[X86][SSE] Add MULHS/MULHU custom lowering for i8 vectors

[RKSimon]

Summary

Currently this is to mainly to prevent scalarization of integer division by constants.

[qcolombet]

lib/Target/X86/X86ISelLowering.cpp

19053 ↗

(On Diff #51137)

Right now, Subtarget.hasInt256() should be true for 256 bit vector, since we add the custom lowering for those only when this predicate is true.
I.e., this is dead code at the moment, right?

19078 ↗

(On Diff #51137)

This returns the low value, right?
I believe we miss a shift before the truncate.

19094 ↗

(On Diff #51137)

A few comments on how we achieve the extraction wouldn’t hurt :).

E.g., Fill the 8 high bits of a v8i16 vector with the low input vector.
Shift that vector right by 8, to put the 8 relevant bits at the right place and fill the 8 high bits with 0 or sign extension.

[RKSimon]

Revised based on Quentin's feedback.

Enabled v32i8 and v16i16 MULHS/MULHU custom lowering on AVX1 targets to prevent scalarization - I can commit the v16i16 support separately if necessary.

Improved the AVX 32i8 lowerings to make use ymm PACKUS (saves 2cy on my Carrizo tests) and added a missing shift to the v16i8 version (I had stupidly only locally tested 32i8 on AVX2 hardware).

Tried to improve comments on what is going on.

[qcolombet]

Hi Simon,

LGTM. Couple of nits inlined.

Cheers,
-Quentin

lib/Target/X86/X86ISelLowering.cpp

19094 ↗

(On Diff #51350)

Add that unlike the smaller PACKUS, the ymm variant interleaves the 128 bits of the both sources.
Without that in mind, the shuffle does not make sense, whereas it is definitely required :).

19117 ↗

(On Diff #51350)

Period.

19133 ↗

(On Diff #51350)

Period.

llvm/trunk/lib/Target/X86/X86ISelLowering.cpp

if (!Subtarget.useSoftFloat() && Subtarget.hasSSE2()) {
     setOperationAction(ISD::ADD,                MVT::v8i16, Legal);
     setOperationAction(ISD::ADD,                MVT::v4i32, Legal);
     setOperationAction(ISD::ADD,                MVT::v2i64, Legal);
     setOperationAction(ISD::MUL,                MVT::v16i8, Custom);
     setOperationAction(ISD::MUL,                MVT::v4i32, Custom);
     setOperationAction(ISD::MUL,                MVT::v2i64, Custom);
     setOperationAction(ISD::UMUL_LOHI,          MVT::v4i32, Custom);
     setOperationAction(ISD::SMUL_LOHI,          MVT::v4i32, Custom);
+    setOperationAction(ISD::MULHU,              MVT::v16i8, Custom);
+    setOperationAction(ISD::MULHS,              MVT::v16i8, Custom);
     setOperationAction(ISD::MULHU,              MVT::v8i16, Legal);
     setOperationAction(ISD::MULHS,              MVT::v8i16, Legal);
     setOperationAction(ISD::SUB,                MVT::v16i8, Legal);
     setOperationAction(ISD::SUB,                MVT::v8i16, Legal);
     setOperationAction(ISD::SUB,                MVT::v4i32, Legal);
     setOperationAction(ISD::SUB,                MVT::v2i64, Legal);
     setOperationAction(ISD::MUL,                MVT::v8i16, Legal);
     setOperationAction(ISD::FADD,               MVT::v2f64, Legal);
if (Subtarget.hasInt256()) {
 
       setOperationAction(ISD::MUL,             MVT::v4i64, Custom);
       setOperationAction(ISD::MUL,             MVT::v8i32, Legal);
       setOperationAction(ISD::MUL,             MVT::v16i16, Legal);
       setOperationAction(ISD::MUL,             MVT::v32i8, Custom);
 
       setOperationAction(ISD::UMUL_LOHI,       MVT::v8i32, Custom);
       setOperationAction(ISD::SMUL_LOHI,       MVT::v8i32, Custom);
+
       setOperationAction(ISD::MULHU,           MVT::v16i16, Legal);
       setOperationAction(ISD::MULHS,           MVT::v16i16, Legal);
+      setOperationAction(ISD::MULHU,           MVT::v32i8, Custom);
+      setOperationAction(ISD::MULHS,           MVT::v32i8, Custom);
 
       setOperationAction(ISD::SMAX,            MVT::v32i8,  Legal);
       setOperationAction(ISD::SMAX,            MVT::v16i16, Legal);
       setOperationAction(ISD::SMAX,            MVT::v8i32,  Legal);
       setOperationAction(ISD::UMAX,            MVT::v32i8,  Legal);
       setOperationAction(ISD::UMAX,            MVT::v16i16, Legal);
       setOperationAction(ISD::UMAX,            MVT::v8i32,  Legal);
       setOperationAction(ISD::SMIN,            MVT::v32i8,  Legal);
if (Subtarget.hasInt256()) {
       setOperationAction(ISD::SUB,             MVT::v16i16, Custom);
       setOperationAction(ISD::SUB,             MVT::v32i8, Custom);
 
       setOperationAction(ISD::MUL,             MVT::v4i64, Custom);
       setOperationAction(ISD::MUL,             MVT::v8i32, Custom);
       setOperationAction(ISD::MUL,             MVT::v16i16, Custom);
       setOperationAction(ISD::MUL,             MVT::v32i8, Custom);
 
