.0:
+; SSE2-NEXT: movdqa {{.*#+}} xmm2 = [2147483648,0,2147483648,0]
+; SSE2-NEXT: movdqa %xmm0, %xmm3
+; SSE2-NEXT: pxor %xmm2, %xmm3
+; SSE2-NEXT: movdqa {{.*#+}} xmm4 = [2147483647,2147483647]
+; SSE2-NEXT: movdqa %xmm4, %xmm5
+; SSE2-NEXT: pcmpgtd %xmm3, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm5[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm4, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm3[1,1,3,3]
+; SSE2-NEXT: pand %xmm6, %xmm7
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm5[1,1,3,3]
+; SSE2-NEXT: por %xmm7, %xmm3
+; SSE2-NEXT: movdqa %xmm1, %xmm5
+; SSE2-NEXT: pxor %xmm2, %xmm5
+; SSE2-NEXT: movdqa %xmm4, %xmm6
+; SSE2-NEXT: pcmpgtd %xmm5, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm4, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE2-NEXT: pand %xmm7, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm6[1,1,3,3]
+; SSE2-NEXT: por %xmm5, %xmm4
+; SSE2-NEXT: movdqa {{.*#+}} xmm5 = [4294967295,4294967295]
+; SSE2-NEXT: pand %xmm4, %xmm1
+; SSE2-NEXT: pandn %xmm5, %xmm4
+; SSE2-NEXT: por %xmm1, %xmm4
+; SSE2-NEXT: pand %xmm3, %xmm0
+; SSE2-NEXT: pandn %xmm5, %xmm3
+; SSE2-NEXT: por %xmm0, %xmm3
+; SSE2-NEXT: movdqa %xmm3, %xmm0
+; SSE2-NEXT: pxor %xmm2, %xmm0
+; SSE2-NEXT: movdqa %xmm0, %xmm1
+; SSE2-NEXT: pcmpgtd %xmm2, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm1[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm2, %xmm0
+; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm0[1,1,3,3]
+; SSE2-NEXT: pand %xmm5, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
+; SSE2-NEXT: por %xmm6, %xmm0
+; SSE2-NEXT: movdqa %xmm4, %xmm1
+; SSE2-NEXT: pxor %xmm2, %xmm1
+; SSE2-NEXT: movdqa %xmm1, %xmm5
+; SSE2-NEXT: pcmpgtd %xmm2, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm5[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm2, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSE2-NEXT: pand %xmm6, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm5[1,1,3,3]
+; SSE2-NEXT: por %xmm1, %xmm2
+; SSE2-NEXT: pand %xmm4, %xmm2
+; SSE2-NEXT: pand %xmm3, %xmm0
+; SSE2-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2],xmm2[0,2]
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: trunc_packus_v4i64_v4i32:
+; SSSE3: # %bb.0:
+; SSSE3-NEXT: movdqa {{.*#+}} xmm2 = [2147483648,0,2147483648,0]
+; SSSE3-NEXT: movdqa %xmm0, %xmm3
+; SSSE3-NEXT: pxor %xmm2, %xmm3
+; SSSE3-NEXT: movdqa {{.*#+}} xmm4 = [2147483647,2147483647]
+; SSSE3-NEXT: movdqa %xmm4, %xmm5
+; SSSE3-NEXT: pcmpgtd %xmm3, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm6 = xmm5[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm4, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm3[1,1,3,3]
+; SSSE3-NEXT: pand %xmm6, %xmm7
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm5[1,1,3,3]
+; SSSE3-NEXT: por %xmm7, %xmm3
+; SSSE3-NEXT: movdqa %xmm1, %xmm5
+; SSSE3-NEXT: pxor %xmm2, %xmm5
+; SSSE3-NEXT: movdqa %xmm4, %xmm6
+; SSSE3-NEXT: pcmpgtd %xmm5, %xmm6
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm4, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSSE3-NEXT: pand %xmm7, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm6[1,1,3,3]
+; SSSE3-NEXT: por %xmm5, %xmm4
+; SSSE3-NEXT: movdqa {{.*#+}} xmm5 = [4294967295,4294967295]
+; SSSE3-NEXT: pand %xmm4, %xmm1
+; SSSE3-NEXT: pandn %xmm5, %xmm4
+; SSSE3-NEXT: por %xmm1, %xmm4
+; SSSE3-NEXT: pand %xmm3, %xmm0
+; SSSE3-NEXT: pandn %xmm5, %xmm3
+; SSSE3-NEXT: por %xmm0, %xmm3
+; SSSE3-NEXT: movdqa %xmm3, %xmm0
+; SSSE3-NEXT: pxor %xmm2, %xmm0
+; SSSE3-NEXT: movdqa %xmm0, %xmm1
+; SSSE3-NEXT: pcmpgtd %xmm2, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm1[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm2, %xmm0
+; SSSE3-NEXT: pshufd {{.*#+}} xmm6 = xmm0[1,1,3,3]
+; SSSE3-NEXT: pand %xmm5, %xmm6
+; SSSE3-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
+; SSSE3-NEXT: por %xmm6, %xmm0
+; SSSE3-NEXT: movdqa %xmm4, %xmm1
+; SSSE3-NEXT: pxor %xmm2, %xmm1
+; SSSE3-NEXT: movdqa %xmm1, %xmm5
+; SSSE3-NEXT: pcmpgtd %xmm2, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm6 = xmm5[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm2, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSSE3-NEXT: pand %xmm6, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm5[1,1,3,3]
+; SSSE3-NEXT: por %xmm1, %xmm2
+; SSSE3-NEXT: pand %xmm4, %xmm2
+; SSSE3-NEXT: pand %xmm3, %xmm0
+; SSSE3-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2],xmm2[0,2]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: trunc_packus_v4i64_v4i32:
+; SSE41: # %bb.0:
+; SSE41-NEXT: movdqa %xmm0, %xmm2
+; SSE41-NEXT: movdqa {{.*#+}} xmm4 = [2147483648,0,2147483648,0]
+; SSE41-NEXT: pxor %xmm4, %xmm0
+; SSE41-NEXT: movdqa {{.*#+}} xmm5 = [2147483647,2147483647]
+; SSE41-NEXT: movdqa %xmm5, %xmm3
+; SSE41-NEXT: pcmpgtd %xmm0, %xmm3
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm3[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm5, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm6, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm3
+; SSE41-NEXT: movdqa %xmm1, %xmm0
+; SSE41-NEXT: pxor %xmm4, %xmm0
+; SSE41-NEXT: movdqa %xmm5, %xmm6
+; SSE41-NEXT: pcmpgtd %xmm0, %xmm6
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm5, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm5 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm6[1,1,3,3]
+; SSE41-NEXT: por %xmm5, %xmm0
+; SSE41-NEXT: movapd {{.*#+}} xmm5 = [4294967295,4294967295]
+; SSE41-NEXT: movapd %xmm5, %xmm6
+; SSE41-NEXT: blendvpd %xmm0, %xmm1, %xmm6
+; SSE41-NEXT: movdqa %xmm3, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm2, %xmm5
+; SSE41-NEXT: movapd %xmm5, %xmm0
+; SSE41-NEXT: xorpd %xmm4, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm1
+; SSE41-NEXT: pcmpgtd %xmm4, %xmm1
+; SSE41-NEXT: pshufd {{.*#+}} xmm2 = xmm1[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm4, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm3 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm2, %xmm3
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
+; SSE41-NEXT: por %xmm3, %xmm0
+; SSE41-NEXT: movapd %xmm6, %xmm1
+; SSE41-NEXT: xorpd %xmm4, %xmm1
+; SSE41-NEXT: movapd %xmm1, %xmm2
+; SSE41-NEXT: pcmpgtd %xmm4, %xmm2
+; SSE41-NEXT: pshufd {{.*#+}} xmm3 = xmm2[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm4, %xmm1
+; SSE41-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSE41-NEXT: pand %xmm3, %xmm1
+; SSE41-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE41-NEXT: por %xmm1, %xmm2
+; SSE41-NEXT: pand %xmm6, %xmm2
+; SSE41-NEXT: pand %xmm5, %xmm0
+; SSE41-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2],xmm2[0,2]
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_packus_v4i64_v4i32:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm2 = [4294967295,4294967295]
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm2, %xmm1
+; AVX1-NEXT: vpcmpgtq %xmm0, %xmm2, %xmm2
+; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm2, %ymm1
+; AVX1-NEXT: vmovapd {{.*#+}} ymm2 = [4294967295,4294967295,4294967295,4294967295]
+; AVX1-NEXT: vblendvpd %ymm1, %ymm0, %ymm2, %ymm0
+; AVX1-NEXT: vxorpd %xmm1, %xmm1, %xmm1
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm0, %xmm2
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm3
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm3, %xmm1
+; AVX1-NEXT: vpand %xmm3, %xmm1, %xmm1
+; AVX1-NEXT: vpand %xmm0, %xmm2, %xmm0
+; AVX1-NEXT: vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm1[0,2]
+; AVX1-NEXT: vzeroupper
+; AVX1-NEXT: retq
+;
+; AVX2-SLOW-LABEL: trunc_packus_v4i64_v4i32:
+; AVX2-SLOW: # %bb.0:
+; AVX2-SLOW-NEXT: vpbroadcastq {{.*#+}} ymm1 = [4294967295,4294967295,4294967295,4294967295]
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm0, %ymm1, %ymm2
+; AVX2-SLOW-NEXT: vblendvpd %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-SLOW-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm1, %ymm0, %ymm1
+; AVX2-SLOW-NEXT: vpand %ymm0, %ymm1, %ymm0
+; AVX2-SLOW-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[0,2,2,3,4,6,6,7]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-SLOW-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX2-SLOW-NEXT: vzeroupper
+; AVX2-SLOW-NEXT: retq
+;
+; AVX2-FAST-LABEL: trunc_packus_v4i64_v4i32:
+; AVX2-FAST: # %bb.0:
+; AVX2-FAST-NEXT: vpbroadcastq {{.*#+}} ymm1 = [4294967295,4294967295,4294967295,4294967295]
+; AVX2-FAST-NEXT: vpcmpgtq %ymm0, %ymm1, %ymm2
+; AVX2-FAST-NEXT: vblendvpd %ymm2, %ymm0, %ymm1, %ymm0
+; AVX2-FAST-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX2-FAST-NEXT: vpcmpgtq %ymm1, %ymm0, %ymm1
+; AVX2-FAST-NEXT: vpand %ymm0, %ymm1, %ymm0
+; AVX2-FAST-NEXT: vmovdqa {{.*#+}} ymm1 = [0,2,4,6,4,6,6,7]
+; AVX2-FAST-NEXT: vpermd %ymm0, %ymm1, %ymm0
+; AVX2-FAST-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX2-FAST-NEXT: vzeroupper
+; AVX2-FAST-NEXT: retq
+;
+; AVX512F-LABEL: trunc_packus_v4i64_v4i32:
+; AVX512F: # %bb.0:
+; AVX512F-NEXT: # kill: def %ymm0 killed %ymm0 def %zmm0
+; AVX512F-NEXT: vpbroadcastq {{.*#+}} ymm1 = [4294967295,4294967295,4294967295,4294967295]
+; AVX512F-NEXT: vpminsq %zmm1, %zmm0, %zmm0
+; AVX512F-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512F-NEXT: vpmaxsq %zmm1, %zmm0, %zmm0
+; AVX512F-NEXT: vpmovqd %zmm0, %ymm0
+; AVX512F-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX512F-NEXT: vzeroupper
+; AVX512F-NEXT: retq
+;
+; AVX512VL-LABEL: trunc_packus_v4i64_v4i32:
+; AVX512VL: # %bb.0:
+; AVX512VL-NEXT: vpminsq {{.*}}(%rip){1to4}, %ymm0, %ymm0
+; AVX512VL-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512VL-NEXT: vpmaxsq %ymm1, %ymm0, %ymm0
+; AVX512VL-NEXT: vpmovqd %ymm0, %xmm0
+; AVX512VL-NEXT: vzeroupper
+; AVX512VL-NEXT: retq
+;
+; AVX512BW-LABEL: trunc_packus_v4i64_v4i32:
+; AVX512BW: # %bb.0:
+; AVX512BW-NEXT: # kill: def %ymm0 killed %ymm0 def %zmm0
+; AVX512BW-NEXT: vpbroadcastq {{.*#+}} ymm1 = [4294967295,4294967295,4294967295,4294967295]
+; AVX512BW-NEXT: vpminsq %zmm1, %zmm0, %zmm0
+; AVX512BW-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512BW-NEXT: vpmaxsq %zmm1, %zmm0, %zmm0
+; AVX512BW-NEXT: vpmovqd %zmm0, %ymm0
+; AVX512BW-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX512BW-NEXT: vzeroupper
+; AVX512BW-NEXT: retq
+;
+; AVX512BWVL-LABEL: trunc_packus_v4i64_v4i32:
+; AVX512BWVL: # %bb.0:
+; AVX512BWVL-NEXT: vpminsq {{.*}}(%rip){1to4}, %ymm0, %ymm0
+; AVX512BWVL-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512BWVL-NEXT: vpmaxsq %ymm1, %ymm0, %ymm0
+; AVX512BWVL-NEXT: vpmovqd %ymm0, %xmm0
+; AVX512BWVL-NEXT: vzeroupper
+; AVX512BWVL-NEXT: retq
+ %1 = icmp slt <4 x i64> %a0, <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295>
+ %2 = select <4 x i1> %1, <4 x i64> %a0, <4 x i64> <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295>
+ %3 = icmp sgt <4 x i64> %2, zeroinitializer
+ %4 = select <4 x i1> %3, <4 x i64> %2, <4 x i64> zeroinitializer
+ %5 = trunc <4 x i64> %4 to <4 x i32>
+ ret <4 x i32> %5
+}
+
+
+define <8 x i32> @trunc_packus_v8i64_v8i32(<8 x i64> %a0) {
+; SSE2-LABEL: trunc_packus_v8i64_v8i32:
+; SSE2: # %bb.0:
+; SSE2-NEXT: movdqa {{.*#+}} xmm8 = [2147483648,0,2147483648,0]
+; SSE2-NEXT: movdqa %xmm0, %xmm4
+; SSE2-NEXT: pxor %xmm8, %xmm4
+; SSE2-NEXT: movdqa {{.*#+}} xmm9 = [2147483647,2147483647]
+; SSE2-NEXT: movdqa %xmm9, %xmm6
+; SSE2-NEXT: pcmpgtd %xmm4, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm9, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm4[1,1,3,3]
+; SSE2-NEXT: pand %xmm7, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm10 = xmm6[1,1,3,3]
+; SSE2-NEXT: por %xmm5, %xmm10
+; SSE2-NEXT: movdqa %xmm1, %xmm5
+; SSE2-NEXT: pxor %xmm8, %xmm5
+; SSE2-NEXT: movdqa %xmm9, %xmm6
+; SSE2-NEXT: pcmpgtd %xmm5, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm9, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE2-NEXT: pand %xmm7, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm6[1,1,3,3]
+; SSE2-NEXT: por %xmm5, %xmm6
+; SSE2-NEXT: movdqa %xmm2, %xmm5
+; SSE2-NEXT: pxor %xmm8, %xmm5
+; SSE2-NEXT: movdqa %xmm9, %xmm7
+; SSE2-NEXT: pcmpgtd %xmm5, %xmm7
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm7[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm9, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE2-NEXT: pand %xmm4, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm7[1,1,3,3]
+; SSE2-NEXT: por %xmm5, %xmm7
+; SSE2-NEXT: movdqa %xmm3, %xmm4
+; SSE2-NEXT: pxor %xmm8, %xmm4
+; SSE2-NEXT: movdqa %xmm9, %xmm5
+; SSE2-NEXT: pcmpgtd %xmm4, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm11 = xmm5[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm9, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSE2-NEXT: pand %xmm11, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE2-NEXT: por %xmm4, %xmm5
+; SSE2-NEXT: movdqa {{.*#+}} xmm4 = [4294967295,4294967295]
+; SSE2-NEXT: pand %xmm5, %xmm3
+; SSE2-NEXT: pandn %xmm4, %xmm5
+; SSE2-NEXT: por %xmm3, %xmm5
+; SSE2-NEXT: pand %xmm7, %xmm2
+; SSE2-NEXT: pandn %xmm4, %xmm7
+; SSE2-NEXT: por %xmm2, %xmm7
+; SSE2-NEXT: pand %xmm6, %xmm1
+; SSE2-NEXT: pandn %xmm4, %xmm6
+; SSE2-NEXT: por %xmm1, %xmm6
+; SSE2-NEXT: pand %xmm10, %xmm0
+; SSE2-NEXT: pandn %xmm4, %xmm10
+; SSE2-NEXT: por %xmm0, %xmm10
+; SSE2-NEXT: movdqa %xmm10, %xmm0
+; SSE2-NEXT: pxor %xmm8, %xmm0
+; SSE2-NEXT: movdqa %xmm0, %xmm1
+; SSE2-NEXT: pcmpgtd %xmm8, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm1[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm8, %xmm0
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm0[1,1,3,3]
+; SSE2-NEXT: pand %xmm2, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
+; SSE2-NEXT: por %xmm3, %xmm0
+; SSE2-NEXT: movdqa %xmm6, %xmm1
+; SSE2-NEXT: pxor %xmm8, %xmm1
+; SSE2-NEXT: movdqa %xmm1, %xmm2
+; SSE2-NEXT: pcmpgtd %xmm8, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm2[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm8, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSE2-NEXT: pand %xmm3, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: por %xmm1, %xmm2
+; SSE2-NEXT: movdqa %xmm7, %xmm1
+; SSE2-NEXT: pxor %xmm8, %xmm1
+; SSE2-NEXT: movdqa %xmm1, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm8, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm9 = xmm3[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm8, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm1[1,1,3,3]
+; SSE2-NEXT: pand %xmm9, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm3[1,1,3,3]
+; SSE2-NEXT: por %xmm4, %xmm1
+; SSE2-NEXT: movdqa %xmm5, %xmm3
+; SSE2-NEXT: pxor %xmm8, %xmm3
+; SSE2-NEXT: movdqa %xmm3, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm8, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm9 = xmm4[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm8, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSE2-NEXT: pand %xmm9, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm3, %xmm4
+; SSE2-NEXT: pand %xmm5, %xmm4
+; SSE2-NEXT: pand %xmm7, %xmm1
+; SSE2-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,2],xmm4[0,2]
+; SSE2-NEXT: pand %xmm6, %xmm2
+; SSE2-NEXT: pand %xmm10, %xmm0
+; SSE2-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2],xmm2[0,2]
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: trunc_packus_v8i64_v8i32:
+; SSSE3: # %bb.0:
+; SSSE3-NEXT: movdqa {{.*#+}} xmm8 = [2147483648,0,2147483648,0]
+; SSSE3-NEXT: movdqa %xmm0, %xmm4
+; SSSE3-NEXT: pxor %xmm8, %xmm4
+; SSSE3-NEXT: movdqa {{.*#+}} xmm9 = [2147483647,2147483647]
+; SSSE3-NEXT: movdqa %xmm9, %xmm6
+; SSSE3-NEXT: pcmpgtd %xmm4, %xmm6
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm9, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm4[1,1,3,3]
+; SSSE3-NEXT: pand %xmm7, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm10 = xmm6[1,1,3,3]
+; SSSE3-NEXT: por %xmm5, %xmm10
+; SSSE3-NEXT: movdqa %xmm1, %xmm5
+; SSSE3-NEXT: pxor %xmm8, %xmm5
+; SSSE3-NEXT: movdqa %xmm9, %xmm6
+; SSSE3-NEXT: pcmpgtd %xmm5, %xmm6
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm9, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSSE3-NEXT: pand %xmm7, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm6 = xmm6[1,1,3,3]
+; SSSE3-NEXT: por %xmm5, %xmm6
+; SSSE3-NEXT: movdqa %xmm2, %xmm5
+; SSSE3-NEXT: pxor %xmm8, %xmm5
+; SSSE3-NEXT: movdqa %xmm9, %xmm7
+; SSSE3-NEXT: pcmpgtd %xmm5, %xmm7
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm7[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm9, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSSE3-NEXT: pand %xmm4, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm7[1,1,3,3]
+; SSSE3-NEXT: por %xmm5, %xmm7
+; SSSE3-NEXT: movdqa %xmm3, %xmm4
+; SSSE3-NEXT: pxor %xmm8, %xmm4
+; SSSE3-NEXT: movdqa %xmm9, %xmm5
+; SSSE3-NEXT: pcmpgtd %xmm4, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm11 = xmm5[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm9, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSSE3-NEXT: pand %xmm11, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSSE3-NEXT: por %xmm4, %xmm5