+      setOperationAction(ISD::MULHU,           MVT::v32i8, Custom);
+      setOperationAction(ISD::MULHS,           MVT::v32i8, Custom);
+
       setOperationAction(ISD::SMAX,            MVT::v32i8,  Custom);
       setOperationAction(ISD::SMAX,            MVT::v16i16, Custom);
       setOperationAction(ISD::SMAX,            MVT::v8i32,  Custom);
       setOperationAction(ISD::UMAX,            MVT::v32i8,  Custom);
       setOperationAction(ISD::UMAX,            MVT::v16i16, Custom);
       setOperationAction(ISD::UMAX,            MVT::v8i32,  Custom);
       setOperationAction(ISD::SMIN,            MVT::v32i8,  Custom);
       setOperationAction(ISD::SMIN,            MVT::v16i16, Custom);
if (!ISD::isBuildVectorAllZeros(Bhi.getNode())) {
     AloBhi = DAG.getNode(X86ISD::PMULUDQ, dl, VT, A, Bhi);
     AloBhi = getTargetVShiftByConstNode(X86ISD::VSHLI, dl, VT, AloBhi, 32, DAG);
   }
 
   SDValue Res = DAG.getNode(ISD::ADD, dl, VT, AloBlo, AloBhi);
   return DAG.getNode(ISD::ADD, dl, VT, Res, AhiBlo);
 }
 
+static SDValue LowerMULH(SDValue Op, const X86Subtarget &Subtarget,
+                         SelectionDAG &DAG) {
+  SDLoc dl(Op);
+  MVT VT = Op.getSimpleValueType();
+
+  // Decompose 256-bit ops into smaller 128-bit ops.
+  if (VT.is256BitVector() && !Subtarget.hasInt256())
+    return Lower256IntArith(Op, DAG);
+
+  // Only i8 vectors should need custom lowering after this.
+  assert((VT == MVT::v16i8 || (VT == MVT::v32i8 && Subtarget.hasInt256())) &&
+         "Unsupported vector type");
+
+  // Lower v16i8/v32i8 as extension to v8i16/v16i16 vector pairs, multiply,
+  // logical shift down the upper half and pack back to i8.
+  SDValue A = Op.getOperand(0);
+  SDValue B = Op.getOperand(1);
+
+  // With SSE41 we can use sign/zero extend, but for pre-SSE41 we unpack
+  // and then ashr/lshr the upper bits down to the lower bits before multiply.
+  unsigned Opcode = Op.getOpcode();
+  unsigned ExShift = (ISD::MULHU == Opcode ? ISD::SRL : ISD::SRA);
+  unsigned ExSSE41 = (ISD::MULHU == Opcode ? X86ISD::VZEXT : X86ISD::VSEXT);
+
+  // AVX2 implementations - extend xmm subvectors to ymm.
+  if (Subtarget.hasInt256()) {
+    SDValue Lo = DAG.getIntPtrConstant(0, dl);
+    SDValue Hi = DAG.getIntPtrConstant(VT.getVectorNumElements() / 2, dl);
+
+    if (VT == MVT::v32i8) {
+      SDValue ALo = DAG.getNode(ISD::EXTRACT_SUBVECTOR, dl, MVT::v16i8, A, Lo);
+      SDValue BLo = DAG.getNode(ISD::EXTRACT_SUBVECTOR, dl, MVT::v16i8, B, Lo);
+      SDValue AHi = DAG.getNode(ISD::EXTRACT_SUBVECTOR, dl, MVT::v16i8, A, Hi);
+      SDValue BHi = DAG.getNode(ISD::EXTRACT_SUBVECTOR, dl, MVT::v16i8, B, Hi);
+      ALo = DAG.getNode(ExSSE41, dl, MVT::v16i16, ALo);
+      BLo = DAG.getNode(ExSSE41, dl, MVT::v16i16, BLo);
+      AHi = DAG.getNode(ExSSE41, dl, MVT::v16i16, AHi);
+      BHi = DAG.getNode(ExSSE41, dl, MVT::v16i16, BHi);
+      Lo = DAG.getNode(ISD::SRL, dl, MVT::v16i16,
+                       DAG.getNode(ISD::MUL, dl, MVT::v16i16, ALo, BLo),
+                       DAG.getConstant(8, dl, MVT::v16i16));
+      Hi = DAG.getNode(ISD::SRL, dl, MVT::v16i16,
+                       DAG.getNode(ISD::MUL, dl, MVT::v16i16, AHi, BHi),
+                       DAG.getConstant(8, dl, MVT::v16i16));
+      // The ymm variant of PACKUS treats the 128-bit lanes separately, so before
+      // using PACKUS we need to permute the inputs to the correct lo/hi xmm lane.
+      const int LoMask[] = {0,  1,  2,  3,  4,  5,  6,  7,
+                            16, 17, 18, 19, 20, 21, 22, 23};
+      const int HiMask[] = {8,  9,  10, 11, 12, 13, 14, 15,
+                            24, 25, 26, 27, 28, 29, 30, 31};
+      return DAG.getNode(X86ISD::PACKUS, dl, VT,
+                         DAG.getVectorShuffle(MVT::v16i16, dl, Lo, Hi, LoMask),
+                         DAG.getVectorShuffle(MVT::v16i16, dl, Lo, Hi, HiMask));
+    }
+
+    SDValue ExA = DAG.getNode(ExSSE41, dl, MVT::v16i16, A);
+    SDValue ExB = DAG.getNode(ExSSE41, dl, MVT::v16i16, B);
+    SDValue Mul = DAG.getNode(ISD::MUL, dl, MVT::v16i16, ExA, ExB);
+    SDValue MulH = DAG.getNode(ISD::SRL, dl, MVT::v16i16, Mul,
+                               DAG.getConstant(8, dl, MVT::v16i16));
+    Lo = DAG.getNode(ISD::EXTRACT_SUBVECTOR, dl, MVT::v8i16, MulH, Lo);
+    Hi = DAG.getNode(ISD::EXTRACT_SUBVECTOR, dl, MVT::v8i16, MulH, Hi);
+    return DAG.getNode(X86ISD::PACKUS, dl, VT, Lo, Hi);
+  }
+
+  assert(VT == MVT::v16i8 &&