+; SSSE3-NEXT: movdqa {{.*#+}} xmm4 = [4294967295,4294967295]
+; SSSE3-NEXT: pand %xmm5, %xmm3
+; SSSE3-NEXT: pandn %xmm4, %xmm5
+; SSSE3-NEXT: por %xmm3, %xmm5
+; SSSE3-NEXT: pand %xmm7, %xmm2
+; SSSE3-NEXT: pandn %xmm4, %xmm7
+; SSSE3-NEXT: por %xmm2, %xmm7
+; SSSE3-NEXT: pand %xmm6, %xmm1
+; SSSE3-NEXT: pandn %xmm4, %xmm6
+; SSSE3-NEXT: por %xmm1, %xmm6
+; SSSE3-NEXT: pand %xmm10, %xmm0
+; SSSE3-NEXT: pandn %xmm4, %xmm10
+; SSSE3-NEXT: por %xmm0, %xmm10
+; SSSE3-NEXT: movdqa %xmm10, %xmm0
+; SSSE3-NEXT: pxor %xmm8, %xmm0
+; SSSE3-NEXT: movdqa %xmm0, %xmm1
+; SSSE3-NEXT: pcmpgtd %xmm8, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm1[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm8, %xmm0
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm0[1,1,3,3]
+; SSSE3-NEXT: pand %xmm2, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
+; SSSE3-NEXT: por %xmm3, %xmm0
+; SSSE3-NEXT: movdqa %xmm6, %xmm1
+; SSSE3-NEXT: pxor %xmm8, %xmm1
+; SSSE3-NEXT: movdqa %xmm1, %xmm2
+; SSSE3-NEXT: pcmpgtd %xmm8, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm2[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm8, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSSE3-NEXT: pand %xmm3, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: por %xmm1, %xmm2
+; SSSE3-NEXT: movdqa %xmm7, %xmm1
+; SSSE3-NEXT: pxor %xmm8, %xmm1
+; SSSE3-NEXT: movdqa %xmm1, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm8, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm9 = xmm3[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm8, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm1[1,1,3,3]
+; SSSE3-NEXT: pand %xmm9, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm3[1,1,3,3]
+; SSSE3-NEXT: por %xmm4, %xmm1
+; SSSE3-NEXT: movdqa %xmm5, %xmm3
+; SSSE3-NEXT: pxor %xmm8, %xmm3
+; SSSE3-NEXT: movdqa %xmm3, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm8, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm9 = xmm4[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm8, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSSE3-NEXT: pand %xmm9, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSSE3-NEXT: por %xmm3, %xmm4
+; SSSE3-NEXT: pand %xmm5, %xmm4
+; SSSE3-NEXT: pand %xmm7, %xmm1
+; SSSE3-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,2],xmm4[0,2]
+; SSSE3-NEXT: pand %xmm6, %xmm2
+; SSSE3-NEXT: pand %xmm10, %xmm0
+; SSSE3-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2],xmm2[0,2]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: trunc_packus_v8i64_v8i32:
+; SSE41: # %bb.0:
+; SSE41-NEXT: movdqa %xmm0, %xmm8
+; SSE41-NEXT: movdqa {{.*#+}} xmm11 = [2147483648,0,2147483648,0]
+; SSE41-NEXT: movdqa %xmm8, %xmm5
+; SSE41-NEXT: pxor %xmm11, %xmm5
+; SSE41-NEXT: movdqa {{.*#+}} xmm0 = [2147483647,2147483647]
+; SSE41-NEXT: movdqa %xmm0, %xmm6
+; SSE41-NEXT: pcmpgtd %xmm5, %xmm6
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm0, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm9 = xmm6[1,1,3,3]
+; SSE41-NEXT: por %xmm5, %xmm9
+; SSE41-NEXT: movdqa %xmm1, %xmm5
+; SSE41-NEXT: pxor %xmm11, %xmm5
+; SSE41-NEXT: movdqa %xmm0, %xmm6
+; SSE41-NEXT: pcmpgtd %xmm5, %xmm6
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm0, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm12 = xmm6[1,1,3,3]
+; SSE41-NEXT: por %xmm5, %xmm12
+; SSE41-NEXT: movdqa %xmm2, %xmm5
+; SSE41-NEXT: pxor %xmm11, %xmm5
+; SSE41-NEXT: movdqa %xmm0, %xmm7
+; SSE41-NEXT: pcmpgtd %xmm5, %xmm7
+; SSE41-NEXT: pshufd {{.*#+}} xmm4 = xmm7[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm0, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE41-NEXT: pand %xmm4, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm7[1,1,3,3]
+; SSE41-NEXT: por %xmm5, %xmm7
+; SSE41-NEXT: movdqa %xmm3, %xmm4
+; SSE41-NEXT: pxor %xmm11, %xmm4
+; SSE41-NEXT: movdqa %xmm0, %xmm5
+; SSE41-NEXT: pcmpgtd %xmm4, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm5[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm0, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSE41-NEXT: pand %xmm6, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm5[1,1,3,3]
+; SSE41-NEXT: por %xmm4, %xmm0
+; SSE41-NEXT: movapd {{.*#+}} xmm5 = [4294967295,4294967295]
+; SSE41-NEXT: movapd %xmm5, %xmm10
+; SSE41-NEXT: blendvpd %xmm0, %xmm3, %xmm10
+; SSE41-NEXT: movapd %xmm5, %xmm3
+; SSE41-NEXT: movdqa %xmm7, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm2, %xmm3
+; SSE41-NEXT: movapd %xmm5, %xmm13
+; SSE41-NEXT: movdqa %xmm12, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm1, %xmm13
+; SSE41-NEXT: movdqa %xmm9, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm8, %xmm5
+; SSE41-NEXT: movapd %xmm5, %xmm0
+; SSE41-NEXT: xorpd %xmm11, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm1
+; SSE41-NEXT: pcmpgtd %xmm11, %xmm1
+; SSE41-NEXT: pshufd {{.*#+}} xmm4 = xmm1[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm11, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm4, %xmm6
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
+; SSE41-NEXT: por %xmm6, %xmm0
+; SSE41-NEXT: movapd %xmm13, %xmm1
+; SSE41-NEXT: xorpd %xmm11, %xmm1
+; SSE41-NEXT: movapd %xmm1, %xmm4
+; SSE41-NEXT: pcmpgtd %xmm11, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm4[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm11, %xmm1
+; SSE41-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSE41-NEXT: pand %xmm6, %xmm1
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm4[1,1,3,3]
+; SSE41-NEXT: por %xmm1, %xmm6
+; SSE41-NEXT: movapd %xmm3, %xmm1
+; SSE41-NEXT: xorpd %xmm11, %xmm1
+; SSE41-NEXT: movapd %xmm1, %xmm4
+; SSE41-NEXT: pcmpgtd %xmm11, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm4[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm11, %xmm1
+; SSE41-NEXT: pshufd {{.*#+}} xmm2 = xmm1[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm2
+; SSE41-NEXT: pshufd {{.*#+}} xmm1 = xmm4[1,1,3,3]
+; SSE41-NEXT: por %xmm2, %xmm1
+; SSE41-NEXT: movapd %xmm10, %xmm2
+; SSE41-NEXT: xorpd %xmm11, %xmm2
+; SSE41-NEXT: movapd %xmm2, %xmm4
+; SSE41-NEXT: pcmpgtd %xmm11, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm4[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm11, %xmm2
+; SSE41-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm2
+; SSE41-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSE41-NEXT: por %xmm2, %xmm4
+; SSE41-NEXT: pand %xmm10, %xmm4
+; SSE41-NEXT: pand %xmm3, %xmm1
+; SSE41-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,2],xmm4[0,2]
+; SSE41-NEXT: pand %xmm13, %xmm6
+; SSE41-NEXT: pand %xmm5, %xmm0
+; SSE41-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2],xmm6[0,2]
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_packus_v8i64_v8i32:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm3 = [4294967295,4294967295]
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm3, %xmm2
+; AVX1-NEXT: vpcmpgtq %xmm0, %xmm3, %xmm4
+; AVX1-NEXT: vinsertf128 $1, %xmm2, %ymm4, %ymm2
+; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm4
+; AVX1-NEXT: vpcmpgtq %xmm4, %xmm3, %xmm4
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm3, %xmm3
+; AVX1-NEXT: vinsertf128 $1, %xmm4, %ymm3, %ymm3
+; AVX1-NEXT: vmovapd {{.*#+}} ymm4 = [4294967295,4294967295,4294967295,4294967295]
+; AVX1-NEXT: vblendvpd %ymm3, %ymm1, %ymm4, %ymm1
+; AVX1-NEXT: vblendvpd %ymm2, %ymm0, %ymm4, %ymm0
+; AVX1-NEXT: vxorpd %xmm2, %xmm2, %xmm2
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm0, %xmm3
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm4
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm4, %xmm5
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm1, %xmm6
+; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm7
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm7, %xmm2
+; AVX1-NEXT: vpand %xmm7, %xmm2, %xmm2
+; AVX1-NEXT: vpand %xmm1, %xmm6, %xmm1
+; AVX1-NEXT: vshufps {{.*#+}} xmm1 = xmm1[0,2],xmm2[0,2]
+; AVX1-NEXT: vpand %xmm4, %xmm5, %xmm2
+; AVX1-NEXT: vpand %xmm0, %xmm3, %xmm0
+; AVX1-NEXT: vshufps {{.*#+}} xmm0 = xmm0[0,2],xmm2[0,2]
+; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT: retq
+;
+; AVX2-SLOW-LABEL: trunc_packus_v8i64_v8i32:
+; AVX2-SLOW: # %bb.0:
+; AVX2-SLOW-NEXT: vpbroadcastq {{.*#+}} ymm2 = [4294967295,4294967295,4294967295,4294967295]
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm0, %ymm2, %ymm3
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm1, %ymm2, %ymm4
+; AVX2-SLOW-NEXT: vblendvpd %ymm4, %ymm1, %ymm2, %ymm1
+; AVX2-SLOW-NEXT: vblendvpd %ymm3, %ymm0, %ymm2, %ymm0
+; AVX2-SLOW-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm2, %ymm0, %ymm3
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm2, %ymm1, %ymm2
+; AVX2-SLOW-NEXT: vpand %ymm1, %ymm2, %ymm1
+; AVX2-SLOW-NEXT: vpand %ymm0, %ymm3, %ymm0
+; AVX2-SLOW-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[0,2,2,3,4,6,6,7]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-SLOW-NEXT: vpshufd {{.*#+}} ymm1 = ymm1[0,2,2,3,4,6,6,7]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,2,2,3]
+; AVX2-SLOW-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-SLOW-NEXT: retq
+;
+; AVX2-FAST-LABEL: trunc_packus_v8i64_v8i32:
+; AVX2-FAST: # %bb.0:
+; AVX2-FAST-NEXT: vpbroadcastq {{.*#+}} ymm2 = [4294967295,4294967295,4294967295,4294967295]
+; AVX2-FAST-NEXT: vpcmpgtq %ymm0, %ymm2, %ymm3
+; AVX2-FAST-NEXT: vpcmpgtq %ymm1, %ymm2, %ymm4
+; AVX2-FAST-NEXT: vblendvpd %ymm4, %ymm1, %ymm2, %ymm1
+; AVX2-FAST-NEXT: vblendvpd %ymm3, %ymm0, %ymm2, %ymm0
+; AVX2-FAST-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX2-FAST-NEXT: vpcmpgtq %ymm2, %ymm0, %ymm3
+; AVX2-FAST-NEXT: vpcmpgtq %ymm2, %ymm1, %ymm2
+; AVX2-FAST-NEXT: vpand %ymm1, %ymm2, %ymm1
+; AVX2-FAST-NEXT: vpand %ymm0, %ymm3, %ymm0
+; AVX2-FAST-NEXT: vmovdqa {{.*#+}} ymm2 = [0,2,4,6,4,6,6,7]
+; AVX2-FAST-NEXT: vpermd %ymm0, %ymm2, %ymm0
+; AVX2-FAST-NEXT: vpermd %ymm1, %ymm2, %ymm1
+; AVX2-FAST-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-FAST-NEXT: retq
+;
+; AVX512-LABEL: trunc_packus_v8i64_v8i32:
+; AVX512: # %bb.0:
+; AVX512-NEXT: vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
+; AVX512-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512-NEXT: vpmaxsq %zmm1, %zmm0, %zmm0
+; AVX512-NEXT: vpmovqd %zmm0, %ymm0
+; AVX512-NEXT: retq
+ %1 = icmp slt <8 x i64> %a0, <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295>
+ %2 = select <8 x i1> %1, <8 x i64> %a0, <8 x i64> <i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295, i64 4294967295>
+ %3 = icmp sgt <8 x i64> %2, zeroinitializer
+ %4 = select <8 x i1> %3, <8 x i64> %2, <8 x i64> zeroinitializer
+ %5 = trunc <8 x i64> %4 to <8 x i32>
+ ret <8 x i32> %5
+}
+
+;
+; PACKUS saturation truncation to vXi16
+;
+
+define <8 x i16> @trunc_packus_v8i64_v8i16(<8 x i64> %a0) {
+; SSE2-LABEL: trunc_packus_v8i64_v8i16:
+; SSE2: # %bb.0:
+; SSE2-NEXT: movdqa {{.*#+}} xmm8 = [2147483648,0,2147483648,0]
+; SSE2-NEXT: movdqa %xmm1, %xmm4
+; SSE2-NEXT: pxor %xmm8, %xmm4
+; SSE2-NEXT: movdqa {{.*#+}} xmm9 = [2147549183,2147549183]
+; SSE2-NEXT: movdqa %xmm9, %xmm6
+; SSE2-NEXT: pcmpgtd %xmm4, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm9, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm4[1,1,3,3]
+; SSE2-NEXT: pand %xmm7, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm11 = xmm6[1,1,3,3]
+; SSE2-NEXT: por %xmm5, %xmm11
+; SSE2-NEXT: movdqa %xmm0, %xmm5
+; SSE2-NEXT: pxor %xmm8, %xmm5
+; SSE2-NEXT: movdqa %xmm9, %xmm6
+; SSE2-NEXT: pcmpgtd %xmm5, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm9, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE2-NEXT: pand %xmm7, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm12 = xmm6[1,1,3,3]
+; SSE2-NEXT: por %xmm5, %xmm12
+; SSE2-NEXT: movdqa %xmm3, %xmm5
+; SSE2-NEXT: pxor %xmm8, %xmm5
+; SSE2-NEXT: movdqa %xmm9, %xmm7
+; SSE2-NEXT: pcmpgtd %xmm5, %xmm7
+; SSE2-NEXT: pshufd {{.*#+}} xmm10 = xmm7[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm9, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE2-NEXT: pand %xmm10, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm7[1,1,3,3]
+; SSE2-NEXT: por %xmm5, %xmm7
+; SSE2-NEXT: movdqa %xmm2, %xmm5
+; SSE2-NEXT: pxor %xmm8, %xmm5
+; SSE2-NEXT: movdqa %xmm9, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm5, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm10 = xmm4[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm9, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm5[1,1,3,3]
+; SSE2-NEXT: pand %xmm10, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm6, %xmm5
+; SSE2-NEXT: movdqa {{.*#+}} xmm4 = [65535,65535]
+; SSE2-NEXT: pand %xmm5, %xmm2
+; SSE2-NEXT: pandn %xmm4, %xmm5
+; SSE2-NEXT: por %xmm2, %xmm5
+; SSE2-NEXT: pand %xmm7, %xmm3
+; SSE2-NEXT: pandn %xmm4, %xmm7
+; SSE2-NEXT: por %xmm3, %xmm7
+; SSE2-NEXT: pand %xmm12, %xmm0
+; SSE2-NEXT: pandn %xmm4, %xmm12
+; SSE2-NEXT: por %xmm0, %xmm12
+; SSE2-NEXT: pand %xmm11, %xmm1
+; SSE2-NEXT: pandn %xmm4, %xmm11
+; SSE2-NEXT: por %xmm1, %xmm11
+; SSE2-NEXT: movdqa %xmm11, %xmm0
+; SSE2-NEXT: pxor %xmm8, %xmm0
+; SSE2-NEXT: movdqa %xmm0, %xmm1
+; SSE2-NEXT: pcmpgtd %xmm8, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm1[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm8, %xmm0
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm0[1,1,3,3]
+; SSE2-NEXT: pand %xmm2, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
+; SSE2-NEXT: por %xmm3, %xmm0
+; SSE2-NEXT: movdqa %xmm12, %xmm1
+; SSE2-NEXT: pxor %xmm8, %xmm1
+; SSE2-NEXT: movdqa %xmm1, %xmm2
+; SSE2-NEXT: pcmpgtd %xmm8, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm2[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm8, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm1[1,1,3,3]
+; SSE2-NEXT: pand %xmm3, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm2[1,1,3,3]
+; SSE2-NEXT: por %xmm4, %xmm1
+; SSE2-NEXT: movdqa %xmm7, %xmm2
+; SSE2-NEXT: pxor %xmm8, %xmm2
+; SSE2-NEXT: movdqa %xmm2, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm8, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm3[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm8, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: pand %xmm4, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm3
+; SSE2-NEXT: movdqa %xmm5, %xmm2
+; SSE2-NEXT: pxor %xmm8, %xmm2
+; SSE2-NEXT: movdqa %xmm2, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm8, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm4[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm8, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: pand %xmm6, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm4
+; SSE2-NEXT: pand %xmm5, %xmm4
+; SSE2-NEXT: pand %xmm7, %xmm3
+; SSE2-NEXT: pand %xmm12, %xmm1
+; SSE2-NEXT: pand %xmm11, %xmm0
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
+; SSE2-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm3[0,2,2,3]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm0[0,1,0,2,4,5,6,7]
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm4[0,2,2,3]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,1,0,2,4,5,6,7]
+; SSE2-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
+; SSE2-NEXT: movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: trunc_packus_v8i64_v8i16:
+; SSSE3: # %bb.0:
+; SSSE3-NEXT: movdqa {{.*#+}} xmm8 = [2147483648,0,2147483648,0]
+; SSSE3-NEXT: movdqa %xmm1, %xmm4
+; SSSE3-NEXT: pxor %xmm8, %xmm4
+; SSSE3-NEXT: movdqa {{.*#+}} xmm9 = [2147549183,2147549183]
+; SSSE3-NEXT: movdqa %xmm9, %xmm6
+; SSSE3-NEXT: pcmpgtd %xmm4, %xmm6
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm9, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm4[1,1,3,3]
+; SSSE3-NEXT: pand %xmm7, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm11 = xmm6[1,1,3,3]
+; SSSE3-NEXT: por %xmm5, %xmm11
+; SSSE3-NEXT: movdqa %xmm0, %xmm5
+; SSSE3-NEXT: pxor %xmm8, %xmm5
+; SSSE3-NEXT: movdqa %xmm9, %xmm6
+; SSSE3-NEXT: pcmpgtd %xmm5, %xmm6
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm9, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSSE3-NEXT: pand %xmm7, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm12 = xmm6[1,1,3,3]
+; SSSE3-NEXT: por %xmm5, %xmm12
+; SSSE3-NEXT: movdqa %xmm3, %xmm5
+; SSSE3-NEXT: pxor %xmm8, %xmm5
+; SSSE3-NEXT: movdqa %xmm9, %xmm7
+; SSSE3-NEXT: pcmpgtd %xmm5, %xmm7
+; SSSE3-NEXT: pshufd {{.*#+}} xmm10 = xmm7[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm9, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSSE3-NEXT: pand %xmm10, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm7[1,1,3,3]
+; SSSE3-NEXT: por %xmm5, %xmm7
+; SSSE3-NEXT: movdqa %xmm2, %xmm5
+; SSSE3-NEXT: pxor %xmm8, %xmm5
+; SSSE3-NEXT: movdqa %xmm9, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm5, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm10 = xmm4[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm9, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm6 = xmm5[1,1,3,3]