+         "Pre-AVX2 support only supports v16i8 multiplication");
+  MVT ExVT = MVT::v8i16;
+
+  // Extract the lo parts and zero/sign extend to i16.
+  SDValue ALo, BLo;
+  if (Subtarget.hasSSE41()) {
+    ALo = DAG.getNode(ExSSE41, dl, ExVT, A);
+    BLo = DAG.getNode(ExSSE41, dl, ExVT, B);
+  } else {
+    const int ShufMask[] = {-1, 0, -1, 1, -1, 2, -1, 3,
+                            -1, 4, -1, 5, -1, 6, -1, 7};
+    ALo = DAG.getVectorShuffle(VT, dl, A, A, ShufMask);
+    BLo = DAG.getVectorShuffle(VT, dl, B, B, ShufMask);
+    ALo = DAG.getBitcast(ExVT, ALo);
+    BLo = DAG.getBitcast(ExVT, BLo);
+    ALo = DAG.getNode(ExShift, dl, ExVT, ALo, DAG.getConstant(8, dl, ExVT));
+    BLo = DAG.getNode(ExShift, dl, ExVT, BLo, DAG.getConstant(8, dl, ExVT));
+  }
+
+  // Extract the hi parts and zero/sign extend to i16.
+  SDValue AHi, BHi;
+  if (Subtarget.hasSSE41()) {
+    const int ShufMask[] = {8,  9,  10, 11, 12, 13, 14, 15,
+                            -1, -1, -1, -1, -1, -1, -1, -1};
+    AHi = DAG.getVectorShuffle(VT, dl, A, A, ShufMask);
+    BHi = DAG.getVectorShuffle(VT, dl, B, B, ShufMask);
+    AHi = DAG.getNode(ExSSE41, dl, ExVT, AHi);
+    BHi = DAG.getNode(ExSSE41, dl, ExVT, BHi);
+  } else {
+    const int ShufMask[] = {-1, 8,  -1, 9,  -1, 10, -1, 11,
+                            -1, 12, -1, 13, -1, 14, -1, 15};
+    AHi = DAG.getVectorShuffle(VT, dl, A, A, ShufMask);
+    BHi = DAG.getVectorShuffle(VT, dl, B, B, ShufMask);
+    AHi = DAG.getBitcast(ExVT, AHi);
+    BHi = DAG.getBitcast(ExVT, BHi);
+    AHi = DAG.getNode(ExShift, dl, ExVT, AHi, DAG.getConstant(8, dl, ExVT));
+    BHi = DAG.getNode(ExShift, dl, ExVT, BHi, DAG.getConstant(8, dl, ExVT));
+  }
+
+  // Multiply, lshr the upper 8bits to the lower 8bits of the lo/hi results and
+  // pack back to v16i8.
+  SDValue RLo = DAG.getNode(ISD::MUL, dl, ExVT, ALo, BLo);
+  SDValue RHi = DAG.getNode(ISD::MUL, dl, ExVT, AHi, BHi);
+  RLo = DAG.getNode(ISD::SRL, dl, ExVT, RLo, DAG.getConstant(8, dl, ExVT));
+  RHi = DAG.getNode(ISD::SRL, dl, ExVT, RHi, DAG.getConstant(8, dl, ExVT));
+  return DAG.getNode(X86ISD::PACKUS, dl, VT, RLo, RHi);
+}
+
 SDValue X86TargetLowering::LowerWin64_i128OP(SDValue Op, SelectionDAG &DAG) const {
   assert(Subtarget.isTargetWin64() && "Unexpected target");
   EVT VT = Op.getValueType();
   assert(VT.isInteger() && VT.getSizeInBits() == 128 &&
          "Unexpected return type for lowering");
 
   RTLIB::Libcall LC;
   bool isSigned;
SDValue X86TargetLowering::LowerOperation(SDValue Op, SelectionDAG &DAG) const {
   case ISD::INIT_TRAMPOLINE:    return LowerINIT_TRAMPOLINE(Op, DAG);
   case ISD::ADJUST_TRAMPOLINE:  return LowerADJUST_TRAMPOLINE(Op, DAG);
   case ISD::FLT_ROUNDS_:        return LowerFLT_ROUNDS_(Op, DAG);
   case ISD::CTLZ:               return LowerCTLZ(Op, Subtarget, DAG);
   case ISD::CTLZ_ZERO_UNDEF:    return LowerCTLZ_ZERO_UNDEF(Op, Subtarget, DAG);
   case ISD::CTTZ:
   case ISD::CTTZ_ZERO_UNDEF:    return LowerCTTZ(Op, DAG);
   case ISD::MUL:                return LowerMUL(Op, Subtarget, DAG);
+  case ISD::MULHS:
+  case ISD::MULHU:              return LowerMULH(Op, Subtarget, DAG);
   case ISD::UMUL_LOHI:
   case ISD::SMUL_LOHI:          return LowerMUL_LOHI(Op, Subtarget, DAG);
   case ISD::ROTL:               return LowerRotate(Op, Subtarget, DAG);
   case ISD::SRA:
   case ISD::SRL:
   case ISD::SHL:                return LowerShift(Op, Subtarget, DAG);
   case ISD::SADDO:
   case ISD::UADDO:

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

 ; AVX-NEXT:    retq
   %res = sdiv <8 x i16> %a, <i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7>
   ret <8 x i16> %res
 }
 
 define <16 x i8> @test_div7_16i8(<16 x i8> %a) nounwind {
 ; SSE2-LABEL: test_div7_16i8:
 ; SSE2:       # BB#0:
-; SSE2-NEXT:    pushq %rbp
-; SSE2-NEXT:    pushq %r14
-; SSE2-NEXT:    pushq %rbx
-; SSE2-NEXT:    movaps %xmm0, -{{[0-9]+}}(%rsp)
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %eax
-; SSE2-NEXT:    imull $-109, %eax, %ecx
-; SSE2-NEXT:    shrl $8, %ecx
-; SSE2-NEXT:    addb %al, %cl
-; SSE2-NEXT:    movb %cl, %al
-; SSE2-NEXT:    shrb $7, %al
-; SSE2-NEXT:    sarb $2, %cl
-; SSE2-NEXT:    addb %al, %cl
-; SSE2-NEXT:    movzbl %cl, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %r14d
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %edx
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %r9d
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %eax
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %r11d
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %ecx
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %r8d
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %esi
-; SSE2-NEXT:    imull $-109, %esi, %edi
-; SSE2-NEXT:    shrl $8, %edi
-; SSE2-NEXT:    addb %sil, %dil
-; SSE2-NEXT:    movb %dil, %bl
-; SSE2-NEXT:    shrb $7, %bl
-; SSE2-NEXT:    sarb $2, %dil
-; SSE2-NEXT:    addb %bl, %dil
-; SSE2-NEXT:    movzbl %dil, %esi
-; SSE2-NEXT:    movd %esi, %xmm1
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
-; SSE2-NEXT:    imull $-109, %eax, %esi
-; SSE2-NEXT:    shrl $8, %esi
-; SSE2-NEXT:    addb %al, %sil
-; SSE2-NEXT:    movb %sil, %al
-; SSE2-NEXT:    shrb $7, %al
-; SSE2-NEXT:    sarb $2, %sil
-; SSE2-NEXT:    addb %al, %sil
-; SSE2-NEXT:    movzbl %sil, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %ebp
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %esi
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %r10d
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %edi
-; SSE2-NEXT:    imull $-109, %edi, %ebx
-; SSE2-NEXT:    shrl $8, %ebx
-; SSE2-NEXT:    addb %dil, %bl
-; SSE2-NEXT:    movb %bl, %al
-; SSE2-NEXT:    shrb $7, %al
-; SSE2-NEXT:    sarb $2, %bl
-; SSE2-NEXT:    addb %al, %bl
-; SSE2-NEXT:    movzbl %bl, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; SSE2-NEXT:    imull $-109, %edx, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm1
-; SSE2-NEXT:    imull $-109, %esi, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %sil, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1],xmm2[2],xmm1[2],xmm2[3],xmm1[3],xmm2[4],xmm1[4],xmm2[5],xmm1[5],xmm2[6],xmm1[6],xmm2[7],xmm1[7]
-; SSE2-NEXT:    imull $-109, %ecx, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    movb %al, %cl
-; SSE2-NEXT:    shrb $7, %cl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm3
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %ecx
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %eax
-; SSE2-NEXT:    imull $-109, %eax, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    addb %al, %dl
-; SSE2-NEXT:    movb %dl, %al
-; SSE2-NEXT:    shrb $7, %al
-; SSE2-NEXT:    sarb $2, %dl
-; SSE2-NEXT:    addb %al, %dl
-; SSE2-NEXT:    movzbl %dl, %eax
-; SSE2-NEXT:    movd %eax, %xmm1
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1],xmm1[2],xmm3[2],xmm1[3],xmm3[3],xmm1[4],xmm3[4],xmm1[5],xmm3[5],xmm1[6],xmm3[6],xmm1[7],xmm3[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3],xmm1[4],xmm2[4],xmm1[5],xmm2[5],xmm1[6],xmm2[6],xmm1[7],xmm2[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
-; SSE2-NEXT:    imull $-109, %r14d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %r14b, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    imull $-109, %ebp, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %bpl, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-NEXT:    imull $-109, %r11d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %r11b, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm3
-; SSE2-NEXT:    imull $-109, %ecx, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    movb %al, %cl
-; SSE2-NEXT:    shrb $7, %cl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1],xmm2[2],xmm3[2],xmm2[3],xmm3[3],xmm2[4],xmm3[4],xmm2[5],xmm3[5],xmm2[6],xmm3[6],xmm2[7],xmm3[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3],xmm2[4],xmm0[4],xmm2[5],xmm0[5],xmm2[6],xmm0[6],xmm2[7],xmm0[7]
-; SSE2-NEXT:    imull $-109, %r9d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %r9b, %al
-; SSE2-NEXT:    movb %al, %cl
-; SSE2-NEXT:    shrb $7, %cl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    imull $-109, %r10d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %r10b, %al
-; SSE2-NEXT:    movb %al, %cl
-; SSE2-NEXT:    shrb $7, %cl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm3
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm3 = xmm3[0],xmm0[0],xmm3[1],xmm0[1],xmm3[2],xmm0[2],xmm3[3],xmm0[3],xmm3[4],xmm0[4],xmm3[5],xmm0[5],xmm3[6],xmm0[6],xmm3[7],xmm0[7]
-; SSE2-NEXT:    imull $-109, %r8d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %r8b, %al
-; SSE2-NEXT:    movb %al, %cl
-; SSE2-NEXT:    shrb $7, %cl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm4
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %eax
-; SSE2-NEXT:    imull $-109, %eax, %ecx
-; SSE2-NEXT:    shrl $8, %ecx
-; SSE2-NEXT:    addb %al, %cl
-; SSE2-NEXT:    movb %cl, %al
-; SSE2-NEXT:    shrb $7, %al
-; SSE2-NEXT:    sarb $2, %cl
-; SSE2-NEXT:    addb %al, %cl
-; SSE2-NEXT:    movzbl %cl, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm4[0],xmm0[1],xmm4[1],xmm0[2],xmm4[2],xmm0[3],xmm4[3],xmm0[4],xmm4[4],xmm0[5],xmm4[5],xmm0[6],xmm4[6],xmm0[7],xmm4[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1],xmm0[2],xmm3[2],xmm0[3],xmm3[3],xmm0[4],xmm3[4],xmm0[5],xmm3[5],xmm0[6],xmm3[6],xmm0[7],xmm3[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; SSE2-NEXT:    popq %rbx
-; SSE2-NEXT:    popq %r14
-; SSE2-NEXT:    popq %rbp
+; SSE2-NEXT:    movdqa {{.*#+}} xmm2 = [147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147]
+; SSE2-NEXT:    psraw $8, %xmm2
+; SSE2-NEXT:    movdqa %xmm0, %xmm3
+; SSE2-NEXT:    punpckhbw {{.*#+}} xmm3 = xmm3[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; SSE2-NEXT:    psraw $8, %xmm3
+; SSE2-NEXT:    pmullw %xmm2, %xmm3
+; SSE2-NEXT:    psrlw $8, %xmm3
+; SSE2-NEXT:    movdqa %xmm0, %xmm1
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    psraw $8, %xmm1
+; SSE2-NEXT:    pmullw %xmm2, %xmm1
+; SSE2-NEXT:    psrlw $8, %xmm1
+; SSE2-NEXT:    packuswb %xmm3, %xmm1
+; SSE2-NEXT:    paddb %xmm0, %xmm1
+; SSE2-NEXT:    movdqa %xmm1, %xmm0
+; SSE2-NEXT:    psrlw $2, %xmm0
+; SSE2-NEXT:    pand {{.*}}(%rip), %xmm0
+; SSE2-NEXT:    movdqa {{.*#+}} xmm2 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; SSE2-NEXT:    pxor %xmm2, %xmm0