+; SSSE3-NEXT: pand %xmm10, %xmm6
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm4[1,1,3,3]
+; SSSE3-NEXT: por %xmm6, %xmm5
+; SSSE3-NEXT: movdqa {{.*#+}} xmm4 = [65535,65535]
+; SSSE3-NEXT: pand %xmm5, %xmm2
+; SSSE3-NEXT: pandn %xmm4, %xmm5
+; SSSE3-NEXT: por %xmm2, %xmm5
+; SSSE3-NEXT: pand %xmm7, %xmm3
+; SSSE3-NEXT: pandn %xmm4, %xmm7
+; SSSE3-NEXT: por %xmm3, %xmm7
+; SSSE3-NEXT: pand %xmm12, %xmm0
+; SSSE3-NEXT: pandn %xmm4, %xmm12
+; SSSE3-NEXT: por %xmm0, %xmm12
+; SSSE3-NEXT: pand %xmm11, %xmm1
+; SSSE3-NEXT: pandn %xmm4, %xmm11
+; SSSE3-NEXT: por %xmm1, %xmm11
+; SSSE3-NEXT: movdqa %xmm11, %xmm0
+; SSSE3-NEXT: pxor %xmm8, %xmm0
+; SSSE3-NEXT: movdqa %xmm0, %xmm1
+; SSSE3-NEXT: pcmpgtd %xmm8, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm1[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm8, %xmm0
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm0[1,1,3,3]
+; SSSE3-NEXT: pand %xmm2, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
+; SSSE3-NEXT: por %xmm3, %xmm0
+; SSSE3-NEXT: movdqa %xmm12, %xmm1
+; SSSE3-NEXT: pxor %xmm8, %xmm1
+; SSSE3-NEXT: movdqa %xmm1, %xmm2
+; SSSE3-NEXT: pcmpgtd %xmm8, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm2[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm8, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm1[1,1,3,3]
+; SSSE3-NEXT: pand %xmm3, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm2[1,1,3,3]
+; SSSE3-NEXT: por %xmm4, %xmm1
+; SSSE3-NEXT: movdqa %xmm7, %xmm2
+; SSSE3-NEXT: pxor %xmm8, %xmm2
+; SSSE3-NEXT: movdqa %xmm2, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm8, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm3[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm8, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: pand %xmm4, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSSE3-NEXT: por %xmm2, %xmm3
+; SSSE3-NEXT: movdqa %xmm5, %xmm2
+; SSSE3-NEXT: pxor %xmm8, %xmm2
+; SSSE3-NEXT: movdqa %xmm2, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm8, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm6 = xmm4[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm8, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: pand %xmm6, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSSE3-NEXT: por %xmm2, %xmm4
+; SSSE3-NEXT: pand %xmm5, %xmm4
+; SSSE3-NEXT: pand %xmm7, %xmm3
+; SSSE3-NEXT: pand %xmm12, %xmm1
+; SSSE3-NEXT: pand %xmm11, %xmm0
+; SSSE3-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
+; SSSE3-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
+; SSSE3-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
+; SSSE3-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm0 = xmm3[0,2,2,3]
+; SSSE3-NEXT: pshuflw {{.*#+}} xmm2 = xmm0[0,1,0,2,4,5,6,7]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm0 = xmm4[0,2,2,3]
+; SSSE3-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,1,0,2,4,5,6,7]
+; SSSE3-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
+; SSSE3-NEXT: movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: trunc_packus_v8i64_v8i16:
+; SSE41: # %bb.0:
+; SSE41-NEXT: movdqa %xmm0, %xmm8
+; SSE41-NEXT: movdqa {{.*#+}} xmm11 = [2147483648,0,2147483648,0]
+; SSE41-NEXT: movdqa %xmm3, %xmm5
+; SSE41-NEXT: pxor %xmm11, %xmm5
+; SSE41-NEXT: movdqa {{.*#+}} xmm0 = [2147549183,2147549183]
+; SSE41-NEXT: movdqa %xmm0, %xmm6
+; SSE41-NEXT: pcmpgtd %xmm5, %xmm6
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm0, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm9 = xmm6[1,1,3,3]
+; SSE41-NEXT: por %xmm5, %xmm9
+; SSE41-NEXT: movdqa %xmm2, %xmm5
+; SSE41-NEXT: pxor %xmm11, %xmm5
+; SSE41-NEXT: movdqa %xmm0, %xmm6
+; SSE41-NEXT: pcmpgtd %xmm5, %xmm6
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm0, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm12 = xmm6[1,1,3,3]
+; SSE41-NEXT: por %xmm5, %xmm12
+; SSE41-NEXT: movdqa %xmm1, %xmm5
+; SSE41-NEXT: pxor %xmm11, %xmm5
+; SSE41-NEXT: movdqa %xmm0, %xmm7
+; SSE41-NEXT: pcmpgtd %xmm5, %xmm7
+; SSE41-NEXT: pshufd {{.*#+}} xmm4 = xmm7[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm0, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE41-NEXT: pand %xmm4, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm7[1,1,3,3]
+; SSE41-NEXT: por %xmm5, %xmm7
+; SSE41-NEXT: movdqa %xmm8, %xmm4
+; SSE41-NEXT: pxor %xmm11, %xmm4
+; SSE41-NEXT: movdqa %xmm0, %xmm5
+; SSE41-NEXT: pcmpgtd %xmm4, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm5[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm0, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSE41-NEXT: pand %xmm6, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm5[1,1,3,3]
+; SSE41-NEXT: por %xmm4, %xmm0
+; SSE41-NEXT: movapd {{.*#+}} xmm5 = [65535,65535]
+; SSE41-NEXT: movapd %xmm5, %xmm10
+; SSE41-NEXT: blendvpd %xmm0, %xmm8, %xmm10
+; SSE41-NEXT: movapd %xmm5, %xmm8
+; SSE41-NEXT: movdqa %xmm7, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm1, %xmm8
+; SSE41-NEXT: movapd %xmm5, %xmm13
+; SSE41-NEXT: movdqa %xmm12, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm2, %xmm13
+; SSE41-NEXT: movdqa %xmm9, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm3, %xmm5
+; SSE41-NEXT: movapd %xmm5, %xmm0
+; SSE41-NEXT: xorpd %xmm11, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm2
+; SSE41-NEXT: pcmpgtd %xmm11, %xmm2
+; SSE41-NEXT: pshufd {{.*#+}} xmm3 = xmm2[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm11, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm3, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm2
+; SSE41-NEXT: movapd %xmm13, %xmm0
+; SSE41-NEXT: xorpd %xmm11, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm3
+; SSE41-NEXT: pcmpgtd %xmm11, %xmm3
+; SSE41-NEXT: pshufd {{.*#+}} xmm4 = xmm3[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm11, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm4, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm3
+; SSE41-NEXT: movapd %xmm8, %xmm0
+; SSE41-NEXT: xorpd %xmm11, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm4
+; SSE41-NEXT: pcmpgtd %xmm11, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm4[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm11, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm6, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm4[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm6
+; SSE41-NEXT: movapd %xmm10, %xmm0
+; SSE41-NEXT: xorpd %xmm11, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm4
+; SSE41-NEXT: pcmpgtd %xmm11, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm4[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm11, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm1
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm4[1,1,3,3]
+; SSE41-NEXT: por %xmm1, %xmm0
+; SSE41-NEXT: pand %xmm10, %xmm0
+; SSE41-NEXT: pand %xmm8, %xmm6
+; SSE41-NEXT: pand %xmm13, %xmm3
+; SSE41-NEXT: pand %xmm5, %xmm2
+; SSE41-NEXT: pxor %xmm1, %xmm1
+; SSE41-NEXT: pblendw {{.*#+}} xmm2 = xmm2[0],xmm1[1,2,3],xmm2[4],xmm1[5,6,7]
+; SSE41-NEXT: pblendw {{.*#+}} xmm3 = xmm3[0],xmm1[1,2,3],xmm3[4],xmm1[5,6,7]
+; SSE41-NEXT: packusdw %xmm2, %xmm3
+; SSE41-NEXT: pblendw {{.*#+}} xmm6 = xmm6[0],xmm1[1,2,3],xmm6[4],xmm1[5,6,7]
+; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3],xmm0[4],xmm1[5,6,7]
+; SSE41-NEXT: packusdw %xmm6, %xmm0
+; SSE41-NEXT: packusdw %xmm3, %xmm0
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_packus_v8i64_v8i16:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm3 = [65535,65535]
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm3, %xmm2
+; AVX1-NEXT: vpcmpgtq %xmm0, %xmm3, %xmm4
+; AVX1-NEXT: vinsertf128 $1, %xmm2, %ymm4, %ymm2
+; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm4
+; AVX1-NEXT: vpcmpgtq %xmm4, %xmm3, %xmm4
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm3, %xmm3
+; AVX1-NEXT: vinsertf128 $1, %xmm4, %ymm3, %ymm3
+; AVX1-NEXT: vmovapd {{.*#+}} ymm4 = [65535,65535,65535,65535]
+; AVX1-NEXT: vblendvpd %ymm3, %ymm1, %ymm4, %ymm1
+; AVX1-NEXT: vblendvpd %ymm2, %ymm0, %ymm4, %ymm0
+; AVX1-NEXT: vxorpd %xmm2, %xmm2, %xmm2
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm0, %xmm8
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm4
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm4, %xmm5
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm1, %xmm6
+; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm7
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm7, %xmm3
+; AVX1-NEXT: vpand %xmm7, %xmm3, %xmm3
+; AVX1-NEXT: vpblendw {{.*#+}} xmm3 = xmm3[0],xmm2[1,2,3],xmm3[4],xmm2[5,6,7]
+; AVX1-NEXT: vpand %xmm1, %xmm6, %xmm1
+; AVX1-NEXT: vpblendw {{.*#+}} xmm1 = xmm1[0],xmm2[1,2,3],xmm1[4],xmm2[5,6,7]
+; AVX1-NEXT: vpackusdw %xmm3, %xmm1, %xmm1
+; AVX1-NEXT: vpand %xmm4, %xmm5, %xmm3
+; AVX1-NEXT: vpblendw {{.*#+}} xmm3 = xmm3[0],xmm2[1,2,3],xmm3[4],xmm2[5,6,7]
+; AVX1-NEXT: vpand %xmm0, %xmm8, %xmm0
+; AVX1-NEXT: vpblendw {{.*#+}} xmm0 = xmm0[0],xmm2[1,2,3],xmm0[4],xmm2[5,6,7]
+; AVX1-NEXT: vpackusdw %xmm3, %xmm0, %xmm0
+; AVX1-NEXT: vpackusdw %xmm1, %xmm0, %xmm0
+; AVX1-NEXT: vzeroupper
+; AVX1-NEXT: retq
+;
+; AVX2-SLOW-LABEL: trunc_packus_v8i64_v8i16:
+; AVX2-SLOW: # %bb.0:
+; AVX2-SLOW-NEXT: vpbroadcastq {{.*#+}} ymm2 = [65535,65535,65535,65535]
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm0, %ymm2, %ymm3
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm1, %ymm2, %ymm4
+; AVX2-SLOW-NEXT: vblendvpd %ymm4, %ymm1, %ymm2, %ymm1
+; AVX2-SLOW-NEXT: vblendvpd %ymm3, %ymm0, %ymm2, %ymm0
+; AVX2-SLOW-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm2, %ymm0, %ymm3
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm2, %ymm1, %ymm2
+; AVX2-SLOW-NEXT: vpand %ymm1, %ymm2, %ymm1
+; AVX2-SLOW-NEXT: vpand %ymm0, %ymm3, %ymm0
+; AVX2-SLOW-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[0,2,2,3,4,6,6,7]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-SLOW-NEXT: vpshufd {{.*#+}} ymm1 = ymm1[0,2,2,3,4,6,6,7]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,2,2,3]
+; AVX2-SLOW-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-SLOW-NEXT: vpshufb {{.*#+}} ymm0 = ymm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-SLOW-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX2-SLOW-NEXT: vzeroupper
+; AVX2-SLOW-NEXT: retq
+;
+; AVX2-FAST-LABEL: trunc_packus_v8i64_v8i16:
+; AVX2-FAST: # %bb.0:
+; AVX2-FAST-NEXT: vpbroadcastq {{.*#+}} ymm2 = [65535,65535,65535,65535]
+; AVX2-FAST-NEXT: vpcmpgtq %ymm0, %ymm2, %ymm3
+; AVX2-FAST-NEXT: vpcmpgtq %ymm1, %ymm2, %ymm4
+; AVX2-FAST-NEXT: vblendvpd %ymm4, %ymm1, %ymm2, %ymm1
+; AVX2-FAST-NEXT: vblendvpd %ymm3, %ymm0, %ymm2, %ymm0
+; AVX2-FAST-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX2-FAST-NEXT: vpcmpgtq %ymm2, %ymm0, %ymm3
+; AVX2-FAST-NEXT: vpcmpgtq %ymm2, %ymm1, %ymm2
+; AVX2-FAST-NEXT: vpand %ymm1, %ymm2, %ymm1
+; AVX2-FAST-NEXT: vpand %ymm0, %ymm3, %ymm0
+; AVX2-FAST-NEXT: vmovdqa {{.*#+}} ymm2 = [0,2,4,6,4,6,6,7]
+; AVX2-FAST-NEXT: vpermd %ymm0, %ymm2, %ymm0
+; AVX2-FAST-NEXT: vpermd %ymm1, %ymm2, %ymm1
+; AVX2-FAST-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-FAST-NEXT: vpshufb {{.*#+}} ymm0 = ymm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-FAST-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-FAST-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX2-FAST-NEXT: vzeroupper
+; AVX2-FAST-NEXT: retq
+;
+; AVX512-LABEL: trunc_packus_v8i64_v8i16:
+; AVX512: # %bb.0:
+; AVX512-NEXT: vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
+; AVX512-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512-NEXT: vpmaxsq %zmm1, %zmm0, %zmm0
+; AVX512-NEXT: vpmovqw %zmm0, %xmm0
+; AVX512-NEXT: vzeroupper
+; AVX512-NEXT: retq
+ %1 = icmp slt <8 x i64> %a0, <i64 65535, i64 65535, i64 65535, i64 65535, i64 65535, i64 65535, i64 65535, i64 65535>
+ %2 = select <8 x i1> %1, <8 x i64> %a0, <8 x i64> <i64 65535, i64 65535, i64 65535, i64 65535, i64 65535, i64 65535, i64 65535, i64 65535>
+ %3 = icmp sgt <8 x i64> %2, zeroinitializer
+ %4 = select <8 x i1> %3, <8 x i64> %2, <8 x i64> zeroinitializer
+ %5 = trunc <8 x i64> %4 to <8 x i16>
+ ret <8 x i16> %5
+}
+
+define <8 x i16> @trunc_packus_v8i32_v8i16(<8 x i32> %a0) {
+; SSE2-LABEL: trunc_packus_v8i32_v8i16:
+; SSE2: # %bb.0:
+; SSE2-NEXT: movdqa {{.*#+}} xmm2 = [65535,65535,65535,65535]
+; SSE2-NEXT: movdqa %xmm2, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm3
+; SSE2-NEXT: movdqa %xmm2, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm0, %xmm4
+; SSE2-NEXT: pand %xmm4, %xmm0
+; SSE2-NEXT: pandn %xmm2, %xmm4
+; SSE2-NEXT: por %xmm0, %xmm4
+; SSE2-NEXT: pand %xmm3, %xmm1
+; SSE2-NEXT: pandn %xmm2, %xmm3
+; SSE2-NEXT: por %xmm1, %xmm3
+; SSE2-NEXT: pxor %xmm1, %xmm1
+; SSE2-NEXT: movdqa %xmm3, %xmm2
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm2
+; SSE2-NEXT: movdqa %xmm4, %xmm0
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm0
+; SSE2-NEXT: pand %xmm4, %xmm0
+; SSE2-NEXT: pand %xmm3, %xmm2
+; SSE2-NEXT: pslld $16, %xmm2
+; SSE2-NEXT: psrad $16, %xmm2
+; SSE2-NEXT: pslld $16, %xmm0
+; SSE2-NEXT: psrad $16, %xmm0
+; SSE2-NEXT: packssdw %xmm2, %xmm0
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: trunc_packus_v8i32_v8i16:
+; SSSE3: # %bb.0:
+; SSSE3-NEXT: movdqa {{.*#+}} xmm2 = [65535,65535,65535,65535]
+; SSSE3-NEXT: movdqa %xmm2, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm3
+; SSSE3-NEXT: movdqa %xmm2, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm0, %xmm4
+; SSSE3-NEXT: pand %xmm4, %xmm0
+; SSSE3-NEXT: pandn %xmm2, %xmm4
+; SSSE3-NEXT: por %xmm0, %xmm4
+; SSSE3-NEXT: pand %xmm3, %xmm1
+; SSSE3-NEXT: pandn %xmm2, %xmm3
+; SSSE3-NEXT: por %xmm1, %xmm3
+; SSSE3-NEXT: pxor %xmm1, %xmm1
+; SSSE3-NEXT: movdqa %xmm3, %xmm2
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm2
+; SSSE3-NEXT: movdqa %xmm4, %xmm0
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm0
+; SSSE3-NEXT: pand %xmm4, %xmm0
+; SSSE3-NEXT: pand %xmm3, %xmm2
+; SSSE3-NEXT: movdqa {{.*#+}} xmm1 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
+; SSSE3-NEXT: pshufb %xmm1, %xmm2
+; SSSE3-NEXT: pshufb %xmm1, %xmm0
+; SSSE3-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: trunc_packus_v8i32_v8i16:
+; SSE41: # %bb.0:
+; SSE41-NEXT: movdqa {{.*#+}} xmm2 = [65535,65535,65535,65535]
+; SSE41-NEXT: pminsd %xmm2, %xmm1
+; SSE41-NEXT: pminsd %xmm2, %xmm0
+; SSE41-NEXT: pxor %xmm2, %xmm2
+; SSE41-NEXT: pmaxsd %xmm2, %xmm0
+; SSE41-NEXT: pmaxsd %xmm2, %xmm1
+; SSE41-NEXT: movdqa {{.*#+}} xmm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
+; SSE41-NEXT: pshufb %xmm2, %xmm1
+; SSE41-NEXT: pshufb %xmm2, %xmm0
+; SSE41-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_packus_v8i32_v8i16:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm2 = [65535,65535,65535,65535]
+; AVX1-NEXT: vpminsd %xmm2, %xmm1, %xmm1
+; AVX1-NEXT: vpminsd %xmm2, %xmm0, %xmm0
+; AVX1-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX1-NEXT: vpmaxsd %xmm2, %xmm0, %xmm0
+; AVX1-NEXT: vpmaxsd %xmm2, %xmm1, %xmm1
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
+; AVX1-NEXT: vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT: vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX1-NEXT: vzeroupper
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: trunc_packus_v8i32_v8i16:
+; AVX2: # %bb.0:
+; AVX2-NEXT: vpbroadcastd {{.*#+}} ymm1 = [65535,65535,65535,65535,65535,65535,65535,65535]
+; AVX2-NEXT: vpminsd %ymm1, %ymm0, %ymm0
+; AVX2-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX2-NEXT: vpmaxsd %ymm1, %ymm0, %ymm0
+; AVX2-NEXT: vpshufb {{.*#+}} ymm0 = ymm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX2-NEXT: vzeroupper
+; AVX2-NEXT: retq
+;
+; AVX512F-LABEL: trunc_packus_v8i32_v8i16:
+; AVX512F: # %bb.0:
+; AVX512F-NEXT: vpbroadcastd {{.*#+}} ymm1 = [65535,65535,65535,65535,65535,65535,65535,65535]
+; AVX512F-NEXT: vpminsd %ymm1, %ymm0, %ymm0
+; AVX512F-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512F-NEXT: vpmaxsd %ymm1, %ymm0, %ymm0
+; AVX512F-NEXT: vpmovdw %zmm0, %ymm0
+; AVX512F-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX512F-NEXT: vzeroupper
+; AVX512F-NEXT: retq
+;
+; AVX512VL-LABEL: trunc_packus_v8i32_v8i16:
+; AVX512VL: # %bb.0:
+; AVX512VL-NEXT: vpminsd {{.*}}(%rip){1to8}, %ymm0, %ymm0
+; AVX512VL-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512VL-NEXT: vpmaxsd %ymm1, %ymm0, %ymm0
+; AVX512VL-NEXT: vpmovdw %ymm0, %xmm0
+; AVX512VL-NEXT: vzeroupper
+; AVX512VL-NEXT: retq
+;
+; AVX512BW-LABEL: trunc_packus_v8i32_v8i16:
+; AVX512BW: # %bb.0:
+; AVX512BW-NEXT: vpbroadcastd {{.*#+}} ymm1 = [65535,65535,65535,65535,65535,65535,65535,65535]
+; AVX512BW-NEXT: vpminsd %ymm1, %ymm0, %ymm0
+; AVX512BW-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512BW-NEXT: vpmaxsd %ymm1, %ymm0, %ymm0
+; AVX512BW-NEXT: vpmovdw %zmm0, %ymm0
+; AVX512BW-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX512BW-NEXT: vzeroupper
+; AVX512BW-NEXT: retq
+;
+; AVX512BWVL-LABEL: trunc_packus_v8i32_v8i16:
+; AVX512BWVL: # %bb.0:
+; AVX512BWVL-NEXT: vpminsd {{.*}}(%rip){1to8}, %ymm0, %ymm0
+; AVX512BWVL-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512BWVL-NEXT: vpmaxsd %ymm1, %ymm0, %ymm0
+; AVX512BWVL-NEXT: vpmovdw %ymm0, %xmm0
+; AVX512BWVL-NEXT: vzeroupper
+; AVX512BWVL-NEXT: retq
+ %1 = icmp slt <8 x i32> %a0, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+ %2 = select <8 x i1> %1, <8 x i32> %a0, <8 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+ %3 = icmp sgt <8 x i32> %2, zeroinitializer
+ %4 = select <8 x i1> %3, <8 x i32> %2, <8 x i32> zeroinitializer
+ %5 = trunc <8 x i32> %4 to <8 x i16>
+ ret <8 x i16> %5
+}
+
+define <16 x i16> @trunc_packus_v16i32_v16i16(<16 x i32> %a0) {
+; SSE2-LABEL: trunc_packus_v16i32_v16i16:
+; SSE2: # %bb.0:
+; SSE2-NEXT: movdqa {{.*#+}} xmm8 = [65535,65535,65535,65535]
+; SSE2-NEXT: movdqa %xmm8, %xmm9
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm9
+; SSE2-NEXT: movdqa %xmm8, %xmm5
+; SSE2-NEXT: pcmpgtd %xmm0, %xmm5
+; SSE2-NEXT: movdqa %xmm8, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm3, %xmm4
+; SSE2-NEXT: movdqa %xmm8, %xmm7
+; SSE2-NEXT: pcmpgtd %xmm2, %xmm7
+; SSE2-NEXT: pand %xmm7, %xmm2
+; SSE2-NEXT: pandn %xmm8, %xmm7
+; SSE2-NEXT: por %xmm2, %xmm7
+; SSE2-NEXT: pand %xmm4, %xmm3
+; SSE2-NEXT: pandn %xmm8, %xmm4
+; SSE2-NEXT: por %xmm3, %xmm4
+; SSE2-NEXT: pand %xmm5, %xmm0
+; SSE2-NEXT: pandn %xmm8, %xmm5
+; SSE2-NEXT: por %xmm0, %xmm5
+; SSE2-NEXT: pand %xmm9, %xmm1
+; SSE2-NEXT: pandn %xmm8, %xmm9
+; SSE2-NEXT: por %xmm1, %xmm9
+; SSE2-NEXT: pxor %xmm2, %xmm2
+; SSE2-NEXT: movdqa %xmm9, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm2, %xmm3
+; SSE2-NEXT: movdqa %xmm5, %xmm0
+; SSE2-NEXT: pcmpgtd %xmm2, %xmm0
+; SSE2-NEXT: movdqa %xmm4, %xmm6
+; SSE2-NEXT: pcmpgtd %xmm2, %xmm6
+; SSE2-NEXT: movdqa %xmm7, %xmm1
+; SSE2-NEXT: pcmpgtd %xmm2, %xmm1
+; SSE2-NEXT: pand %xmm7, %xmm1
+; SSE2-NEXT: pand %xmm4, %xmm6
+; SSE2-NEXT: pand %xmm5, %xmm0
+; SSE2-NEXT: pand %xmm9, %xmm3
+; SSE2-NEXT: pslld $16, %xmm3
+; SSE2-NEXT: psrad $16, %xmm3
+; SSE2-NEXT: pslld $16, %xmm0
+; SSE2-NEXT: psrad $16, %xmm0
+; SSE2-NEXT: packssdw %xmm3, %xmm0
+; SSE2-NEXT: pslld $16, %xmm6
+; SSE2-NEXT: psrad $16, %xmm6
+; SSE2-NEXT: pslld $16, %xmm1
+; SSE2-NEXT: psrad $16, %xmm1
+; SSE2-NEXT: packssdw %xmm6, %xmm1
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: trunc_packus_v16i32_v16i16:
+; SSSE3: # %bb.0:
+; SSSE3-NEXT: movdqa {{.*#+}} xmm8 = [65535,65535,65535,65535]
+; SSSE3-NEXT: movdqa %xmm8, %xmm9
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm9
+; SSSE3-NEXT: movdqa %xmm8, %xmm5
+; SSSE3-NEXT: pcmpgtd %xmm0, %xmm5
+; SSSE3-NEXT: movdqa %xmm8, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm3, %xmm4
+; SSSE3-NEXT: movdqa %xmm8, %xmm7
+; SSSE3-NEXT: pcmpgtd %xmm2, %xmm7
+; SSSE3-NEXT: pand %xmm7, %xmm2
+; SSSE3-NEXT: pandn %xmm8, %xmm7
+; SSSE3-NEXT: por %xmm2, %xmm7
+; SSSE3-NEXT: pand %xmm4, %xmm3
+; SSSE3-NEXT: pandn %xmm8, %xmm4
+; SSSE3-NEXT: por %xmm3, %xmm4
+; SSSE3-NEXT: pand %xmm5, %xmm0
+; SSSE3-NEXT: pandn %xmm8, %xmm5
+; SSSE3-NEXT: por %xmm0, %xmm5
+; SSSE3-NEXT: pand %xmm9, %xmm1
+; SSSE3-NEXT: pandn %xmm8, %xmm9
+; SSSE3-NEXT: por %xmm1, %xmm9
+; SSSE3-NEXT: pxor %xmm2, %xmm2
+; SSSE3-NEXT: movdqa %xmm9, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm2, %xmm3
+; SSSE3-NEXT: movdqa %xmm5, %xmm0
+; SSSE3-NEXT: pcmpgtd %xmm2, %xmm0
+; SSSE3-NEXT: movdqa %xmm4, %xmm6
+; SSSE3-NEXT: pcmpgtd %xmm2, %xmm6
+; SSSE3-NEXT: movdqa %xmm7, %xmm1
+; SSSE3-NEXT: pcmpgtd %xmm2, %xmm1
+; SSSE3-NEXT: pand %xmm7, %xmm1
+; SSSE3-NEXT: pand %xmm4, %xmm6
+; SSSE3-NEXT: pand %xmm5, %xmm0
+; SSSE3-NEXT: pand %xmm9, %xmm3
+; SSSE3-NEXT: pslld $16, %xmm3
+; SSSE3-NEXT: psrad $16, %xmm3
+; SSSE3-NEXT: pslld $16, %xmm0
+; SSSE3-NEXT: psrad $16, %xmm0
+; SSSE3-NEXT: packssdw %xmm3, %xmm0
+; SSSE3-NEXT: pslld $16, %xmm6
+; SSSE3-NEXT: psrad $16, %xmm6
+; SSSE3-NEXT: pslld $16, %xmm1
+; SSSE3-NEXT: psrad $16, %xmm1
+; SSSE3-NEXT: packssdw %xmm6, %xmm1
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: trunc_packus_v16i32_v16i16:
+; SSE41: # %bb.0:
+; SSE41-NEXT: movdqa {{.*#+}} xmm4 = [65535,65535,65535,65535]
+; SSE41-NEXT: pminsd %xmm4, %xmm1
+; SSE41-NEXT: pminsd %xmm4, %xmm0
+; SSE41-NEXT: pminsd %xmm4, %xmm3
+; SSE41-NEXT: pminsd %xmm2, %xmm4
+; SSE41-NEXT: pxor %xmm2, %xmm2
+; SSE41-NEXT: pmaxsd %xmm2, %xmm4
+; SSE41-NEXT: pmaxsd %xmm2, %xmm3
+; SSE41-NEXT: pmaxsd %xmm2, %xmm0
+; SSE41-NEXT: pmaxsd %xmm2, %xmm1
+; SSE41-NEXT: pblendw {{.*#+}} xmm1 = xmm1[0],xmm2[1],xmm1[2],xmm2[3],xmm1[4],xmm2[5],xmm1[6],xmm2[7]
+; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm0[0],xmm2[1],xmm0[2],xmm2[3],xmm0[4],xmm2[5],xmm0[6],xmm2[7]
+; SSE41-NEXT: packusdw %xmm1, %xmm0
+; SSE41-NEXT: pblendw {{.*#+}} xmm3 = xmm3[0],xmm2[1],xmm3[2],xmm2[3],xmm3[4],xmm2[5],xmm3[6],xmm2[7]
+; SSE41-NEXT: pblendw {{.*#+}} xmm4 = xmm4[0],xmm2[1],xmm4[2],xmm2[3],xmm4[4],xmm2[5],xmm4[6],xmm2[7]
+; SSE41-NEXT: packusdw %xmm3, %xmm4
+; SSE41-NEXT: movdqa %xmm4, %xmm1
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_packus_v16i32_v16i16:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm3 = [65535,65535,65535,65535]
+; AVX1-NEXT: vpminsd %xmm3, %xmm2, %xmm2
+; AVX1-NEXT: vpminsd %xmm3, %xmm1, %xmm1
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm4
+; AVX1-NEXT: vpminsd %xmm3, %xmm4, %xmm4
+; AVX1-NEXT: vpminsd %xmm3, %xmm0, %xmm0
+; AVX1-NEXT: vpxor %xmm3, %xmm3, %xmm3
+; AVX1-NEXT: vpmaxsd %xmm3, %xmm0, %xmm0
+; AVX1-NEXT: vpmaxsd %xmm3, %xmm4, %xmm4
+; AVX1-NEXT: vpmaxsd %xmm3, %xmm1, %xmm1
+; AVX1-NEXT: vpmaxsd %xmm3, %xmm2, %xmm2
+; AVX1-NEXT: vpblendw {{.*#+}} xmm2 = xmm2[0],xmm3[1],xmm2[2],xmm3[3],xmm2[4],xmm3[5],xmm2[6],xmm3[7]
+; AVX1-NEXT: vpblendw {{.*#+}} xmm1 = xmm1[0],xmm3[1],xmm1[2],xmm3[3],xmm1[4],xmm3[5],xmm1[6],xmm3[7]
+; AVX1-NEXT: vpackusdw %xmm2, %xmm1, %xmm1
+; AVX1-NEXT: vpblendw {{.*#+}} xmm2 = xmm4[0],xmm3[1],xmm4[2],xmm3[3],xmm4[4],xmm3[5],xmm4[6],xmm3[7]
+; AVX1-NEXT: vpblendw {{.*#+}} xmm0 = xmm0[0],xmm3[1],xmm0[2],xmm3[3],xmm0[4],xmm3[5],xmm0[6],xmm3[7]
+; AVX1-NEXT: vpackusdw %xmm2, %xmm0, %xmm0
+; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: trunc_packus_v16i32_v16i16:
+; AVX2: # %bb.0:
+; AVX2-NEXT: vpbroadcastd {{.*#+}} ymm2 = [65535,65535,65535,65535,65535,65535,65535,65535]
+; AVX2-NEXT: vpminsd %ymm2, %ymm0, %ymm0
+; AVX2-NEXT: vpminsd %ymm2, %ymm1, %ymm1
+; AVX2-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX2-NEXT: vpmaxsd %ymm2, %ymm1, %ymm1
+; AVX2-NEXT: vpmaxsd %ymm2, %ymm0, %ymm0
+; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-NEXT: vpshufb %ymm2, %ymm0, %ymm0
+; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-NEXT: vpshufb %ymm2, %ymm1, %ymm1
+; AVX2-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,2,2,3]
+; AVX2-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT: retq
+;
+; AVX512-LABEL: trunc_packus_v16i32_v16i16:
+; AVX512: # %bb.0:
+; AVX512-NEXT: vpminsd {{.*}}(%rip){1to16}, %zmm0, %zmm0
+; AVX512-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512-NEXT: vpmaxsd %zmm1, %zmm0, %zmm0
+; AVX512-NEXT: vpmovdw %zmm0, %ymm0
+; AVX512-NEXT: retq
+ %1 = icmp slt <16 x i32> %a0, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+ %2 = select <16 x i1> %1, <16 x i32> %a0, <16 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+ %3 = icmp sgt <16 x i32> %2, zeroinitializer
+ %4 = select <16 x i1> %3, <16 x i32> %2, <16 x i32> zeroinitializer
+ %5 = trunc <16 x i32> %4 to <16 x i16>
+ ret <16 x i16> %5
+}
+
+;
+; PACKUS saturation truncation to v16i8
+;
+
+define <8 x i8> @trunc_packus_v8i64_v8i8(<8 x i64> %a0) {
+; SSE2-LABEL: trunc_packus_v8i64_v8i8:
+; SSE2: # %bb.0:
+; SSE2-NEXT: movdqa {{.*#+}} xmm8 = [2147483648,0,2147483648,0]
+; SSE2-NEXT: movdqa %xmm1, %xmm4
+; SSE2-NEXT: pxor %xmm8, %xmm4
+; SSE2-NEXT: movdqa {{.*#+}} xmm9 = [2147483903,2147483903]
+; SSE2-NEXT: movdqa %xmm9, %xmm6
+; SSE2-NEXT: pcmpgtd %xmm4, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm9, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm4[1,1,3,3]
+; SSE2-NEXT: pand %xmm7, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm11 = xmm6[1,1,3,3]
+; SSE2-NEXT: por %xmm5, %xmm11
+; SSE2-NEXT: movdqa %xmm0, %xmm5
+; SSE2-NEXT: pxor %xmm8, %xmm5
+; SSE2-NEXT: movdqa %xmm9, %xmm6
+; SSE2-NEXT: pcmpgtd %xmm5, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm9, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE2-NEXT: pand %xmm7, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm12 = xmm6[1,1,3,3]
+; SSE2-NEXT: por %xmm5, %xmm12
+; SSE2-NEXT: movdqa %xmm3, %xmm5
+; SSE2-NEXT: pxor %xmm8, %xmm5
+; SSE2-NEXT: movdqa %xmm9, %xmm7
+; SSE2-NEXT: pcmpgtd %xmm5, %xmm7
+; SSE2-NEXT: pshufd {{.*#+}} xmm10 = xmm7[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm9, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE2-NEXT: pand %xmm10, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm7[1,1,3,3]
+; SSE2-NEXT: por %xmm5, %xmm7
+; SSE2-NEXT: movdqa %xmm2, %xmm5
+; SSE2-NEXT: pxor %xmm8, %xmm5
+; SSE2-NEXT: movdqa %xmm9, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm5, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm10 = xmm4[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm9, %xmm5
+; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm5[1,1,3,3]
+; SSE2-NEXT: pand %xmm10, %xmm6
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm6, %xmm5
+; SSE2-NEXT: movdqa {{.*#+}} xmm4 = [255,255]
+; SSE2-NEXT: pand %xmm5, %xmm2
+; SSE2-NEXT: pandn %xmm4, %xmm5
+; SSE2-NEXT: por %xmm2, %xmm5
+; SSE2-NEXT: pand %xmm7, %xmm3
+; SSE2-NEXT: pandn %xmm4, %xmm7
+; SSE2-NEXT: por %xmm3, %xmm7
+; SSE2-NEXT: pand %xmm12, %xmm0
+; SSE2-NEXT: pandn %xmm4, %xmm12
+; SSE2-NEXT: por %xmm0, %xmm12
+; SSE2-NEXT: pand %xmm11, %xmm1
+; SSE2-NEXT: pandn %xmm4, %xmm11
+; SSE2-NEXT: por %xmm1, %xmm11
+; SSE2-NEXT: movdqa %xmm11, %xmm0
+; SSE2-NEXT: pxor %xmm8, %xmm0
+; SSE2-NEXT: movdqa %xmm0, %xmm1
+; SSE2-NEXT: pcmpgtd %xmm8, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm1[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm8, %xmm0
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm0[1,1,3,3]
+; SSE2-NEXT: pand %xmm2, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
+; SSE2-NEXT: por %xmm3, %xmm0
+; SSE2-NEXT: movdqa %xmm12, %xmm1
+; SSE2-NEXT: pxor %xmm8, %xmm1
+; SSE2-NEXT: movdqa %xmm1, %xmm2
+; SSE2-NEXT: pcmpgtd %xmm8, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm2[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm8, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm1[1,1,3,3]
+; SSE2-NEXT: pand %xmm3, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm2[1,1,3,3]
+; SSE2-NEXT: por %xmm4, %xmm1
+; SSE2-NEXT: movdqa %xmm7, %xmm2
+; SSE2-NEXT: pxor %xmm8, %xmm2
+; SSE2-NEXT: movdqa %xmm2, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm8, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm3[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm8, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: pand %xmm4, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm3
+; SSE2-NEXT: movdqa %xmm5, %xmm2
+; SSE2-NEXT: pxor %xmm8, %xmm2
+; SSE2-NEXT: movdqa %xmm2, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm8, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm4[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm8, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: pand %xmm6, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm4
+; SSE2-NEXT: pand %xmm5, %xmm4
+; SSE2-NEXT: pand %xmm7, %xmm3
+; SSE2-NEXT: pand %xmm12, %xmm1
+; SSE2-NEXT: pand %xmm11, %xmm0
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
+; SSE2-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm3[0,2,2,3]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm2 = xmm0[0,1,0,2,4,5,6,7]
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm4[0,2,2,3]
+; SSE2-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,1,0,2,4,5,6,7]
+; SSE2-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
+; SSE2-NEXT: movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: trunc_packus_v8i64_v8i8:
+; SSSE3: # %bb.0:
+; SSSE3-NEXT: movdqa {{.*#+}} xmm8 = [2147483648,0,2147483648,0]
+; SSSE3-NEXT: movdqa %xmm1, %xmm4
+; SSSE3-NEXT: pxor %xmm8, %xmm4
+; SSSE3-NEXT: movdqa {{.*#+}} xmm9 = [2147483903,2147483903]
+; SSSE3-NEXT: movdqa %xmm9, %xmm6
+; SSSE3-NEXT: pcmpgtd %xmm4, %xmm6
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm9, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm4[1,1,3,3]
+; SSSE3-NEXT: pand %xmm7, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm11 = xmm6[1,1,3,3]
+; SSSE3-NEXT: por %xmm5, %xmm11
+; SSSE3-NEXT: movdqa %xmm0, %xmm5
+; SSSE3-NEXT: pxor %xmm8, %xmm5
+; SSSE3-NEXT: movdqa %xmm9, %xmm6
+; SSSE3-NEXT: pcmpgtd %xmm5, %xmm6
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm9, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSSE3-NEXT: pand %xmm7, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm12 = xmm6[1,1,3,3]
+; SSSE3-NEXT: por %xmm5, %xmm12
+; SSSE3-NEXT: movdqa %xmm3, %xmm5
+; SSSE3-NEXT: pxor %xmm8, %xmm5
+; SSSE3-NEXT: movdqa %xmm9, %xmm7
+; SSSE3-NEXT: pcmpgtd %xmm5, %xmm7
+; SSSE3-NEXT: pshufd {{.*#+}} xmm10 = xmm7[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm9, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSSE3-NEXT: pand %xmm10, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm7[1,1,3,3]
+; SSSE3-NEXT: por %xmm5, %xmm7
+; SSSE3-NEXT: movdqa %xmm2, %xmm5
+; SSSE3-NEXT: pxor %xmm8, %xmm5
+; SSSE3-NEXT: movdqa %xmm9, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm5, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm10 = xmm4[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm9, %xmm5
+; SSSE3-NEXT: pshufd {{.*#+}} xmm6 = xmm5[1,1,3,3]
+; SSSE3-NEXT: pand %xmm10, %xmm6
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm4[1,1,3,3]
+; SSSE3-NEXT: por %xmm6, %xmm5
+; SSSE3-NEXT: movdqa {{.*#+}} xmm4 = [255,255]
+; SSSE3-NEXT: pand %xmm5, %xmm2
+; SSSE3-NEXT: pandn %xmm4, %xmm5
+; SSSE3-NEXT: por %xmm2, %xmm5
+; SSSE3-NEXT: pand %xmm7, %xmm3
+; SSSE3-NEXT: pandn %xmm4, %xmm7
+; SSSE3-NEXT: por %xmm3, %xmm7
+; SSSE3-NEXT: pand %xmm12, %xmm0
+; SSSE3-NEXT: pandn %xmm4, %xmm12
+; SSSE3-NEXT: por %xmm0, %xmm12
+; SSSE3-NEXT: pand %xmm11, %xmm1
+; SSSE3-NEXT: pandn %xmm4, %xmm11
+; SSSE3-NEXT: por %xmm1, %xmm11
+; SSSE3-NEXT: movdqa %xmm11, %xmm0
+; SSSE3-NEXT: pxor %xmm8, %xmm0
+; SSSE3-NEXT: movdqa %xmm0, %xmm1
+; SSSE3-NEXT: pcmpgtd %xmm8, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm1[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm8, %xmm0
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm0[1,1,3,3]
+; SSSE3-NEXT: pand %xmm2, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
+; SSSE3-NEXT: por %xmm3, %xmm0
+; SSSE3-NEXT: movdqa %xmm12, %xmm1
+; SSSE3-NEXT: pxor %xmm8, %xmm1
+; SSSE3-NEXT: movdqa %xmm1, %xmm2
+; SSSE3-NEXT: pcmpgtd %xmm8, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm2[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm8, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm1[1,1,3,3]
+; SSSE3-NEXT: pand %xmm3, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm2[1,1,3,3]
+; SSSE3-NEXT: por %xmm4, %xmm1
+; SSSE3-NEXT: movdqa %xmm7, %xmm2
+; SSSE3-NEXT: pxor %xmm8, %xmm2
+; SSSE3-NEXT: movdqa %xmm2, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm8, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm3[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm8, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: pand %xmm4, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSSE3-NEXT: por %xmm2, %xmm3
+; SSSE3-NEXT: movdqa %xmm5, %xmm2
+; SSSE3-NEXT: pxor %xmm8, %xmm2
+; SSSE3-NEXT: movdqa %xmm2, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm8, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm6 = xmm4[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm8, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: pand %xmm6, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSSE3-NEXT: por %xmm2, %xmm4
+; SSSE3-NEXT: pand %xmm5, %xmm4
+; SSSE3-NEXT: pand %xmm7, %xmm3
+; SSSE3-NEXT: pand %xmm12, %xmm1
+; SSSE3-NEXT: pand %xmm11, %xmm0
+; SSSE3-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
+; SSSE3-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,2,2,3,4,5,6,7]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3]
+; SSSE3-NEXT: pshuflw {{.*#+}} xmm1 = xmm1[0,2,2,3,4,5,6,7]
+; SSSE3-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm0 = xmm3[0,2,2,3]
+; SSSE3-NEXT: pshuflw {{.*#+}} xmm2 = xmm0[0,1,0,2,4,5,6,7]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm0 = xmm4[0,2,2,3]
+; SSSE3-NEXT: pshuflw {{.*#+}} xmm0 = xmm0[0,1,0,2,4,5,6,7]
+; SSSE3-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
+; SSSE3-NEXT: movsd {{.*#+}} xmm0 = xmm1[0],xmm0[1]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: trunc_packus_v8i64_v8i8:
+; SSE41: # %bb.0:
+; SSE41-NEXT: movdqa %xmm0, %xmm8
+; SSE41-NEXT: movdqa {{.*#+}} xmm11 = [2147483648,0,2147483648,0]
+; SSE41-NEXT: movdqa %xmm3, %xmm5
+; SSE41-NEXT: pxor %xmm11, %xmm5
+; SSE41-NEXT: movdqa {{.*#+}} xmm0 = [2147483903,2147483903]
+; SSE41-NEXT: movdqa %xmm0, %xmm6
+; SSE41-NEXT: pcmpgtd %xmm5, %xmm6
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm0, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm9 = xmm6[1,1,3,3]
+; SSE41-NEXT: por %xmm5, %xmm9
+; SSE41-NEXT: movdqa %xmm2, %xmm5
+; SSE41-NEXT: pxor %xmm11, %xmm5
+; SSE41-NEXT: movdqa %xmm0, %xmm6
+; SSE41-NEXT: pcmpgtd %xmm5, %xmm6
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm0, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm12 = xmm6[1,1,3,3]
+; SSE41-NEXT: por %xmm5, %xmm12
+; SSE41-NEXT: movdqa %xmm1, %xmm5
+; SSE41-NEXT: pxor %xmm11, %xmm5
+; SSE41-NEXT: movdqa %xmm0, %xmm7
+; SSE41-NEXT: pcmpgtd %xmm5, %xmm7
+; SSE41-NEXT: pshufd {{.*#+}} xmm4 = xmm7[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm0, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3]
+; SSE41-NEXT: pand %xmm4, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm7[1,1,3,3]
+; SSE41-NEXT: por %xmm5, %xmm7
+; SSE41-NEXT: movdqa %xmm8, %xmm4
+; SSE41-NEXT: pxor %xmm11, %xmm4
+; SSE41-NEXT: movdqa %xmm0, %xmm5
+; SSE41-NEXT: pcmpgtd %xmm4, %xmm5
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm5[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm0, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSE41-NEXT: pand %xmm6, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm5[1,1,3,3]
+; SSE41-NEXT: por %xmm4, %xmm0
+; SSE41-NEXT: movapd {{.*#+}} xmm5 = [255,255]
+; SSE41-NEXT: movapd %xmm5, %xmm10
+; SSE41-NEXT: blendvpd %xmm0, %xmm8, %xmm10
+; SSE41-NEXT: movapd %xmm5, %xmm8
+; SSE41-NEXT: movdqa %xmm7, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm1, %xmm8
+; SSE41-NEXT: movapd %xmm5, %xmm13
+; SSE41-NEXT: movdqa %xmm12, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm2, %xmm13
+; SSE41-NEXT: movdqa %xmm9, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm3, %xmm5
+; SSE41-NEXT: movapd %xmm5, %xmm0
+; SSE41-NEXT: xorpd %xmm11, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm2
+; SSE41-NEXT: pcmpgtd %xmm11, %xmm2
+; SSE41-NEXT: pshufd {{.*#+}} xmm3 = xmm2[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm11, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm3, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm2
+; SSE41-NEXT: movapd %xmm13, %xmm0
+; SSE41-NEXT: xorpd %xmm11, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm3
+; SSE41-NEXT: pcmpgtd %xmm11, %xmm3
+; SSE41-NEXT: pshufd {{.*#+}} xmm4 = xmm3[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm11, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm4, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm3
+; SSE41-NEXT: movapd %xmm8, %xmm0
+; SSE41-NEXT: xorpd %xmm11, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm4
+; SSE41-NEXT: pcmpgtd %xmm11, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm4[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm11, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm6, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm4[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm6
+; SSE41-NEXT: movapd %xmm10, %xmm0
+; SSE41-NEXT: xorpd %xmm11, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm4
+; SSE41-NEXT: pcmpgtd %xmm11, %xmm4
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm4[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm11, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm1
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm4[1,1,3,3]
+; SSE41-NEXT: por %xmm1, %xmm0
+; SSE41-NEXT: pand %xmm10, %xmm0
+; SSE41-NEXT: pand %xmm8, %xmm6
+; SSE41-NEXT: pand %xmm13, %xmm3
+; SSE41-NEXT: pand %xmm5, %xmm2
+; SSE41-NEXT: pxor %xmm1, %xmm1
+; SSE41-NEXT: pblendw {{.*#+}} xmm2 = xmm2[0],xmm1[1,2,3],xmm2[4],xmm1[5,6,7]
+; SSE41-NEXT: pblendw {{.*#+}} xmm3 = xmm3[0],xmm1[1,2,3],xmm3[4],xmm1[5,6,7]
+; SSE41-NEXT: packusdw %xmm2, %xmm3
+; SSE41-NEXT: pblendw {{.*#+}} xmm6 = xmm6[0],xmm1[1,2,3],xmm6[4],xmm1[5,6,7]
+; SSE41-NEXT: pblendw {{.*#+}} xmm0 = xmm0[0],xmm1[1,2,3],xmm0[4],xmm1[5,6,7]
+; SSE41-NEXT: packusdw %xmm6, %xmm0
+; SSE41-NEXT: packusdw %xmm3, %xmm0
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_packus_v8i64_v8i8:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm3 = [255,255]
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm3, %xmm2
+; AVX1-NEXT: vpcmpgtq %xmm0, %xmm3, %xmm4
+; AVX1-NEXT: vinsertf128 $1, %xmm2, %ymm4, %ymm2
+; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm4
+; AVX1-NEXT: vpcmpgtq %xmm4, %xmm3, %xmm4
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm3, %xmm3
+; AVX1-NEXT: vinsertf128 $1, %xmm4, %ymm3, %ymm3
+; AVX1-NEXT: vmovapd {{.*#+}} ymm4 = [255,255,255,255]
+; AVX1-NEXT: vblendvpd %ymm3, %ymm1, %ymm4, %ymm1
+; AVX1-NEXT: vblendvpd %ymm2, %ymm0, %ymm4, %ymm0
+; AVX1-NEXT: vxorpd %xmm2, %xmm2, %xmm2
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm0, %xmm8
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm4
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm4, %xmm5
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm1, %xmm6
+; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm7
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm7, %xmm3
+; AVX1-NEXT: vpand %xmm7, %xmm3, %xmm3
+; AVX1-NEXT: vpblendw {{.*#+}} xmm3 = xmm3[0],xmm2[1,2,3],xmm3[4],xmm2[5,6,7]
+; AVX1-NEXT: vpand %xmm1, %xmm6, %xmm1
+; AVX1-NEXT: vpblendw {{.*#+}} xmm1 = xmm1[0],xmm2[1,2,3],xmm1[4],xmm2[5,6,7]
+; AVX1-NEXT: vpackusdw %xmm3, %xmm1, %xmm1
+; AVX1-NEXT: vpand %xmm4, %xmm5, %xmm3
+; AVX1-NEXT: vpblendw {{.*#+}} xmm3 = xmm3[0],xmm2[1,2,3],xmm3[4],xmm2[5,6,7]
+; AVX1-NEXT: vpand %xmm0, %xmm8, %xmm0
+; AVX1-NEXT: vpblendw {{.*#+}} xmm0 = xmm0[0],xmm2[1,2,3],xmm0[4],xmm2[5,6,7]
+; AVX1-NEXT: vpackusdw %xmm3, %xmm0, %xmm0
+; AVX1-NEXT: vpackusdw %xmm1, %xmm0, %xmm0
+; AVX1-NEXT: vzeroupper
+; AVX1-NEXT: retq
+;
+; AVX2-SLOW-LABEL: trunc_packus_v8i64_v8i8:
+; AVX2-SLOW: # %bb.0:
+; AVX2-SLOW-NEXT: vpbroadcastq {{.*#+}} ymm2 = [255,255,255,255]
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm0, %ymm2, %ymm3
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm1, %ymm2, %ymm4
+; AVX2-SLOW-NEXT: vblendvpd %ymm4, %ymm1, %ymm2, %ymm1
+; AVX2-SLOW-NEXT: vblendvpd %ymm3, %ymm0, %ymm2, %ymm0
+; AVX2-SLOW-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm2, %ymm0, %ymm3
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm2, %ymm1, %ymm2
+; AVX2-SLOW-NEXT: vpand %ymm1, %ymm2, %ymm1
+; AVX2-SLOW-NEXT: vpand %ymm0, %ymm3, %ymm0
+; AVX2-SLOW-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[0,2,2,3,4,6,6,7]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-SLOW-NEXT: vpshufd {{.*#+}} ymm1 = ymm1[0,2,2,3,4,6,6,7]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,2,2,3]
+; AVX2-SLOW-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-SLOW-NEXT: vpshufb {{.*#+}} ymm0 = ymm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-SLOW-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX2-SLOW-NEXT: vzeroupper
+; AVX2-SLOW-NEXT: retq
+;
+; AVX2-FAST-LABEL: trunc_packus_v8i64_v8i8:
+; AVX2-FAST: # %bb.0:
+; AVX2-FAST-NEXT: vpbroadcastq {{.*#+}} ymm2 = [255,255,255,255]
+; AVX2-FAST-NEXT: vpcmpgtq %ymm0, %ymm2, %ymm3
+; AVX2-FAST-NEXT: vpcmpgtq %ymm1, %ymm2, %ymm4
+; AVX2-FAST-NEXT: vblendvpd %ymm4, %ymm1, %ymm2, %ymm1
+; AVX2-FAST-NEXT: vblendvpd %ymm3, %ymm0, %ymm2, %ymm0
+; AVX2-FAST-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX2-FAST-NEXT: vpcmpgtq %ymm2, %ymm0, %ymm3
+; AVX2-FAST-NEXT: vpcmpgtq %ymm2, %ymm1, %ymm2
+; AVX2-FAST-NEXT: vpand %ymm1, %ymm2, %ymm1
+; AVX2-FAST-NEXT: vpand %ymm0, %ymm3, %ymm0
+; AVX2-FAST-NEXT: vmovdqa {{.*#+}} ymm2 = [0,2,4,6,4,6,6,7]
+; AVX2-FAST-NEXT: vpermd %ymm0, %ymm2, %ymm0
+; AVX2-FAST-NEXT: vpermd %ymm1, %ymm2, %ymm1
+; AVX2-FAST-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-FAST-NEXT: vpshufb {{.*#+}} ymm0 = ymm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-FAST-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-FAST-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX2-FAST-NEXT: vzeroupper
+; AVX2-FAST-NEXT: retq
+;
+; AVX512-LABEL: trunc_packus_v8i64_v8i8:
+; AVX512: # %bb.0:
+; AVX512-NEXT: vpminsq {{.*}}(%rip){1to8}, %zmm0, %zmm0
+; AVX512-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512-NEXT: vpmaxsq %zmm1, %zmm0, %zmm0
+; AVX512-NEXT: vpmovqw %zmm0, %xmm0
+; AVX512-NEXT: vzeroupper
+; AVX512-NEXT: retq
+ %1 = icmp slt <8 x i64> %a0, <i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255>
+ %2 = select <8 x i1> %1, <8 x i64> %a0, <8 x i64> <i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255>
+ %3 = icmp sgt <8 x i64> %2, zeroinitializer
+ %4 = select <8 x i1> %3, <8 x i64> %2, <8 x i64> zeroinitializer
+ %5 = trunc <8 x i64> %4 to <8 x i8>
+ ret <8 x i8> %5
+}
+
+define <16 x i8> @trunc_packus_v16i64_v16i8(<16 x i64> %a0) {
+; SSE2-LABEL: trunc_packus_v16i64_v16i8:
+; SSE2: # %bb.0:
+; SSE2-NEXT: movdqa %xmm7, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSE2-NEXT: movdqa %xmm6, %xmm10
+; SSE2-NEXT: movdqa %xmm10, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSE2-NEXT: movdqa %xmm5, %xmm11
+; SSE2-NEXT: movdqa %xmm11, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSE2-NEXT: movdqa %xmm4, %xmm6
+; SSE2-NEXT: movdqa %xmm3, %xmm5
+; SSE2-NEXT: movdqa %xmm1, %xmm3
+; SSE2-NEXT: movdqa %xmm0, %xmm8
+; SSE2-NEXT: movdqa {{.*#+}} xmm12 = [2147483648,0,2147483648,0]
+; SSE2-NEXT: movdqa %xmm7, %xmm1
+; SSE2-NEXT: pxor %xmm12, %xmm1
+; SSE2-NEXT: movdqa {{.*#+}} xmm0 = [2147483903,2147483903]
+; SSE2-NEXT: movdqa %xmm0, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm9 = xmm4[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm0, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSE2-NEXT: pand %xmm9, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm9 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm1, %xmm9
+; SSE2-NEXT: movdqa %xmm10, %xmm1
+; SSE2-NEXT: pxor %xmm12, %xmm1
+; SSE2-NEXT: movdqa %xmm0, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm10 = xmm4[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm0, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSE2-NEXT: pand %xmm10, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm10 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm1, %xmm10
+; SSE2-NEXT: movdqa %xmm11, %xmm1
+; SSE2-NEXT: pxor %xmm12, %xmm1
+; SSE2-NEXT: movdqa %xmm0, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm11 = xmm4[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm0, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSE2-NEXT: pand %xmm11, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm13 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm1, %xmm13
+; SSE2-NEXT: movdqa %xmm6, %xmm1
+; SSE2-NEXT: pxor %xmm12, %xmm1
+; SSE2-NEXT: movdqa %xmm0, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm11 = xmm4[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm0, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSE2-NEXT: pand %xmm11, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm11 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm1, %xmm11
+; SSE2-NEXT: movdqa %xmm5, %xmm1
+; SSE2-NEXT: pxor %xmm12, %xmm1
+; SSE2-NEXT: movdqa %xmm0, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm4
+; SSE2-NEXT: pcmpeqd %xmm0, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm4[0,0,2,2]
+; SSE2-NEXT: pand %xmm7, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm14 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm1, %xmm14
+; SSE2-NEXT: movdqa %xmm2, %xmm1
+; SSE2-NEXT: pxor %xmm12, %xmm1
+; SSE2-NEXT: movdqa %xmm0, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm4
+; SSE2-NEXT: pcmpeqd %xmm0, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm4[0,0,2,2]
+; SSE2-NEXT: pand %xmm7, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm15 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm1, %xmm15
+; SSE2-NEXT: movdqa %xmm3, %xmm1
+; SSE2-NEXT: pxor %xmm12, %xmm1
+; SSE2-NEXT: movdqa %xmm0, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm4
+; SSE2-NEXT: pcmpeqd %xmm0, %xmm1
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm1[1,1,3,3]
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm4[0,0,2,2]
+; SSE2-NEXT: pand %xmm1, %xmm7
+; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm4[1,1,3,3]
+; SSE2-NEXT: por %xmm7, %xmm1
+; SSE2-NEXT: movdqa %xmm8, %xmm4
+; SSE2-NEXT: pxor %xmm12, %xmm4
+; SSE2-NEXT: movdqa %xmm0, %xmm7
+; SSE2-NEXT: pcmpgtd %xmm4, %xmm7
+; SSE2-NEXT: pcmpeqd %xmm0, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm7[0,0,2,2]
+; SSE2-NEXT: pand %xmm0, %xmm4
+; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm7[1,1,3,3]
+; SSE2-NEXT: por %xmm4, %xmm0
+; SSE2-NEXT: pand %xmm0, %xmm8
+; SSE2-NEXT: movdqa {{.*#+}} xmm4 = [255,255]
+; SSE2-NEXT: pandn %xmm4, %xmm0
+; SSE2-NEXT: por %xmm8, %xmm0
+; SSE2-NEXT: pand %xmm1, %xmm3
+; SSE2-NEXT: pandn %xmm4, %xmm1
+; SSE2-NEXT: por %xmm3, %xmm1
+; SSE2-NEXT: pand %xmm15, %xmm2
+; SSE2-NEXT: pandn %xmm4, %xmm15
+; SSE2-NEXT: por %xmm2, %xmm15
+; SSE2-NEXT: pand %xmm14, %xmm5
+; SSE2-NEXT: pandn %xmm4, %xmm14
+; SSE2-NEXT: por %xmm5, %xmm14
+; SSE2-NEXT: pand %xmm11, %xmm6
+; SSE2-NEXT: pandn %xmm4, %xmm11
+; SSE2-NEXT: por %xmm6, %xmm11
+; SSE2-NEXT: movdqa -{{[0-9]+}}(%rsp), %xmm2 # 16-byte Reload
+; SSE2-NEXT: pand %xmm13, %xmm2
+; SSE2-NEXT: pandn %xmm4, %xmm13
+; SSE2-NEXT: por %xmm2, %xmm13
+; SSE2-NEXT: movdqa -{{[0-9]+}}(%rsp), %xmm2 # 16-byte Reload
+; SSE2-NEXT: pand %xmm10, %xmm2
+; SSE2-NEXT: pandn %xmm4, %xmm10
+; SSE2-NEXT: por %xmm2, %xmm10
+; SSE2-NEXT: movdqa -{{[0-9]+}}(%rsp), %xmm2 # 16-byte Reload
+; SSE2-NEXT: pand %xmm9, %xmm2
+; SSE2-NEXT: pandn %xmm4, %xmm9
+; SSE2-NEXT: por %xmm2, %xmm9
+; SSE2-NEXT: movdqa %xmm9, %xmm2
+; SSE2-NEXT: pxor %xmm12, %xmm2
+; SSE2-NEXT: movdqa %xmm2, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm12, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm3[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm12, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: pand %xmm5, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm3
+; SSE2-NEXT: movdqa %xmm3, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSE2-NEXT: movdqa %xmm10, %xmm2
+; SSE2-NEXT: pxor %xmm12, %xmm2
+; SSE2-NEXT: movdqa %xmm2, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm12, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm3[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm12, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: pand %xmm5, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm3
+; SSE2-NEXT: movdqa %xmm3, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSE2-NEXT: movdqa %xmm13, %xmm2
+; SSE2-NEXT: pxor %xmm12, %xmm2
+; SSE2-NEXT: movdqa %xmm2, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm12, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm3[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm12, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: pand %xmm5, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm3
+; SSE2-NEXT: movdqa %xmm3, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSE2-NEXT: movdqa %xmm11, %xmm2
+; SSE2-NEXT: pxor %xmm12, %xmm2
+; SSE2-NEXT: movdqa %xmm2, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm12, %xmm3
+; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm3[0,0,2,2]
+; SSE2-NEXT: pcmpeqd %xmm12, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: pand %xmm6, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm3[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm6
+; SSE2-NEXT: movdqa %xmm14, %xmm2
+; SSE2-NEXT: pxor %xmm12, %xmm2
+; SSE2-NEXT: movdqa %xmm2, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm12, %xmm3
+; SSE2-NEXT: pcmpeqd %xmm12, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm3[0,0,2,2]
+; SSE2-NEXT: pand %xmm7, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm7 = xmm3[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm7
+; SSE2-NEXT: movdqa %xmm15, %xmm2
+; SSE2-NEXT: pxor %xmm12, %xmm2
+; SSE2-NEXT: movdqa %xmm2, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm12, %xmm3
+; SSE2-NEXT: pcmpeqd %xmm12, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: pshufd {{.*#+}} xmm8 = xmm3[0,0,2,2]
+; SSE2-NEXT: pand %xmm8, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm8 = xmm3[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm8
+; SSE2-NEXT: movdqa %xmm1, %xmm2
+; SSE2-NEXT: pxor %xmm12, %xmm2
+; SSE2-NEXT: movdqa %xmm2, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm12, %xmm3
+; SSE2-NEXT: pcmpeqd %xmm12, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm3[0,0,2,2]
+; SSE2-NEXT: pand %xmm5, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm3
+; SSE2-NEXT: movdqa %xmm0, %xmm2
+; SSE2-NEXT: pxor %xmm12, %xmm2
+; SSE2-NEXT: movdqa %xmm2, %xmm5
+; SSE2-NEXT: pcmpgtd %xmm12, %xmm5
+; SSE2-NEXT: pcmpeqd %xmm12, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm5[0,0,2,2]
+; SSE2-NEXT: pand %xmm4, %xmm2
+; SSE2-NEXT: pshufd {{.*#+}} xmm12 = xmm5[1,1,3,3]
+; SSE2-NEXT: por %xmm2, %xmm12
+; SSE2-NEXT: movdqa {{.*#+}} xmm2 = [255,255]
+; SSE2-NEXT: pand %xmm2, %xmm9
+; SSE2-NEXT: pand -{{[0-9]+}}(%rsp), %xmm9 # 16-byte Folded Reload
+; SSE2-NEXT: pand %xmm2, %xmm10
+; SSE2-NEXT: pand -{{[0-9]+}}(%rsp), %xmm10 # 16-byte Folded Reload
+; SSE2-NEXT: packuswb %xmm9, %xmm10
+; SSE2-NEXT: pand %xmm2, %xmm13
+; SSE2-NEXT: pand -{{[0-9]+}}(%rsp), %xmm13 # 16-byte Folded Reload
+; SSE2-NEXT: pand %xmm2, %xmm11
+; SSE2-NEXT: pand %xmm6, %xmm11
+; SSE2-NEXT: packuswb %xmm13, %xmm11
+; SSE2-NEXT: packuswb %xmm10, %xmm11
+; SSE2-NEXT: pand %xmm2, %xmm14
+; SSE2-NEXT: pand %xmm7, %xmm14
+; SSE2-NEXT: pand %xmm2, %xmm15
+; SSE2-NEXT: pand %xmm8, %xmm15
+; SSE2-NEXT: packuswb %xmm14, %xmm15
+; SSE2-NEXT: pand %xmm2, %xmm1
+; SSE2-NEXT: pand %xmm3, %xmm1
+; SSE2-NEXT: pand %xmm2, %xmm0
+; SSE2-NEXT: pand %xmm12, %xmm0
+; SSE2-NEXT: packuswb %xmm1, %xmm0
+; SSE2-NEXT: packuswb %xmm15, %xmm0
+; SSE2-NEXT: packuswb %xmm11, %xmm0
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: trunc_packus_v16i64_v16i8:
+; SSSE3: # %bb.0:
+; SSSE3-NEXT: movdqa %xmm7, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSSE3-NEXT: movdqa %xmm6, %xmm10
+; SSSE3-NEXT: movdqa %xmm10, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSSE3-NEXT: movdqa %xmm5, %xmm11
+; SSSE3-NEXT: movdqa %xmm11, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSSE3-NEXT: movdqa %xmm4, %xmm6
+; SSSE3-NEXT: movdqa %xmm3, %xmm5
+; SSSE3-NEXT: movdqa %xmm1, %xmm3
+; SSSE3-NEXT: movdqa %xmm0, %xmm8
+; SSSE3-NEXT: movdqa {{.*#+}} xmm12 = [2147483648,0,2147483648,0]
+; SSSE3-NEXT: movdqa %xmm7, %xmm1
+; SSSE3-NEXT: pxor %xmm12, %xmm1
+; SSSE3-NEXT: movdqa {{.*#+}} xmm0 = [2147483903,2147483903]
+; SSSE3-NEXT: movdqa %xmm0, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm9 = xmm4[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm0, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSSE3-NEXT: pand %xmm9, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm9 = xmm4[1,1,3,3]
+; SSSE3-NEXT: por %xmm1, %xmm9
+; SSSE3-NEXT: movdqa %xmm10, %xmm1
+; SSSE3-NEXT: pxor %xmm12, %xmm1
+; SSSE3-NEXT: movdqa %xmm0, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm10 = xmm4[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm0, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSSE3-NEXT: pand %xmm10, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm10 = xmm4[1,1,3,3]
+; SSSE3-NEXT: por %xmm1, %xmm10
+; SSSE3-NEXT: movdqa %xmm11, %xmm1
+; SSSE3-NEXT: pxor %xmm12, %xmm1
+; SSSE3-NEXT: movdqa %xmm0, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm11 = xmm4[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm0, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSSE3-NEXT: pand %xmm11, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm13 = xmm4[1,1,3,3]
+; SSSE3-NEXT: por %xmm1, %xmm13
+; SSSE3-NEXT: movdqa %xmm6, %xmm1
+; SSSE3-NEXT: pxor %xmm12, %xmm1
+; SSSE3-NEXT: movdqa %xmm0, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm11 = xmm4[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm0, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSSE3-NEXT: pand %xmm11, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm11 = xmm4[1,1,3,3]
+; SSSE3-NEXT: por %xmm1, %xmm11
+; SSSE3-NEXT: movdqa %xmm5, %xmm1
+; SSSE3-NEXT: pxor %xmm12, %xmm1
+; SSSE3-NEXT: movdqa %xmm0, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm4
+; SSSE3-NEXT: pcmpeqd %xmm0, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm4[0,0,2,2]
+; SSSE3-NEXT: pand %xmm7, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm14 = xmm4[1,1,3,3]
+; SSSE3-NEXT: por %xmm1, %xmm14
+; SSSE3-NEXT: movdqa %xmm2, %xmm1
+; SSSE3-NEXT: pxor %xmm12, %xmm1
+; SSSE3-NEXT: movdqa %xmm0, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm4
+; SSSE3-NEXT: pcmpeqd %xmm0, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm4[0,0,2,2]
+; SSSE3-NEXT: pand %xmm7, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm15 = xmm4[1,1,3,3]
+; SSSE3-NEXT: por %xmm1, %xmm15
+; SSSE3-NEXT: movdqa %xmm3, %xmm1
+; SSSE3-NEXT: pxor %xmm12, %xmm1
+; SSSE3-NEXT: movdqa %xmm0, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm4
+; SSSE3-NEXT: pcmpeqd %xmm0, %xmm1
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm1[1,1,3,3]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm4[0,0,2,2]
+; SSSE3-NEXT: pand %xmm1, %xmm7
+; SSSE3-NEXT: pshufd {{.*#+}} xmm1 = xmm4[1,1,3,3]
+; SSSE3-NEXT: por %xmm7, %xmm1
+; SSSE3-NEXT: movdqa %xmm8, %xmm4
+; SSSE3-NEXT: pxor %xmm12, %xmm4
+; SSSE3-NEXT: movdqa %xmm0, %xmm7
+; SSSE3-NEXT: pcmpgtd %xmm4, %xmm7
+; SSSE3-NEXT: pcmpeqd %xmm0, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm0 = xmm7[0,0,2,2]
+; SSSE3-NEXT: pand %xmm0, %xmm4
+; SSSE3-NEXT: pshufd {{.*#+}} xmm0 = xmm7[1,1,3,3]
+; SSSE3-NEXT: por %xmm4, %xmm0
+; SSSE3-NEXT: pand %xmm0, %xmm8
+; SSSE3-NEXT: movdqa {{.*#+}} xmm4 = [255,255]
+; SSSE3-NEXT: pandn %xmm4, %xmm0
+; SSSE3-NEXT: por %xmm8, %xmm0
+; SSSE3-NEXT: pand %xmm1, %xmm3
+; SSSE3-NEXT: pandn %xmm4, %xmm1
+; SSSE3-NEXT: por %xmm3, %xmm1
+; SSSE3-NEXT: pand %xmm15, %xmm2
+; SSSE3-NEXT: pandn %xmm4, %xmm15
+; SSSE3-NEXT: por %xmm2, %xmm15
+; SSSE3-NEXT: pand %xmm14, %xmm5
+; SSSE3-NEXT: pandn %xmm4, %xmm14
+; SSSE3-NEXT: por %xmm5, %xmm14
+; SSSE3-NEXT: pand %xmm11, %xmm6
+; SSSE3-NEXT: pandn %xmm4, %xmm11
+; SSSE3-NEXT: por %xmm6, %xmm11
+; SSSE3-NEXT: movdqa -{{[0-9]+}}(%rsp), %xmm2 # 16-byte Reload
+; SSSE3-NEXT: pand %xmm13, %xmm2
+; SSSE3-NEXT: pandn %xmm4, %xmm13
+; SSSE3-NEXT: por %xmm2, %xmm13
+; SSSE3-NEXT: movdqa -{{[0-9]+}}(%rsp), %xmm2 # 16-byte Reload
+; SSSE3-NEXT: pand %xmm10, %xmm2
+; SSSE3-NEXT: pandn %xmm4, %xmm10
+; SSSE3-NEXT: por %xmm2, %xmm10
+; SSSE3-NEXT: movdqa -{{[0-9]+}}(%rsp), %xmm2 # 16-byte Reload
+; SSSE3-NEXT: pand %xmm9, %xmm2
+; SSSE3-NEXT: pandn %xmm4, %xmm9
+; SSSE3-NEXT: por %xmm2, %xmm9
+; SSSE3-NEXT: movdqa %xmm9, %xmm2
+; SSSE3-NEXT: pxor %xmm12, %xmm2
+; SSSE3-NEXT: movdqa %xmm2, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm12, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm3[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm12, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: pand %xmm5, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSSE3-NEXT: por %xmm2, %xmm3
+; SSSE3-NEXT: movdqa %xmm3, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSSE3-NEXT: movdqa %xmm10, %xmm2
+; SSSE3-NEXT: pxor %xmm12, %xmm2
+; SSSE3-NEXT: movdqa %xmm2, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm12, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm3[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm12, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: pand %xmm5, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSSE3-NEXT: por %xmm2, %xmm3
+; SSSE3-NEXT: movdqa %xmm3, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSSE3-NEXT: movdqa %xmm13, %xmm2
+; SSSE3-NEXT: pxor %xmm12, %xmm2
+; SSSE3-NEXT: movdqa %xmm2, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm12, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm3[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm12, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: pand %xmm5, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSSE3-NEXT: por %xmm2, %xmm3
+; SSSE3-NEXT: movdqa %xmm3, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSSE3-NEXT: movdqa %xmm11, %xmm2
+; SSSE3-NEXT: pxor %xmm12, %xmm2
+; SSSE3-NEXT: movdqa %xmm2, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm12, %xmm3
+; SSSE3-NEXT: pshufd {{.*#+}} xmm6 = xmm3[0,0,2,2]
+; SSSE3-NEXT: pcmpeqd %xmm12, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: pand %xmm6, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm6 = xmm3[1,1,3,3]
+; SSSE3-NEXT: por %xmm2, %xmm6
+; SSSE3-NEXT: movdqa %xmm14, %xmm2
+; SSSE3-NEXT: pxor %xmm12, %xmm2
+; SSSE3-NEXT: movdqa %xmm2, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm12, %xmm3
+; SSSE3-NEXT: pcmpeqd %xmm12, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm3[0,0,2,2]
+; SSSE3-NEXT: pand %xmm7, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm7 = xmm3[1,1,3,3]
+; SSSE3-NEXT: por %xmm2, %xmm7
+; SSSE3-NEXT: movdqa %xmm15, %xmm2
+; SSSE3-NEXT: pxor %xmm12, %xmm2
+; SSSE3-NEXT: movdqa %xmm2, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm12, %xmm3
+; SSSE3-NEXT: pcmpeqd %xmm12, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm8 = xmm3[0,0,2,2]
+; SSSE3-NEXT: pand %xmm8, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm8 = xmm3[1,1,3,3]
+; SSSE3-NEXT: por %xmm2, %xmm8
+; SSSE3-NEXT: movdqa %xmm1, %xmm2
+; SSSE3-NEXT: pxor %xmm12, %xmm2
+; SSSE3-NEXT: movdqa %xmm2, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm12, %xmm3
+; SSSE3-NEXT: pcmpeqd %xmm12, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm5 = xmm3[0,0,2,2]
+; SSSE3-NEXT: pand %xmm5, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3]
+; SSSE3-NEXT: por %xmm2, %xmm3
+; SSSE3-NEXT: movdqa %xmm0, %xmm2
+; SSSE3-NEXT: pxor %xmm12, %xmm2
+; SSSE3-NEXT: movdqa %xmm2, %xmm5
+; SSSE3-NEXT: pcmpgtd %xmm12, %xmm5
+; SSSE3-NEXT: pcmpeqd %xmm12, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3]
+; SSSE3-NEXT: pshufd {{.*#+}} xmm4 = xmm5[0,0,2,2]
+; SSSE3-NEXT: pand %xmm4, %xmm2
+; SSSE3-NEXT: pshufd {{.*#+}} xmm12 = xmm5[1,1,3,3]
+; SSSE3-NEXT: por %xmm2, %xmm12
+; SSSE3-NEXT: movdqa {{.*#+}} xmm2 = [255,255]
+; SSSE3-NEXT: pand %xmm2, %xmm9
+; SSSE3-NEXT: pand -{{[0-9]+}}(%rsp), %xmm9 # 16-byte Folded Reload
+; SSSE3-NEXT: pand %xmm2, %xmm10
+; SSSE3-NEXT: pand -{{[0-9]+}}(%rsp), %xmm10 # 16-byte Folded Reload
+; SSSE3-NEXT: packuswb %xmm9, %xmm10
+; SSSE3-NEXT: pand %xmm2, %xmm13
+; SSSE3-NEXT: pand -{{[0-9]+}}(%rsp), %xmm13 # 16-byte Folded Reload
+; SSSE3-NEXT: pand %xmm2, %xmm11
+; SSSE3-NEXT: pand %xmm6, %xmm11
+; SSSE3-NEXT: packuswb %xmm13, %xmm11
+; SSSE3-NEXT: packuswb %xmm10, %xmm11
+; SSSE3-NEXT: pand %xmm2, %xmm14
+; SSSE3-NEXT: pand %xmm7, %xmm14
+; SSSE3-NEXT: pand %xmm2, %xmm15
+; SSSE3-NEXT: pand %xmm8, %xmm15
+; SSSE3-NEXT: packuswb %xmm14, %xmm15
+; SSSE3-NEXT: pand %xmm2, %xmm1
+; SSSE3-NEXT: pand %xmm3, %xmm1
+; SSSE3-NEXT: pand %xmm2, %xmm0
+; SSSE3-NEXT: pand %xmm12, %xmm0
+; SSSE3-NEXT: packuswb %xmm1, %xmm0
+; SSSE3-NEXT: packuswb %xmm15, %xmm0
+; SSSE3-NEXT: packuswb %xmm11, %xmm0
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: trunc_packus_v16i64_v16i8:
+; SSE41: # %bb.0:
+; SSE41-NEXT: movdqa %xmm7, %xmm8
+; SSE41-NEXT: movdqa %xmm8, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSE41-NEXT: movdqa %xmm6, %xmm7
+; SSE41-NEXT: movdqa %xmm5, %xmm6
+; SSE41-NEXT: movdqa %xmm4, %xmm5
+; SSE41-NEXT: movdqa %xmm3, %xmm4
+; SSE41-NEXT: movdqa %xmm0, %xmm11
+; SSE41-NEXT: movdqa {{.*#+}} xmm13 = [2147483648,0,2147483648,0]
+; SSE41-NEXT: movdqa %xmm8, %xmm0
+; SSE41-NEXT: pxor %xmm13, %xmm0
+; SSE41-NEXT: movdqa {{.*#+}} xmm14 = [2147483903,2147483903]
+; SSE41-NEXT: movdqa %xmm14, %xmm3
+; SSE41-NEXT: pcmpgtd %xmm0, %xmm3
+; SSE41-NEXT: pshufd {{.*#+}} xmm8 = xmm3[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm14, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm8, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm8 = xmm3[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm8
+; SSE41-NEXT: movdqa %xmm7, %xmm0
+; SSE41-NEXT: pxor %xmm13, %xmm0
+; SSE41-NEXT: movdqa %xmm14, %xmm3
+; SSE41-NEXT: pcmpgtd %xmm0, %xmm3
+; SSE41-NEXT: pshufd {{.*#+}} xmm9 = xmm3[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm14, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm9, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm10 = xmm3[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm10
+; SSE41-NEXT: movdqa %xmm6, %xmm0
+; SSE41-NEXT: pxor %xmm13, %xmm0
+; SSE41-NEXT: movdqa %xmm14, %xmm3
+; SSE41-NEXT: pcmpgtd %xmm0, %xmm3
+; SSE41-NEXT: pshufd {{.*#+}} xmm9 = xmm3[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm14, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm9, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm12 = xmm3[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm12
+; SSE41-NEXT: movdqa %xmm11, %xmm0
+; SSE41-NEXT: pxor %xmm13, %xmm0
+; SSE41-NEXT: movdqa %xmm14, %xmm3
+; SSE41-NEXT: pcmpgtd %xmm0, %xmm3
+; SSE41-NEXT: pshufd {{.*#+}} xmm9 = xmm3[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm14, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm15 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm9, %xmm15
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm3[1,1,3,3]
+; SSE41-NEXT: por %xmm15, %xmm0
+; SSE41-NEXT: movapd {{.*#+}} xmm3 = [255,255]
+; SSE41-NEXT: movapd %xmm3, %xmm9
+; SSE41-NEXT: blendvpd %xmm0, %xmm11, %xmm9
+; SSE41-NEXT: movdqa %xmm1, %xmm0
+; SSE41-NEXT: pxor %xmm13, %xmm0
+; SSE41-NEXT: movdqa %xmm14, %xmm11
+; SSE41-NEXT: pcmpgtd %xmm0, %xmm11
+; SSE41-NEXT: pcmpeqd %xmm14, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm15 = xmm0[1,1,3,3]
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm11[0,0,2,2]
+; SSE41-NEXT: pand %xmm0, %xmm15
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm11[1,1,3,3]
+; SSE41-NEXT: por %xmm15, %xmm0
+; SSE41-NEXT: movapd %xmm3, %xmm11
+; SSE41-NEXT: blendvpd %xmm0, %xmm1, %xmm11
+; SSE41-NEXT: movdqa %xmm2, %xmm0
+; SSE41-NEXT: pxor %xmm13, %xmm0
+; SSE41-NEXT: movdqa %xmm14, %xmm1
+; SSE41-NEXT: pcmpgtd %xmm0, %xmm1
+; SSE41-NEXT: pcmpeqd %xmm14, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm15 = xmm0[1,1,3,3]
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm1[0,0,2,2]
+; SSE41-NEXT: pand %xmm0, %xmm15
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,3,3]
+; SSE41-NEXT: por %xmm15, %xmm0
+; SSE41-NEXT: movapd %xmm3, %xmm15
+; SSE41-NEXT: blendvpd %xmm0, %xmm2, %xmm15
+; SSE41-NEXT: movdqa %xmm5, %xmm0
+; SSE41-NEXT: pxor %xmm13, %xmm0
+; SSE41-NEXT: movdqa %xmm14, %xmm1
+; SSE41-NEXT: pcmpgtd %xmm0, %xmm1
+; SSE41-NEXT: pcmpeqd %xmm14, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pshufd {{.*#+}} xmm2 = xmm1[0,0,2,2]
+; SSE41-NEXT: pand %xmm2, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm1
+; SSE41-NEXT: movdqa %xmm4, %xmm0
+; SSE41-NEXT: pxor %xmm13, %xmm0
+; SSE41-NEXT: movdqa %xmm14, %xmm2
+; SSE41-NEXT: pcmpgtd %xmm0, %xmm2
+; SSE41-NEXT: pcmpeqd %xmm14, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm14 = xmm0[1,1,3,3]
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm2[0,0,2,2]
+; SSE41-NEXT: pand %xmm0, %xmm14
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm2[1,1,3,3]
+; SSE41-NEXT: por %xmm14, %xmm0
+; SSE41-NEXT: movapd %xmm3, %xmm14
+; SSE41-NEXT: blendvpd %xmm0, %xmm4, %xmm14
+; SSE41-NEXT: movapd %xmm3, %xmm2
+; SSE41-NEXT: movdqa %xmm1, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm5, %xmm2
+; SSE41-NEXT: movapd %xmm3, %xmm4
+; SSE41-NEXT: movdqa %xmm12, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm6, %xmm4
+; SSE41-NEXT: movapd %xmm3, %xmm1
+; SSE41-NEXT: movdqa %xmm10, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, %xmm7, %xmm1
+; SSE41-NEXT: movapd %xmm3, %xmm5
+; SSE41-NEXT: movdqa %xmm8, %xmm0
+; SSE41-NEXT: blendvpd %xmm0, -{{[0-9]+}}(%rsp), %xmm5 # 16-byte Folded Reload
+; SSE41-NEXT: movapd %xmm5, %xmm0
+; SSE41-NEXT: xorpd %xmm13, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm6
+; SSE41-NEXT: pcmpgtd %xmm13, %xmm6
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm13, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm6[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm6
+; SSE41-NEXT: movdqa %xmm6, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSE41-NEXT: movapd %xmm1, %xmm0
+; SSE41-NEXT: xorpd %xmm13, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm6
+; SSE41-NEXT: pcmpgtd %xmm13, %xmm6
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm13, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm6[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm6
+; SSE41-NEXT: movdqa %xmm6, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSE41-NEXT: movapd %xmm4, %xmm0
+; SSE41-NEXT: xorpd %xmm13, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm6
+; SSE41-NEXT: pcmpgtd %xmm13, %xmm6
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm13, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm6[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm6
+; SSE41-NEXT: movdqa %xmm6, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSE41-NEXT: movapd %xmm2, %xmm0
+; SSE41-NEXT: xorpd %xmm13, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm6
+; SSE41-NEXT: pcmpgtd %xmm13, %xmm6
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE41-NEXT: pcmpeqd %xmm13, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pand %xmm7, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm6[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm6