+; SSE2-NEXT:    psubb %xmm2, %xmm0
+; SSE2-NEXT:    psrlw $7, %xmm1
+; SSE2-NEXT:    pand {{.*}}(%rip), %xmm1
+; SSE2-NEXT:    paddb %xmm0, %xmm1
+; SSE2-NEXT:    movdqa %xmm1, %xmm0
 ; SSE2-NEXT:    retq
 ;
 ; SSE41-LABEL: test_div7_16i8:
 ; SSE41:       # BB#0:
-; SSE41-NEXT:    pextrb $1, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pextrb $0, %xmm0, %ecx
-; SSE41-NEXT:    movsbl %cl, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %edx
-; SSE41-NEXT:    shrl $8, %edx
-; SSE41-NEXT:    addb %dl, %cl
-; SSE41-NEXT:    movb %cl, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %cl
-; SSE41-NEXT:    addb %dl, %cl
-; SSE41-NEXT:    movzbl %cl, %ecx
-; SSE41-NEXT:    movd %ecx, %xmm1
-; SSE41-NEXT:    pinsrb $1, %eax, %xmm1
-; SSE41-NEXT:    pextrb $2, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $2, %eax, %xmm1
-; SSE41-NEXT:    pextrb $3, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $3, %eax, %xmm1
-; SSE41-NEXT:    pextrb $4, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $4, %eax, %xmm1
-; SSE41-NEXT:    pextrb $5, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $5, %eax, %xmm1
-; SSE41-NEXT:    pextrb $6, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $6, %eax, %xmm1
-; SSE41-NEXT:    pextrb $7, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $7, %eax, %xmm1
-; SSE41-NEXT:    pextrb $8, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $8, %eax, %xmm1
-; SSE41-NEXT:    pextrb $9, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $9, %eax, %xmm1
-; SSE41-NEXT:    pextrb $10, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $10, %eax, %xmm1
-; SSE41-NEXT:    pextrb $11, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $11, %eax, %xmm1
-; SSE41-NEXT:    pextrb $12, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $12, %eax, %xmm1
-; SSE41-NEXT:    pextrb $13, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $13, %eax, %xmm1
-; SSE41-NEXT:    pextrb $14, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $14, %eax, %xmm1
-; SSE41-NEXT:    pextrb $15, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %eax
-; SSE41-NEXT:    imull $-109, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $15, %eax, %xmm1
+; SSE41-NEXT:    pmovsxbw %xmm0, %xmm1
+; SSE41-NEXT:    pmovsxbw {{.*}}(%rip), %xmm2
+; SSE41-NEXT:    pmullw %xmm2, %xmm1
+; SSE41-NEXT:    psrlw $8, %xmm1
+; SSE41-NEXT:    pshufd {{.*#+}} xmm3 = xmm0[2,3,0,1]
+; SSE41-NEXT:    pmovsxbw %xmm3, %xmm3
+; SSE41-NEXT:    pmullw %xmm2, %xmm3
+; SSE41-NEXT:    psrlw $8, %xmm3
+; SSE41-NEXT:    packuswb %xmm3, %xmm1
+; SSE41-NEXT:    paddb %xmm0, %xmm1
+; SSE41-NEXT:    movdqa %xmm1, %xmm0
+; SSE41-NEXT:    psrlw $2, %xmm0
+; SSE41-NEXT:    pand {{.*}}(%rip), %xmm0
+; SSE41-NEXT:    movdqa {{.*#+}} xmm2 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; SSE41-NEXT:    pxor %xmm2, %xmm0
+; SSE41-NEXT:    psubb %xmm2, %xmm0
+; SSE41-NEXT:    psrlw $7, %xmm1
+; SSE41-NEXT:    pand {{.*}}(%rip), %xmm1
+; SSE41-NEXT:    paddb %xmm0, %xmm1
 ; SSE41-NEXT:    movdqa %xmm1, %xmm0
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: test_div7_16i8:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpextrb $1, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpextrb $0, %xmm0, %ecx
-; AVX-NEXT:    movsbl %cl, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %edx
-; AVX-NEXT:    shrl $8, %edx
-; AVX-NEXT:    addb %dl, %cl
-; AVX-NEXT:    movb %cl, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %cl
-; AVX-NEXT:    addb %dl, %cl
-; AVX-NEXT:    movzbl %cl, %ecx
-; AVX-NEXT:    vmovd %ecx, %xmm1
-; AVX-NEXT:    vpinsrb $1, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $2, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $3, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $3, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $5, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $5, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $6, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $7, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $7, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $8, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $9, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $9, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $10, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $10, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $11, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $11, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $12, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $13, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $13, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $14, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $14, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $15, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %eax
-; AVX-NEXT:    imull $-109, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $15, %eax, %xmm1, %xmm0
-; AVX-NEXT:    retq
+; AVX1-LABEL: test_div7_16i8:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpmovsxbw %xmm0, %xmm1
+; AVX1-NEXT:    vpmovsxbw {{.*}}(%rip), %xmm2
+; AVX1-NEXT:    vpmullw %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpsrlw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[2,3,0,1]
+; AVX1-NEXT:    vpmovsxbw %xmm3, %xmm3
+; AVX1-NEXT:    vpmullw %xmm2, %xmm3, %xmm2
+; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
+; AVX1-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpaddb %xmm0, %xmm1, %xmm0
+; AVX1-NEXT:    vpsrlw $2, %xmm0, %xmm1
+; AVX1-NEXT:    vpand {{.*}}(%rip), %xmm1, %xmm1
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm2 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; AVX1-NEXT:    vpxor %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpsubb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpsrlw $7, %xmm0, %xmm0
+; AVX1-NEXT:    vpand {{.*}}(%rip), %xmm0, %xmm0
+; AVX1-NEXT:    vpaddb %xmm0, %xmm1, %xmm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: test_div7_16i8:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpmovsxbw %xmm0, %ymm1
+; AVX2-NEXT:    vpmovsxbw {{.*}}(%rip), %ymm2
+; AVX2-NEXT:    vpmullw %ymm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpaddb %xmm0, %xmm1, %xmm0
+; AVX2-NEXT:    vpsrlw $2, %xmm0, %xmm1
+; AVX2-NEXT:    vpand {{.*}}(%rip), %xmm1, %xmm1
+; AVX2-NEXT:    vmovdqa {{.*#+}} xmm2 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; AVX2-NEXT:    vpxor %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpsubb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpsrlw $7, %xmm0, %xmm0
+; AVX2-NEXT:    vpand {{.*}}(%rip), %xmm0, %xmm0
+; AVX2-NEXT:    vpaddb %xmm0, %xmm1, %xmm0
+; AVX2-NEXT:    vzeroupper
+; AVX2-NEXT:    retq
   %res = sdiv <16 x i8> %a, <i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7>
   ret <16 x i8> %res
 }
 