+; SSE41-NEXT: movdqa %xmm6, -{{[0-9]+}}(%rsp) # 16-byte Spill
+; SSE41-NEXT: movapd %xmm14, %xmm0
+; SSE41-NEXT: xorpd %xmm13, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm6
+; SSE41-NEXT: pcmpgtd %xmm13, %xmm6
+; SSE41-NEXT: pcmpeqd %xmm13, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pshufd {{.*#+}} xmm7 = xmm6[0,0,2,2]
+; SSE41-NEXT: pand %xmm7, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm8 = xmm6[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm8
+; SSE41-NEXT: movapd %xmm15, %xmm0
+; SSE41-NEXT: xorpd %xmm13, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm7
+; SSE41-NEXT: pcmpgtd %xmm13, %xmm7
+; SSE41-NEXT: pcmpeqd %xmm13, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pshufd {{.*#+}} xmm10 = xmm7[0,0,2,2]
+; SSE41-NEXT: pand %xmm10, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm10 = xmm7[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm10
+; SSE41-NEXT: movapd %xmm11, %xmm0
+; SSE41-NEXT: xorpd %xmm13, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm7
+; SSE41-NEXT: pcmpgtd %xmm13, %xmm7
+; SSE41-NEXT: pcmpeqd %xmm13, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pshufd {{.*#+}} xmm12 = xmm7[0,0,2,2]
+; SSE41-NEXT: pand %xmm12, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm12 = xmm7[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm12
+; SSE41-NEXT: movapd %xmm9, %xmm0
+; SSE41-NEXT: xorpd %xmm13, %xmm0
+; SSE41-NEXT: movapd %xmm0, %xmm7
+; SSE41-NEXT: pcmpgtd %xmm13, %xmm7
+; SSE41-NEXT: pcmpeqd %xmm13, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm0 = xmm0[1,1,3,3]
+; SSE41-NEXT: pshufd {{.*#+}} xmm6 = xmm7[0,0,2,2]
+; SSE41-NEXT: pand %xmm6, %xmm0
+; SSE41-NEXT: pshufd {{.*#+}} xmm13 = xmm7[1,1,3,3]
+; SSE41-NEXT: por %xmm0, %xmm13
+; SSE41-NEXT: andpd %xmm3, %xmm5
+; SSE41-NEXT: andpd -{{[0-9]+}}(%rsp), %xmm5 # 16-byte Folded Reload
+; SSE41-NEXT: andpd %xmm3, %xmm1
+; SSE41-NEXT: andpd -{{[0-9]+}}(%rsp), %xmm1 # 16-byte Folded Reload
+; SSE41-NEXT: packuswb %xmm5, %xmm1
+; SSE41-NEXT: andpd %xmm3, %xmm4
+; SSE41-NEXT: andpd -{{[0-9]+}}(%rsp), %xmm4 # 16-byte Folded Reload
+; SSE41-NEXT: andpd %xmm3, %xmm2
+; SSE41-NEXT: andpd -{{[0-9]+}}(%rsp), %xmm2 # 16-byte Folded Reload
+; SSE41-NEXT: packuswb %xmm4, %xmm2
+; SSE41-NEXT: packuswb %xmm1, %xmm2
+; SSE41-NEXT: andpd %xmm3, %xmm14
+; SSE41-NEXT: andpd %xmm8, %xmm14
+; SSE41-NEXT: andpd %xmm3, %xmm15
+; SSE41-NEXT: andpd %xmm10, %xmm15
+; SSE41-NEXT: packuswb %xmm14, %xmm15
+; SSE41-NEXT: andpd %xmm3, %xmm11
+; SSE41-NEXT: andpd %xmm12, %xmm11
+; SSE41-NEXT: andpd %xmm3, %xmm9
+; SSE41-NEXT: andpd %xmm13, %xmm9
+; SSE41-NEXT: packuswb %xmm11, %xmm9
+; SSE41-NEXT: packuswb %xmm15, %xmm9
+; SSE41-NEXT: packuswb %xmm2, %xmm9
+; SSE41-NEXT: movdqa %xmm9, %xmm0
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_packus_v16i64_v16i8:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm5
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm4 = [255,255]
+; AVX1-NEXT: vpcmpgtq %xmm5, %xmm4, %xmm5
+; AVX1-NEXT: vpcmpgtq %xmm0, %xmm4, %xmm6
+; AVX1-NEXT: vinsertf128 $1, %xmm5, %ymm6, %ymm9
+; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm6
+; AVX1-NEXT: vpcmpgtq %xmm6, %xmm4, %xmm6
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm4, %xmm7
+; AVX1-NEXT: vinsertf128 $1, %xmm6, %ymm7, %ymm6
+; AVX1-NEXT: vextractf128 $1, %ymm2, %xmm7
+; AVX1-NEXT: vpcmpgtq %xmm7, %xmm4, %xmm7
+; AVX1-NEXT: vpcmpgtq %xmm2, %xmm4, %xmm8
+; AVX1-NEXT: vinsertf128 $1, %xmm7, %ymm8, %ymm7
+; AVX1-NEXT: vextractf128 $1, %ymm3, %xmm5
+; AVX1-NEXT: vpcmpgtq %xmm5, %xmm4, %xmm5
+; AVX1-NEXT: vpcmpgtq %xmm3, %xmm4, %xmm8
+; AVX1-NEXT: vinsertf128 $1, %xmm5, %ymm8, %ymm5
+; AVX1-NEXT: vmovapd {{.*#+}} ymm8 = [255,255,255,255]
+; AVX1-NEXT: vblendvpd %ymm5, %ymm3, %ymm8, %ymm3
+; AVX1-NEXT: vblendvpd %ymm7, %ymm2, %ymm8, %ymm2
+; AVX1-NEXT: vblendvpd %ymm6, %ymm1, %ymm8, %ymm5
+; AVX1-NEXT: vblendvpd %ymm9, %ymm0, %ymm8, %ymm0
+; AVX1-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX1-NEXT: vextractf128 $1, %ymm3, %xmm6
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm6, %xmm7
+; AVX1-NEXT: vpand %xmm4, %xmm6, %xmm6
+; AVX1-NEXT: vpand %xmm6, %xmm7, %xmm6
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm3, %xmm7
+; AVX1-NEXT: vpand %xmm4, %xmm3, %xmm3
+; AVX1-NEXT: vpand %xmm3, %xmm7, %xmm3
+; AVX1-NEXT: vextractf128 $1, %ymm2, %xmm7
+; AVX1-NEXT: vpackuswb %xmm6, %xmm3, %xmm3
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm7, %xmm6
+; AVX1-NEXT: vpand %xmm4, %xmm7, %xmm7
+; AVX1-NEXT: vpand %xmm7, %xmm6, %xmm6
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm2, %xmm7
+; AVX1-NEXT: vpand %xmm4, %xmm2, %xmm2
+; AVX1-NEXT: vpand %xmm2, %xmm7, %xmm2
+; AVX1-NEXT: vpackuswb %xmm6, %xmm2, %xmm2
+; AVX1-NEXT: vextractf128 $1, %ymm5, %xmm6
+; AVX1-NEXT: vpackuswb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm6, %xmm3
+; AVX1-NEXT: vpand %xmm4, %xmm6, %xmm6
+; AVX1-NEXT: vpand %xmm6, %xmm3, %xmm3
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm5, %xmm6
+; AVX1-NEXT: vpand %xmm4, %xmm5, %xmm5
+; AVX1-NEXT: vpand %xmm5, %xmm6, %xmm5
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm6
+; AVX1-NEXT: vpackuswb %xmm3, %xmm5, %xmm3
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm6, %xmm5
+; AVX1-NEXT: vpand %xmm4, %xmm6, %xmm6
+; AVX1-NEXT: vpand %xmm6, %xmm5, %xmm5
+; AVX1-NEXT: vpcmpgtq %xmm1, %xmm0, %xmm1
+; AVX1-NEXT: vpand %xmm4, %xmm0, %xmm0
+; AVX1-NEXT: vpand %xmm0, %xmm1, %xmm0
+; AVX1-NEXT: vpackuswb %xmm5, %xmm0, %xmm0
+; AVX1-NEXT: vpackuswb %xmm3, %xmm0, %xmm0
+; AVX1-NEXT: vpackuswb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT: vzeroupper
+; AVX1-NEXT: retq
+;
+; AVX2-SLOW-LABEL: trunc_packus_v16i64_v16i8:
+; AVX2-SLOW: # %bb.0:
+; AVX2-SLOW-NEXT: vpbroadcastq {{.*#+}} ymm4 = [255,255,255,255]
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm2, %ymm4, %ymm5
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm3, %ymm4, %ymm6
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm0, %ymm4, %ymm7
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm1, %ymm4, %ymm8
+; AVX2-SLOW-NEXT: vblendvpd %ymm8, %ymm1, %ymm4, %ymm1
+; AVX2-SLOW-NEXT: vblendvpd %ymm7, %ymm0, %ymm4, %ymm0
+; AVX2-SLOW-NEXT: vblendvpd %ymm6, %ymm3, %ymm4, %ymm3
+; AVX2-SLOW-NEXT: vblendvpd %ymm5, %ymm2, %ymm4, %ymm2
+; AVX2-SLOW-NEXT: vpxor %xmm4, %xmm4, %xmm4
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm4, %ymm2, %ymm5
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm4, %ymm3, %ymm6
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm4, %ymm0, %ymm7
+; AVX2-SLOW-NEXT: vpcmpgtq %ymm4, %ymm1, %ymm4
+; AVX2-SLOW-NEXT: vpand %ymm1, %ymm4, %ymm1
+; AVX2-SLOW-NEXT: vpand %ymm0, %ymm7, %ymm0
+; AVX2-SLOW-NEXT: vpand %ymm3, %ymm6, %ymm3
+; AVX2-SLOW-NEXT: vpand %ymm2, %ymm5, %ymm2
+; AVX2-SLOW-NEXT: vpshufd {{.*#+}} ymm2 = ymm2[0,2,2,3,4,6,6,7]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm2 = ymm2[0,2,2,3]
+; AVX2-SLOW-NEXT: vpshufd {{.*#+}} ymm3 = ymm3[0,2,2,3,4,6,6,7]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm3 = ymm3[0,2,2,3]
+; AVX2-SLOW-NEXT: vinserti128 $1, %xmm3, %ymm2, %ymm2
+; AVX2-SLOW-NEXT: vmovdqa {{.*#+}} ymm3 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-SLOW-NEXT: vpshufb %ymm3, %ymm2, %ymm2
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm2 = ymm2[0,2,2,3]
+; AVX2-SLOW-NEXT: vmovdqa {{.*#+}} xmm4 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX2-SLOW-NEXT: vpshufb %xmm4, %xmm2, %xmm2
+; AVX2-SLOW-NEXT: vpshufd {{.*#+}} ymm0 = ymm0[0,2,2,3,4,6,6,7]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-SLOW-NEXT: vpshufd {{.*#+}} ymm1 = ymm1[0,2,2,3,4,6,6,7]
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,2,2,3]
+; AVX2-SLOW-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-SLOW-NEXT: vpshufb %ymm3, %ymm0, %ymm0
+; AVX2-SLOW-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-SLOW-NEXT: vpshufb %xmm4, %xmm0, %xmm0
+; AVX2-SLOW-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; AVX2-SLOW-NEXT: vzeroupper
+; AVX2-SLOW-NEXT: retq
+;
+; AVX2-FAST-LABEL: trunc_packus_v16i64_v16i8:
+; AVX2-FAST: # %bb.0:
+; AVX2-FAST-NEXT: vpbroadcastq {{.*#+}} ymm4 = [255,255,255,255]
+; AVX2-FAST-NEXT: vpcmpgtq %ymm2, %ymm4, %ymm5
+; AVX2-FAST-NEXT: vpcmpgtq %ymm3, %ymm4, %ymm6
+; AVX2-FAST-NEXT: vpcmpgtq %ymm0, %ymm4, %ymm7
+; AVX2-FAST-NEXT: vpcmpgtq %ymm1, %ymm4, %ymm8
+; AVX2-FAST-NEXT: vblendvpd %ymm8, %ymm1, %ymm4, %ymm1
+; AVX2-FAST-NEXT: vblendvpd %ymm7, %ymm0, %ymm4, %ymm0
+; AVX2-FAST-NEXT: vblendvpd %ymm6, %ymm3, %ymm4, %ymm3
+; AVX2-FAST-NEXT: vblendvpd %ymm5, %ymm2, %ymm4, %ymm2
+; AVX2-FAST-NEXT: vpxor %xmm4, %xmm4, %xmm4
+; AVX2-FAST-NEXT: vpcmpgtq %ymm4, %ymm2, %ymm5
+; AVX2-FAST-NEXT: vpcmpgtq %ymm4, %ymm3, %ymm6
+; AVX2-FAST-NEXT: vpcmpgtq %ymm4, %ymm0, %ymm7
+; AVX2-FAST-NEXT: vpcmpgtq %ymm4, %ymm1, %ymm4
+; AVX2-FAST-NEXT: vpand %ymm1, %ymm4, %ymm1
+; AVX2-FAST-NEXT: vpand %ymm0, %ymm7, %ymm0
+; AVX2-FAST-NEXT: vpand %ymm3, %ymm6, %ymm3
+; AVX2-FAST-NEXT: vpand %ymm2, %ymm5, %ymm2
+; AVX2-FAST-NEXT: vmovdqa {{.*#+}} ymm4 = [0,2,4,6,4,6,6,7]
+; AVX2-FAST-NEXT: vpermd %ymm2, %ymm4, %ymm2
+; AVX2-FAST-NEXT: vpermd %ymm3, %ymm4, %ymm3
+; AVX2-FAST-NEXT: vinserti128 $1, %xmm3, %ymm2, %ymm2
+; AVX2-FAST-NEXT: vmovdqa {{.*#+}} ymm3 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-FAST-NEXT: vpshufb %ymm3, %ymm2, %ymm2
+; AVX2-FAST-NEXT: vpermq {{.*#+}} ymm2 = ymm2[0,2,2,3]
+; AVX2-FAST-NEXT: vmovdqa {{.*#+}} xmm5 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX2-FAST-NEXT: vpshufb %xmm5, %xmm2, %xmm2
+; AVX2-FAST-NEXT: vpermd %ymm0, %ymm4, %ymm0
+; AVX2-FAST-NEXT: vpermd %ymm1, %ymm4, %ymm1
+; AVX2-FAST-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-FAST-NEXT: vpshufb %ymm3, %ymm0, %ymm0
+; AVX2-FAST-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-FAST-NEXT: vpshufb %xmm5, %xmm0, %xmm0
+; AVX2-FAST-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; AVX2-FAST-NEXT: vzeroupper
+; AVX2-FAST-NEXT: retq
+;
+; AVX512-LABEL: trunc_packus_v16i64_v16i8:
+; AVX512: # %bb.0:
+; AVX512-NEXT: vpbroadcastq {{.*#+}} zmm2 = [255,255,255,255,255,255,255,255]
+; AVX512-NEXT: vpminsq %zmm2, %zmm0, %zmm0
+; AVX512-NEXT: vpminsq %zmm2, %zmm1, %zmm1
+; AVX512-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX512-NEXT: vpmaxsq %zmm2, %zmm1, %zmm1
+; AVX512-NEXT: vpmaxsq %zmm2, %zmm0, %zmm0
+; AVX512-NEXT: vpmovqd %zmm0, %ymm0
+; AVX512-NEXT: vpmovqd %zmm1, %ymm1
+; AVX512-NEXT: vinserti64x4 $1, %ymm1, %zmm0, %zmm0
+; AVX512-NEXT: vpmovdb %zmm0, %xmm0
+; AVX512-NEXT: vzeroupper
+; AVX512-NEXT: retq
+ %1 = icmp slt <16 x i64> %a0, <i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255>
+ %2 = select <16 x i1> %1, <16 x i64> %a0, <16 x i64> <i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255, i64 255>
+ %3 = icmp sgt <16 x i64> %2, zeroinitializer
+ %4 = select <16 x i1> %3, <16 x i64> %2, <16 x i64> zeroinitializer
+ %5 = trunc <16 x i64> %4 to <16 x i8>
+ ret <16 x i8> %5
+}
+
+define <8 x i8> @trunc_packus_v8i32_v8i8(<8 x i32> %a0) {
+; SSE2-LABEL: trunc_packus_v8i32_v8i8:
+; SSE2: # %bb.0:
+; SSE2-NEXT: movdqa {{.*#+}} xmm2 = [255,255,255,255]
+; SSE2-NEXT: movdqa %xmm2, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm3
+; SSE2-NEXT: movdqa %xmm2, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm0, %xmm4
+; SSE2-NEXT: pand %xmm4, %xmm0
+; SSE2-NEXT: pandn %xmm2, %xmm4
+; SSE2-NEXT: por %xmm0, %xmm4
+; SSE2-NEXT: pand %xmm3, %xmm1
+; SSE2-NEXT: pandn %xmm2, %xmm3
+; SSE2-NEXT: por %xmm1, %xmm3
+; SSE2-NEXT: pxor %xmm1, %xmm1
+; SSE2-NEXT: movdqa %xmm3, %xmm2
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm2
+; SSE2-NEXT: movdqa %xmm4, %xmm0
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm0
+; SSE2-NEXT: pand %xmm4, %xmm0
+; SSE2-NEXT: pand %xmm3, %xmm2
+; SSE2-NEXT: pslld $16, %xmm2
+; SSE2-NEXT: psrad $16, %xmm2
+; SSE2-NEXT: pslld $16, %xmm0
+; SSE2-NEXT: psrad $16, %xmm0
+; SSE2-NEXT: packssdw %xmm2, %xmm0
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: trunc_packus_v8i32_v8i8:
+; SSSE3: # %bb.0:
+; SSSE3-NEXT: movdqa {{.*#+}} xmm2 = [255,255,255,255]
+; SSSE3-NEXT: movdqa %xmm2, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm3
+; SSSE3-NEXT: movdqa %xmm2, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm0, %xmm4
+; SSSE3-NEXT: pand %xmm4, %xmm0
+; SSSE3-NEXT: pandn %xmm2, %xmm4
+; SSSE3-NEXT: por %xmm0, %xmm4
+; SSSE3-NEXT: pand %xmm3, %xmm1
+; SSSE3-NEXT: pandn %xmm2, %xmm3
+; SSSE3-NEXT: por %xmm1, %xmm3
+; SSSE3-NEXT: pxor %xmm1, %xmm1
+; SSSE3-NEXT: movdqa %xmm3, %xmm2
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm2
+; SSSE3-NEXT: movdqa %xmm4, %xmm0
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm0
+; SSSE3-NEXT: pand %xmm4, %xmm0
+; SSSE3-NEXT: pand %xmm3, %xmm2
+; SSSE3-NEXT: movdqa {{.*#+}} xmm1 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
+; SSSE3-NEXT: pshufb %xmm1, %xmm2
+; SSSE3-NEXT: pshufb %xmm1, %xmm0
+; SSSE3-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: trunc_packus_v8i32_v8i8:
+; SSE41: # %bb.0:
+; SSE41-NEXT: movdqa {{.*#+}} xmm2 = [255,255,255,255]
+; SSE41-NEXT: pminsd %xmm2, %xmm1
+; SSE41-NEXT: pminsd %xmm2, %xmm0
+; SSE41-NEXT: pxor %xmm2, %xmm2
+; SSE41-NEXT: pmaxsd %xmm2, %xmm0
+; SSE41-NEXT: pmaxsd %xmm2, %xmm1
+; SSE41-NEXT: movdqa {{.*#+}} xmm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
+; SSE41-NEXT: pshufb %xmm2, %xmm1
+; SSE41-NEXT: pshufb %xmm2, %xmm0
+; SSE41-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_packus_v8i32_v8i8:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm2 = [255,255,255,255]
+; AVX1-NEXT: vpminsd %xmm2, %xmm1, %xmm1
+; AVX1-NEXT: vpminsd %xmm2, %xmm0, %xmm0
+; AVX1-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX1-NEXT: vpmaxsd %xmm2, %xmm0, %xmm0
+; AVX1-NEXT: vpmaxsd %xmm2, %xmm1, %xmm1
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
+; AVX1-NEXT: vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT: vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX1-NEXT: vzeroupper
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: trunc_packus_v8i32_v8i8:
+; AVX2: # %bb.0:
+; AVX2-NEXT: vpbroadcastd {{.*#+}} ymm1 = [255,255,255,255,255,255,255,255]
+; AVX2-NEXT: vpminsd %ymm1, %ymm0, %ymm0
+; AVX2-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX2-NEXT: vpmaxsd %ymm1, %ymm0, %ymm0
+; AVX2-NEXT: vpshufb {{.*#+}} ymm0 = ymm0[0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX2-NEXT: vzeroupper
+; AVX2-NEXT: retq
+;
+; AVX512F-LABEL: trunc_packus_v8i32_v8i8:
+; AVX512F: # %bb.0:
+; AVX512F-NEXT: vpbroadcastd {{.*#+}} ymm1 = [255,255,255,255,255,255,255,255]
+; AVX512F-NEXT: vpminsd %ymm1, %ymm0, %ymm0
+; AVX512F-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512F-NEXT: vpmaxsd %ymm1, %ymm0, %ymm0
+; AVX512F-NEXT: vpmovdw %zmm0, %ymm0
+; AVX512F-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX512F-NEXT: vzeroupper
+; AVX512F-NEXT: retq
+;
+; AVX512VL-LABEL: trunc_packus_v8i32_v8i8:
+; AVX512VL: # %bb.0:
+; AVX512VL-NEXT: vpminsd {{.*}}(%rip){1to8}, %ymm0, %ymm0
+; AVX512VL-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512VL-NEXT: vpmaxsd %ymm1, %ymm0, %ymm0
+; AVX512VL-NEXT: vpmovdw %ymm0, %xmm0
+; AVX512VL-NEXT: vzeroupper
+; AVX512VL-NEXT: retq
+;
+; AVX512BW-LABEL: trunc_packus_v8i32_v8i8:
+; AVX512BW: # %bb.0:
+; AVX512BW-NEXT: vpbroadcastd {{.*#+}} ymm1 = [255,255,255,255,255,255,255,255]
+; AVX512BW-NEXT: vpminsd %ymm1, %ymm0, %ymm0
+; AVX512BW-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512BW-NEXT: vpmaxsd %ymm1, %ymm0, %ymm0
+; AVX512BW-NEXT: vpmovdw %zmm0, %ymm0
+; AVX512BW-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX512BW-NEXT: vzeroupper
+; AVX512BW-NEXT: retq
+;
+; AVX512BWVL-LABEL: trunc_packus_v8i32_v8i8:
+; AVX512BWVL: # %bb.0:
+; AVX512BWVL-NEXT: vpminsd {{.*}}(%rip){1to8}, %ymm0, %ymm0
+; AVX512BWVL-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512BWVL-NEXT: vpmaxsd %ymm1, %ymm0, %ymm0
+; AVX512BWVL-NEXT: vpmovdw %ymm0, %xmm0
+; AVX512BWVL-NEXT: vzeroupper
+; AVX512BWVL-NEXT: retq
+ %1 = icmp slt <8 x i32> %a0, <i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255>
+ %2 = select <8 x i1> %1, <8 x i32> %a0, <8 x i32> <i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255>
+ %3 = icmp sgt <8 x i32> %2, zeroinitializer
+ %4 = select <8 x i1> %3, <8 x i32> %2, <8 x i32> zeroinitializer
+ %5 = trunc <8 x i32> %4 to <8 x i8>
+ ret <8 x i8> %5
+}
+
+define <16 x i8> @trunc_packus_v16i32_v16i8(<16 x i32> %a0) {
+; SSE2-LABEL: trunc_packus_v16i32_v16i8:
+; SSE2: # %bb.0:
+; SSE2-NEXT: movdqa %xmm0, %xmm4
+; SSE2-NEXT: movdqa {{.*#+}} xmm8 = [255,255,255,255]
+; SSE2-NEXT: movdqa %xmm8, %xmm6
+; SSE2-NEXT: pcmpgtd %xmm3, %xmm6
+; SSE2-NEXT: movdqa %xmm8, %xmm5
+; SSE2-NEXT: pcmpgtd %xmm2, %xmm5
+; SSE2-NEXT: movdqa %xmm8, %xmm7
+; SSE2-NEXT: pcmpgtd %xmm1, %xmm7
+; SSE2-NEXT: movdqa %xmm8, %xmm0
+; SSE2-NEXT: pcmpgtd %xmm4, %xmm0
+; SSE2-NEXT: pand %xmm0, %xmm4
+; SSE2-NEXT: pandn %xmm8, %xmm0
+; SSE2-NEXT: por %xmm4, %xmm0