 ;
 ; srem by 7
 ;
 

 ; AVX-NEXT:    retq
   %res = srem <8 x i16> %a, <i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7>
   ret <8 x i16> %res
 }
 
 define <16 x i8> @test_rem7_16i8(<16 x i8> %a) nounwind {
 ; SSE2-LABEL: test_rem7_16i8:
 ; SSE2:       # BB#0:
-; SSE2-NEXT:    pushq %rbp
-; SSE2-NEXT:    pushq %r15
-; SSE2-NEXT:    pushq %r14
-; SSE2-NEXT:    pushq %rbx
-; SSE2-NEXT:    movaps %xmm0, -{{[0-9]+}}(%rsp)
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %ecx
-; SSE2-NEXT:    imull $-109, %ecx, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    movb $7, %r11b
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %cl
-; SSE2-NEXT:    movzbl %cl, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %r15d
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %ebx
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %r9d
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %edi
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %r14d
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %esi
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %r8d
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %ebp
-; SSE2-NEXT:    imull $-109, %ebp, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %bpl, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %bpl
-; SSE2-NEXT:    movzbl %bpl, %eax
-; SSE2-NEXT:    movd %eax, %xmm1
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
-; SSE2-NEXT:    imull $-109, %edi, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %dil, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %dil
-; SSE2-NEXT:    movzbl %dil, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %edi
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %ebp
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %r10d
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %edx
-; SSE2-NEXT:    imull $-109, %edx, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    movb %al, %cl
-; SSE2-NEXT:    shrb $7, %cl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %dl
-; SSE2-NEXT:    movzbl %dl, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; SSE2-NEXT:    imull $-109, %ebx, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %bl, %al
-; SSE2-NEXT:    movb %al, %cl
-; SSE2-NEXT:    shrb $7, %cl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %bl
-; SSE2-NEXT:    movzbl %bl, %eax
-; SSE2-NEXT:    movd %eax, %xmm1
-; SSE2-NEXT:    imull $-109, %ebp, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %bpl, %al
-; SSE2-NEXT:    movb %al, %cl
-; SSE2-NEXT:    shrb $7, %cl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %bpl
-; SSE2-NEXT:    movzbl %bpl, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1],xmm2[2],xmm1[2],xmm2[3],xmm1[3],xmm2[4],xmm1[4],xmm2[5],xmm1[5],xmm2[6],xmm1[6],xmm2[7],xmm1[7]
-; SSE2-NEXT:    imull $-109, %esi, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %sil, %al
-; SSE2-NEXT:    movb %al, %cl
-; SSE2-NEXT:    shrb $7, %cl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %sil
-; SSE2-NEXT:    movzbl %sil, %eax
-; SSE2-NEXT:    movd %eax, %xmm3
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %ecx
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %edx
-; SSE2-NEXT:    imull $-109, %edx, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    movb %al, %bl
-; SSE2-NEXT:    shrb $7, %bl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %bl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %dl
-; SSE2-NEXT:    movzbl %dl, %eax
-; SSE2-NEXT:    movd %eax, %xmm1
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1],xmm1[2],xmm3[2],xmm1[3],xmm3[3],xmm1[4],xmm3[4],xmm1[5],xmm3[5],xmm1[6],xmm3[6],xmm1[7],xmm3[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3],xmm1[4],xmm2[4],xmm1[5],xmm2[5],xmm1[6],xmm2[6],xmm1[7],xmm2[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
-; SSE2-NEXT:    imull $-109, %r15d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %r15b, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %r15b
-; SSE2-NEXT:    movzbl %r15b, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    imull $-109, %edi, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %dil, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %dil
-; SSE2-NEXT:    movzbl %dil, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-NEXT:    imull $-109, %r14d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %r14b, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %r14b
-; SSE2-NEXT:    movzbl %r14b, %eax
-; SSE2-NEXT:    movd %eax, %xmm3
-; SSE2-NEXT:    imull $-109, %ecx, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %cl
-; SSE2-NEXT:    movzbl %cl, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1],xmm2[2],xmm3[2],xmm2[3],xmm3[3],xmm2[4],xmm3[4],xmm2[5],xmm3[5],xmm2[6],xmm3[6],xmm2[7],xmm3[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3],xmm2[4],xmm0[4],xmm2[5],xmm0[5],xmm2[6],xmm0[6],xmm2[7],xmm0[7]
-; SSE2-NEXT:    imull $-109, %r9d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %r9b, %al
-; SSE2-NEXT:    movb %al, %cl
-; SSE2-NEXT:    shrb $7, %cl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %r9b
-; SSE2-NEXT:    movzbl %r9b, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    imull $-109, %r10d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %r10b, %al
-; SSE2-NEXT:    movb %al, %cl
-; SSE2-NEXT:    shrb $7, %cl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %r10b
-; SSE2-NEXT:    movzbl %r10b, %eax