+; SSE2-NEXT: pand %xmm7, %xmm1
+; SSE2-NEXT: pandn %xmm8, %xmm7
+; SSE2-NEXT: por %xmm1, %xmm7
+; SSE2-NEXT: pand %xmm5, %xmm2
+; SSE2-NEXT: pandn %xmm8, %xmm5
+; SSE2-NEXT: por %xmm2, %xmm5
+; SSE2-NEXT: pand %xmm6, %xmm3
+; SSE2-NEXT: pandn %xmm8, %xmm6
+; SSE2-NEXT: por %xmm3, %xmm6
+; SSE2-NEXT: pxor %xmm9, %xmm9
+; SSE2-NEXT: movdqa %xmm6, %xmm2
+; SSE2-NEXT: pcmpgtd %xmm9, %xmm2
+; SSE2-NEXT: movdqa %xmm5, %xmm3
+; SSE2-NEXT: pcmpgtd %xmm9, %xmm3
+; SSE2-NEXT: movdqa %xmm7, %xmm4
+; SSE2-NEXT: pcmpgtd %xmm9, %xmm4
+; SSE2-NEXT: movdqa %xmm0, %xmm1
+; SSE2-NEXT: pcmpgtd %xmm9, %xmm1
+; SSE2-NEXT: pand %xmm8, %xmm6
+; SSE2-NEXT: pand %xmm2, %xmm6
+; SSE2-NEXT: pand %xmm8, %xmm5
+; SSE2-NEXT: pand %xmm3, %xmm5
+; SSE2-NEXT: packuswb %xmm6, %xmm5
+; SSE2-NEXT: pand %xmm8, %xmm7
+; SSE2-NEXT: pand %xmm4, %xmm7
+; SSE2-NEXT: pand %xmm8, %xmm0
+; SSE2-NEXT: pand %xmm1, %xmm0
+; SSE2-NEXT: packuswb %xmm7, %xmm0
+; SSE2-NEXT: packuswb %xmm5, %xmm0
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: trunc_packus_v16i32_v16i8:
+; SSSE3: # %bb.0:
+; SSSE3-NEXT: movdqa %xmm0, %xmm4
+; SSSE3-NEXT: movdqa {{.*#+}} xmm8 = [255,255,255,255]
+; SSSE3-NEXT: movdqa %xmm8, %xmm6
+; SSSE3-NEXT: pcmpgtd %xmm3, %xmm6
+; SSSE3-NEXT: movdqa %xmm8, %xmm5
+; SSSE3-NEXT: pcmpgtd %xmm2, %xmm5
+; SSSE3-NEXT: movdqa %xmm8, %xmm7
+; SSSE3-NEXT: pcmpgtd %xmm1, %xmm7
+; SSSE3-NEXT: movdqa %xmm8, %xmm0
+; SSSE3-NEXT: pcmpgtd %xmm4, %xmm0
+; SSSE3-NEXT: pand %xmm0, %xmm4
+; SSSE3-NEXT: pandn %xmm8, %xmm0
+; SSSE3-NEXT: por %xmm4, %xmm0
+; SSSE3-NEXT: pand %xmm7, %xmm1
+; SSSE3-NEXT: pandn %xmm8, %xmm7
+; SSSE3-NEXT: por %xmm1, %xmm7
+; SSSE3-NEXT: pand %xmm5, %xmm2
+; SSSE3-NEXT: pandn %xmm8, %xmm5
+; SSSE3-NEXT: por %xmm2, %xmm5
+; SSSE3-NEXT: pand %xmm6, %xmm3
+; SSSE3-NEXT: pandn %xmm8, %xmm6
+; SSSE3-NEXT: por %xmm3, %xmm6
+; SSSE3-NEXT: pxor %xmm9, %xmm9
+; SSSE3-NEXT: movdqa %xmm6, %xmm2
+; SSSE3-NEXT: pcmpgtd %xmm9, %xmm2
+; SSSE3-NEXT: movdqa %xmm5, %xmm3
+; SSSE3-NEXT: pcmpgtd %xmm9, %xmm3
+; SSSE3-NEXT: movdqa %xmm7, %xmm4
+; SSSE3-NEXT: pcmpgtd %xmm9, %xmm4
+; SSSE3-NEXT: movdqa %xmm0, %xmm1
+; SSSE3-NEXT: pcmpgtd %xmm9, %xmm1
+; SSSE3-NEXT: pand %xmm8, %xmm6
+; SSSE3-NEXT: pand %xmm2, %xmm6
+; SSSE3-NEXT: pand %xmm8, %xmm5
+; SSSE3-NEXT: pand %xmm3, %xmm5
+; SSSE3-NEXT: packuswb %xmm6, %xmm5
+; SSSE3-NEXT: pand %xmm8, %xmm7
+; SSSE3-NEXT: pand %xmm4, %xmm7
+; SSSE3-NEXT: pand %xmm8, %xmm0
+; SSSE3-NEXT: pand %xmm1, %xmm0
+; SSSE3-NEXT: packuswb %xmm7, %xmm0
+; SSSE3-NEXT: packuswb %xmm5, %xmm0
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: trunc_packus_v16i32_v16i8:
+; SSE41: # %bb.0:
+; SSE41-NEXT: movdqa {{.*#+}} xmm4 = [255,255,255,255]
+; SSE41-NEXT: pminsd %xmm4, %xmm3
+; SSE41-NEXT: pminsd %xmm4, %xmm2
+; SSE41-NEXT: pminsd %xmm4, %xmm1
+; SSE41-NEXT: pminsd %xmm4, %xmm0
+; SSE41-NEXT: pxor %xmm5, %xmm5
+; SSE41-NEXT: pmaxsd %xmm5, %xmm0
+; SSE41-NEXT: pmaxsd %xmm5, %xmm1
+; SSE41-NEXT: pmaxsd %xmm5, %xmm2
+; SSE41-NEXT: pmaxsd %xmm5, %xmm3
+; SSE41-NEXT: pand %xmm4, %xmm3
+; SSE41-NEXT: pand %xmm4, %xmm2
+; SSE41-NEXT: packuswb %xmm3, %xmm2
+; SSE41-NEXT: pand %xmm4, %xmm1
+; SSE41-NEXT: pand %xmm4, %xmm0
+; SSE41-NEXT: packuswb %xmm1, %xmm0
+; SSE41-NEXT: packuswb %xmm2, %xmm0
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_packus_v16i32_v16i8:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm3 = [255,255,255,255]
+; AVX1-NEXT: vpminsd %xmm3, %xmm2, %xmm2
+; AVX1-NEXT: vpminsd %xmm3, %xmm1, %xmm1
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm4
+; AVX1-NEXT: vpminsd %xmm3, %xmm4, %xmm4
+; AVX1-NEXT: vpminsd %xmm3, %xmm0, %xmm0
+; AVX1-NEXT: vpxor %xmm5, %xmm5, %xmm5
+; AVX1-NEXT: vpmaxsd %xmm5, %xmm0, %xmm0
+; AVX1-NEXT: vpmaxsd %xmm5, %xmm4, %xmm4
+; AVX1-NEXT: vpmaxsd %xmm5, %xmm1, %xmm1
+; AVX1-NEXT: vpmaxsd %xmm5, %xmm2, %xmm2
+; AVX1-NEXT: vpand %xmm3, %xmm2, %xmm2
+; AVX1-NEXT: vpand %xmm3, %xmm1, %xmm1
+; AVX1-NEXT: vpackuswb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT: vpand %xmm3, %xmm4, %xmm2
+; AVX1-NEXT: vpand %xmm3, %xmm0, %xmm0
+; AVX1-NEXT: vpackuswb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT: vpackuswb %xmm1, %xmm0, %xmm0
+; AVX1-NEXT: vzeroupper
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: trunc_packus_v16i32_v16i8:
+; AVX2: # %bb.0:
+; AVX2-NEXT: vpbroadcastd {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255]
+; AVX2-NEXT: vpminsd %ymm2, %ymm1, %ymm1
+; AVX2-NEXT: vpminsd %ymm2, %ymm0, %ymm0
+; AVX2-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX2-NEXT: vpmaxsd %ymm2, %ymm0, %ymm0
+; AVX2-NEXT: vpmaxsd %ymm2, %ymm1, %ymm1
+; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-NEXT: vpshufb %ymm2, %ymm1, %ymm1
+; AVX2-NEXT: vpermq {{.*#+}} ymm1 = ymm1[0,2,2,3]
+; AVX2-NEXT: vmovdqa {{.*#+}} xmm3 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX2-NEXT: vpshufb %xmm3, %xmm1, %xmm1
+; AVX2-NEXT: vpshufb %ymm2, %ymm0, %ymm0
+; AVX2-NEXT: vpermq {{.*#+}} ymm0 = ymm0[0,2,2,3]
+; AVX2-NEXT: vpshufb %xmm3, %xmm0, %xmm0
+; AVX2-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX2-NEXT: vzeroupper
+; AVX2-NEXT: retq
+;
+; AVX512-LABEL: trunc_packus_v16i32_v16i8:
+; AVX512: # %bb.0:
+; AVX512-NEXT: vpminsd {{.*}}(%rip){1to16}, %zmm0, %zmm0
+; AVX512-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512-NEXT: vpmaxsd %zmm1, %zmm0, %zmm0
+; AVX512-NEXT: vpmovdb %zmm0, %xmm0
+; AVX512-NEXT: vzeroupper
+; AVX512-NEXT: retq
+ %1 = icmp slt <16 x i32> %a0, <i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255>
+ %2 = select <16 x i1> %1, <16 x i32> %a0, <16 x i32> <i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255, i32 255>
+ %3 = icmp sgt <16 x i32> %2, zeroinitializer
+ %4 = select <16 x i1> %3, <16 x i32> %2, <16 x i32> zeroinitializer
+ %5 = trunc <16 x i32> %4 to <16 x i8>
+ ret <16 x i8> %5
+}
+
+define <16 x i8> @trunc_packus_v16i16_v16i8(<16 x i16> %a0) {
+; SSE2-LABEL: trunc_packus_v16i16_v16i8:
+; SSE2: # %bb.0:
+; SSE2-NEXT: movdqa {{.*#+}} xmm2 = [255,255,255,255,255,255,255,255]
+; SSE2-NEXT: pminsw %xmm2, %xmm1
+; SSE2-NEXT: pminsw %xmm2, %xmm0
+; SSE2-NEXT: pxor %xmm3, %xmm3
+; SSE2-NEXT: pmaxsw %xmm3, %xmm0
+; SSE2-NEXT: pmaxsw %xmm3, %xmm1
+; SSE2-NEXT: pand %xmm2, %xmm1
+; SSE2-NEXT: pand %xmm2, %xmm0
+; SSE2-NEXT: packuswb %xmm1, %xmm0
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: trunc_packus_v16i16_v16i8:
+; SSSE3: # %bb.0:
+; SSSE3-NEXT: movdqa {{.*#+}} xmm2 = [255,255,255,255,255,255,255,255]
+; SSSE3-NEXT: pminsw %xmm2, %xmm1
+; SSSE3-NEXT: pminsw %xmm2, %xmm0
+; SSSE3-NEXT: pxor %xmm2, %xmm2
+; SSSE3-NEXT: pmaxsw %xmm2, %xmm0
+; SSSE3-NEXT: pmaxsw %xmm2, %xmm1
+; SSSE3-NEXT: movdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; SSSE3-NEXT: pshufb %xmm2, %xmm1
+; SSSE3-NEXT: pshufb %xmm2, %xmm0
+; SSSE3-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: trunc_packus_v16i16_v16i8:
+; SSE41: # %bb.0:
+; SSE41-NEXT: movdqa {{.*#+}} xmm2 = [255,255,255,255,255,255,255,255]
+; SSE41-NEXT: pminsw %xmm2, %xmm1
+; SSE41-NEXT: pminsw %xmm2, %xmm0
+; SSE41-NEXT: pxor %xmm2, %xmm2
+; SSE41-NEXT: pmaxsw %xmm2, %xmm0
+; SSE41-NEXT: pmaxsw %xmm2, %xmm1
+; SSE41-NEXT: movdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; SSE41-NEXT: pshufb %xmm2, %xmm1
+; SSE41-NEXT: pshufb %xmm2, %xmm0
+; SSE41-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_packus_v16i16_v16i8:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm1
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm2 = [255,255,255,255,255,255,255,255]
+; AVX1-NEXT: vpminsw %xmm2, %xmm1, %xmm1
+; AVX1-NEXT: vpminsw %xmm2, %xmm0, %xmm0
+; AVX1-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX1-NEXT: vpmaxsw %xmm2, %xmm0, %xmm0
+; AVX1-NEXT: vpmaxsw %xmm2, %xmm1, %xmm1
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX1-NEXT: vpshufb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT: vpshufb %xmm2, %xmm0, %xmm0
+; AVX1-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX1-NEXT: vzeroupper
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: trunc_packus_v16i16_v16i8:
+; AVX2: # %bb.0:
+; AVX2-NEXT: vpminsw {{.*}}(%rip), %ymm0, %ymm0
+; AVX2-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX2-NEXT: vpmaxsw %ymm1, %ymm0, %ymm0
+; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1
+; AVX2-NEXT: vmovdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX2-NEXT: vpshufb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT: vpshufb %xmm2, %xmm0, %xmm0
+; AVX2-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; AVX2-NEXT: vzeroupper
+; AVX2-NEXT: retq
+;
+; AVX512F-LABEL: trunc_packus_v16i16_v16i8:
+; AVX512F: # %bb.0:
+; AVX512F-NEXT: vpminsw {{.*}}(%rip), %ymm0, %ymm0
+; AVX512F-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512F-NEXT: vpmaxsw %ymm1, %ymm0, %ymm0
+; AVX512F-NEXT: vpmovsxwd %ymm0, %zmm0
+; AVX512F-NEXT: vpmovdb %zmm0, %xmm0
+; AVX512F-NEXT: vzeroupper
+; AVX512F-NEXT: retq
+;
+; AVX512VL-LABEL: trunc_packus_v16i16_v16i8:
+; AVX512VL: # %bb.0:
+; AVX512VL-NEXT: vpminsw {{.*}}(%rip), %ymm0, %ymm0
+; AVX512VL-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512VL-NEXT: vpmaxsw %ymm1, %ymm0, %ymm0
+; AVX512VL-NEXT: vpmovsxwd %ymm0, %zmm0
+; AVX512VL-NEXT: vpmovdb %zmm0, %xmm0
+; AVX512VL-NEXT: vzeroupper
+; AVX512VL-NEXT: retq
+;
+; AVX512BW-LABEL: trunc_packus_v16i16_v16i8:
+; AVX512BW: # %bb.0:
+; AVX512BW-NEXT: vpminsw {{.*}}(%rip), %ymm0, %ymm0
+; AVX512BW-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512BW-NEXT: vpmaxsw %ymm1, %ymm0, %ymm0
+; AVX512BW-NEXT: vpmovwb %zmm0, %ymm0
+; AVX512BW-NEXT: # kill: def %xmm0 killed %xmm0 killed %ymm0
+; AVX512BW-NEXT: vzeroupper
+; AVX512BW-NEXT: retq
+;
+; AVX512BWVL-LABEL: trunc_packus_v16i16_v16i8:
+; AVX512BWVL: # %bb.0:
+; AVX512BWVL-NEXT: vpminsw {{.*}}(%rip), %ymm0, %ymm0
+; AVX512BWVL-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512BWVL-NEXT: vpmaxsw %ymm1, %ymm0, %ymm0
+; AVX512BWVL-NEXT: vpmovwb %ymm0, %xmm0
+; AVX512BWVL-NEXT: vzeroupper
+; AVX512BWVL-NEXT: retq
+ %1 = icmp slt <16 x i16> %a0, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+ %2 = select <16 x i1> %1, <16 x i16> %a0, <16 x i16> <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+ %3 = icmp sgt <16 x i16> %2, zeroinitializer
+ %4 = select <16 x i1> %3, <16 x i16> %2, <16 x i16> zeroinitializer
+ %5 = trunc <16 x i16> %4 to <16 x i8>
+ ret <16 x i8> %5
+}
+
+define <32 x i8> @trunc_packus_v32i16_v32i8(<32 x i16> %a0) {
+; SSE2-LABEL: trunc_packus_v32i16_v32i8:
+; SSE2: # %bb.0:
+; SSE2-NEXT: movdqa {{.*#+}} xmm4 = [255,255,255,255,255,255,255,255]
+; SSE2-NEXT: pminsw %xmm4, %xmm1
+; SSE2-NEXT: pminsw %xmm4, %xmm0
+; SSE2-NEXT: pminsw %xmm4, %xmm3
+; SSE2-NEXT: pminsw %xmm4, %xmm2
+; SSE2-NEXT: pxor %xmm5, %xmm5
+; SSE2-NEXT: pmaxsw %xmm5, %xmm2
+; SSE2-NEXT: pmaxsw %xmm5, %xmm3
+; SSE2-NEXT: pmaxsw %xmm5, %xmm0
+; SSE2-NEXT: pmaxsw %xmm5, %xmm1
+; SSE2-NEXT: pand %xmm4, %xmm1
+; SSE2-NEXT: pand %xmm4, %xmm0
+; SSE2-NEXT: packuswb %xmm1, %xmm0
+; SSE2-NEXT: pand %xmm4, %xmm3
+; SSE2-NEXT: pand %xmm2, %xmm4
+; SSE2-NEXT: packuswb %xmm3, %xmm4
+; SSE2-NEXT: movdqa %xmm4, %xmm1
+; SSE2-NEXT: retq
+;
+; SSSE3-LABEL: trunc_packus_v32i16_v32i8:
+; SSSE3: # %bb.0:
+; SSSE3-NEXT: movdqa {{.*#+}} xmm4 = [255,255,255,255,255,255,255,255]
+; SSSE3-NEXT: pminsw %xmm4, %xmm1
+; SSSE3-NEXT: pminsw %xmm4, %xmm0
+; SSSE3-NEXT: pminsw %xmm4, %xmm3
+; SSSE3-NEXT: pminsw %xmm2, %xmm4
+; SSSE3-NEXT: pxor %xmm2, %xmm2
+; SSSE3-NEXT: pmaxsw %xmm2, %xmm4
+; SSSE3-NEXT: pmaxsw %xmm2, %xmm3
+; SSSE3-NEXT: pmaxsw %xmm2, %xmm0
+; SSSE3-NEXT: pmaxsw %xmm2, %xmm1
+; SSSE3-NEXT: movdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; SSSE3-NEXT: pshufb %xmm2, %xmm1
+; SSSE3-NEXT: pshufb %xmm2, %xmm0
+; SSSE3-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSSE3-NEXT: pshufb %xmm2, %xmm3
+; SSSE3-NEXT: pshufb %xmm2, %xmm4
+; SSSE3-NEXT: punpcklqdq {{.*#+}} xmm4 = xmm4[0],xmm3[0]
+; SSSE3-NEXT: movdqa %xmm4, %xmm1
+; SSSE3-NEXT: retq
+;
+; SSE41-LABEL: trunc_packus_v32i16_v32i8:
+; SSE41: # %bb.0:
+; SSE41-NEXT: movdqa {{.*#+}} xmm4 = [255,255,255,255,255,255,255,255]
+; SSE41-NEXT: pminsw %xmm4, %xmm1
+; SSE41-NEXT: pminsw %xmm4, %xmm0
+; SSE41-NEXT: pminsw %xmm4, %xmm3
+; SSE41-NEXT: pminsw %xmm2, %xmm4
+; SSE41-NEXT: pxor %xmm2, %xmm2
+; SSE41-NEXT: pmaxsw %xmm2, %xmm4
+; SSE41-NEXT: pmaxsw %xmm2, %xmm3
+; SSE41-NEXT: pmaxsw %xmm2, %xmm0
+; SSE41-NEXT: pmaxsw %xmm2, %xmm1
+; SSE41-NEXT: movdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; SSE41-NEXT: pshufb %xmm2, %xmm1
+; SSE41-NEXT: pshufb %xmm2, %xmm0
+; SSE41-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
+; SSE41-NEXT: pshufb %xmm2, %xmm3
+; SSE41-NEXT: pshufb %xmm2, %xmm4
+; SSE41-NEXT: punpcklqdq {{.*#+}} xmm4 = xmm4[0],xmm3[0]
+; SSE41-NEXT: movdqa %xmm4, %xmm1
+; SSE41-NEXT: retq
+;
+; AVX1-LABEL: trunc_packus_v32i16_v32i8:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm3 = [255,255,255,255,255,255,255,255]
+; AVX1-NEXT: vpminsw %xmm3, %xmm2, %xmm2
+; AVX1-NEXT: vpminsw %xmm3, %xmm1, %xmm1
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm4
+; AVX1-NEXT: vpminsw %xmm3, %xmm4, %xmm4
+; AVX1-NEXT: vpminsw %xmm3, %xmm0, %xmm0
+; AVX1-NEXT: vpxor %xmm3, %xmm3, %xmm3
+; AVX1-NEXT: vpmaxsw %xmm3, %xmm0, %xmm0
+; AVX1-NEXT: vpmaxsw %xmm3, %xmm4, %xmm4
+; AVX1-NEXT: vpmaxsw %xmm3, %xmm1, %xmm1
+; AVX1-NEXT: vpmaxsw %xmm3, %xmm2, %xmm2
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm3 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX1-NEXT: vpshufb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT: vpshufb %xmm3, %xmm1, %xmm1
+; AVX1-NEXT: vpunpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
+; AVX1-NEXT: vpshufb %xmm3, %xmm4, %xmm2
+; AVX1-NEXT: vpshufb %xmm3, %xmm0, %xmm0
+; AVX1-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: trunc_packus_v32i16_v32i8:
+; AVX2: # %bb.0:
+; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; AVX2-NEXT: vpminsw %ymm2, %ymm1, %ymm1
+; AVX2-NEXT: vpminsw %ymm2, %ymm0, %ymm0
+; AVX2-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX2-NEXT: vpmaxsw %ymm2, %ymm0, %ymm0
+; AVX2-NEXT: vpmaxsw %ymm2, %ymm1, %ymm1
+; AVX2-NEXT: vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT: vmovdqa {{.*#+}} xmm3 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX2-NEXT: vpshufb %xmm3, %xmm2, %xmm2
+; AVX2-NEXT: vpshufb %xmm3, %xmm1, %xmm1
+; AVX2-NEXT: vpunpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm2[0]
+; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm2
+; AVX2-NEXT: vpshufb %xmm3, %xmm2, %xmm2
+; AVX2-NEXT: vpshufb %xmm3, %xmm0, %xmm0
+; AVX2-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
+; AVX2-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT: retq
+;
+; AVX512F-LABEL: trunc_packus_v32i16_v32i8:
+; AVX512F: # %bb.0:
+; AVX512F-NEXT: vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; AVX512F-NEXT: vpminsw %ymm2, %ymm0, %ymm0
+; AVX512F-NEXT: vpminsw %ymm2, %ymm1, %ymm1
+; AVX512F-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX512F-NEXT: vpmaxsw %ymm2, %ymm1, %ymm1
+; AVX512F-NEXT: vpmaxsw %ymm2, %ymm0, %ymm0
+; AVX512F-NEXT: vpmovsxwd %ymm0, %zmm0
+; AVX512F-NEXT: vpmovdb %zmm0, %xmm0
+; AVX512F-NEXT: vpmovsxwd %ymm1, %zmm1
+; AVX512F-NEXT: vpmovdb %zmm1, %xmm1
+; AVX512F-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX512F-NEXT: retq
+;
+; AVX512VL-LABEL: trunc_packus_v32i16_v32i8:
+; AVX512VL: # %bb.0:
+; AVX512VL-NEXT: vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; AVX512VL-NEXT: vpminsw %ymm2, %ymm0, %ymm0
+; AVX512VL-NEXT: vpminsw %ymm2, %ymm1, %ymm1
+; AVX512VL-NEXT: vpxor %xmm2, %xmm2, %xmm2
+; AVX512VL-NEXT: vpmaxsw %ymm2, %ymm1, %ymm1
+; AVX512VL-NEXT: vpmaxsw %ymm2, %ymm0, %ymm0
+; AVX512VL-NEXT: vpmovsxwd %ymm0, %zmm0
+; AVX512VL-NEXT: vpmovdb %zmm0, %xmm0
+; AVX512VL-NEXT: vpmovsxwd %ymm1, %zmm1
+; AVX512VL-NEXT: vpmovdb %zmm1, %xmm1
+; AVX512VL-NEXT: vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX512VL-NEXT: retq
+;
+; AVX512BW-LABEL: trunc_packus_v32i16_v32i8:
+; AVX512BW: # %bb.0:
+; AVX512BW-NEXT: vpminsw {{.*}}(%rip), %zmm0, %zmm0
+; AVX512BW-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512BW-NEXT: vpmaxsw %zmm1, %zmm0, %zmm0
+; AVX512BW-NEXT: vpmovwb %zmm0, %ymm0
+; AVX512BW-NEXT: retq
+;
+; AVX512BWVL-LABEL: trunc_packus_v32i16_v32i8:
+; AVX512BWVL: # %bb.0:
+; AVX512BWVL-NEXT: vpminsw {{.*}}(%rip), %zmm0, %zmm0
+; AVX512BWVL-NEXT: vpxor %xmm1, %xmm1, %xmm1
+; AVX512BWVL-NEXT: vpmaxsw %zmm1, %zmm0, %zmm0
+; AVX512BWVL-NEXT: vpmovwb %zmm0, %ymm0
+; AVX512BWVL-NEXT: retq
+ %1 = icmp slt <32 x i16> %a0, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+ %2 = select <32 x i1> %1, <32 x i16> %a0, <32 x i16> <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+ %3 = icmp sgt <32 x i16> %2, zeroinitializer
+ %4 = select <32 x i1> %3, <32 x i16> %2, <32 x i16> zeroinitializer
+ %5 = trunc <32 x i16> %4 to <32 x i8>
+ ret <32 x i8> %5
+}
More information about the llvm-commits
mailing list