-; SSE2-NEXT:    movd %eax, %xmm3
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm3 = xmm3[0],xmm0[0],xmm3[1],xmm0[1],xmm3[2],xmm0[2],xmm3[3],xmm0[3],xmm3[4],xmm0[4],xmm3[5],xmm0[5],xmm3[6],xmm0[6],xmm3[7],xmm0[7]
-; SSE2-NEXT:    imull $-109, %r8d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %r8b, %al
-; SSE2-NEXT:    movb %al, %cl
-; SSE2-NEXT:    shrb $7, %cl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %r8b
-; SSE2-NEXT:    movzbl %r8b, %eax
-; SSE2-NEXT:    movd %eax, %xmm4
-; SSE2-NEXT:    movsbl -{{[0-9]+}}(%rsp), %ecx
-; SSE2-NEXT:    imull $-109, %ecx, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    movb %al, %dl
-; SSE2-NEXT:    shrb $7, %dl
-; SSE2-NEXT:    sarb $2, %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %cl
-; SSE2-NEXT:    movzbl %cl, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm4[0],xmm0[1],xmm4[1],xmm0[2],xmm4[2],xmm0[3],xmm4[3],xmm0[4],xmm4[4],xmm0[5],xmm4[5],xmm0[6],xmm4[6],xmm0[7],xmm4[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1],xmm0[2],xmm3[2],xmm0[3],xmm3[3],xmm0[4],xmm3[4],xmm0[5],xmm3[5],xmm0[6],xmm3[6],xmm0[7],xmm3[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; SSE2-NEXT:    popq %rbx
-; SSE2-NEXT:    popq %r14
-; SSE2-NEXT:    popq %r15
-; SSE2-NEXT:    popq %rbp
+; SSE2-NEXT:    movdqa {{.*#+}} xmm2 = [147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147]
+; SSE2-NEXT:    psraw $8, %xmm2
+; SSE2-NEXT:    movdqa %xmm0, %xmm3
+; SSE2-NEXT:    punpckhbw {{.*#+}} xmm3 = xmm3[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; SSE2-NEXT:    psraw $8, %xmm3
+; SSE2-NEXT:    pmullw %xmm2, %xmm3
+; SSE2-NEXT:    psrlw $8, %xmm3
+; SSE2-NEXT:    movdqa %xmm0, %xmm1
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    psraw $8, %xmm1
+; SSE2-NEXT:    pmullw %xmm2, %xmm1
+; SSE2-NEXT:    psrlw $8, %xmm1
+; SSE2-NEXT:    packuswb %xmm3, %xmm1
+; SSE2-NEXT:    paddb %xmm0, %xmm1
+; SSE2-NEXT:    movdqa %xmm1, %xmm2
+; SSE2-NEXT:    psrlw $2, %xmm2
+; SSE2-NEXT:    pand {{.*}}(%rip), %xmm2
+; SSE2-NEXT:    movdqa {{.*#+}} xmm3 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; SSE2-NEXT:    pxor %xmm3, %xmm2
+; SSE2-NEXT:    psubb %xmm3, %xmm2
+; SSE2-NEXT:    psrlw $7, %xmm1
+; SSE2-NEXT:    pand {{.*}}(%rip), %xmm1
+; SSE2-NEXT:    paddb %xmm2, %xmm1
+; SSE2-NEXT:    movdqa %xmm1, %xmm2
+; SSE2-NEXT:    punpckhbw {{.*#+}} xmm2 = xmm2[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; SSE2-NEXT:    psraw $8, %xmm2
+; SSE2-NEXT:    movdqa {{.*#+}} xmm3 = [7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7]
+; SSE2-NEXT:    psraw $8, %xmm3
+; SSE2-NEXT:    pmullw %xmm3, %xmm2
+; SSE2-NEXT:    movdqa {{.*#+}} xmm4 = [255,255,255,255,255,255,255,255]
+; SSE2-NEXT:    pand %xmm4, %xmm2
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    psraw $8, %xmm1
+; SSE2-NEXT:    pmullw %xmm3, %xmm1
+; SSE2-NEXT:    pand %xmm4, %xmm1
+; SSE2-NEXT:    packuswb %xmm2, %xmm1
+; SSE2-NEXT:    psubb %xmm1, %xmm0
 ; SSE2-NEXT:    retq
 ;
 ; SSE41-LABEL: test_rem7_16i8:
 ; SSE41:       # BB#0:
-; SSE41-NEXT:    pextrb $1, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %edx
-; SSE41-NEXT:    imull $-109, %edx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb $7, %dil
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %edx
-; SSE41-NEXT:    pextrb $0, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %esi
-; SSE41-NEXT:    imull $-109, %esi, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    movb %al, %cl
-; SSE41-NEXT:    shrb $7, %cl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %sil
-; SSE41-NEXT:    movzbl %sil, %eax
-; SSE41-NEXT:    movd %eax, %xmm1
-; SSE41-NEXT:    pinsrb $1, %edx, %xmm1
-; SSE41-NEXT:    pextrb $2, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $2, %eax, %xmm1
-; SSE41-NEXT:    pextrb $3, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $3, %eax, %xmm1
-; SSE41-NEXT:    pextrb $4, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $4, %eax, %xmm1
-; SSE41-NEXT:    pextrb $5, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $5, %eax, %xmm1
-; SSE41-NEXT:    pextrb $6, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $6, %eax, %xmm1
-; SSE41-NEXT:    pextrb $7, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $7, %eax, %xmm1
-; SSE41-NEXT:    pextrb $8, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $8, %eax, %xmm1
-; SSE41-NEXT:    pextrb $9, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $9, %eax, %xmm1
-; SSE41-NEXT:    pextrb $10, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $10, %eax, %xmm1
-; SSE41-NEXT:    pextrb $11, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $11, %eax, %xmm1
-; SSE41-NEXT:    pextrb $12, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $12, %eax, %xmm1
-; SSE41-NEXT:    pextrb $13, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $13, %eax, %xmm1
-; SSE41-NEXT:    pextrb $14, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $14, %eax, %xmm1
-; SSE41-NEXT:    pextrb $15, %xmm0, %eax
-; SSE41-NEXT:    movsbl %al, %ecx
-; SSE41-NEXT:    imull $-109, %ecx, %eax
-; SSE41-NEXT:    shrl $8, %eax
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    movb %al, %dl
-; SSE41-NEXT:    shrb $7, %dl
-; SSE41-NEXT:    sarb $2, %al
-; SSE41-NEXT:    addb %dl, %al
-; SSE41-NEXT:    mulb %dil
-; SSE41-NEXT:    subb %al, %cl
-; SSE41-NEXT:    movzbl %cl, %eax
-; SSE41-NEXT:    pinsrb $15, %eax, %xmm1
-; SSE41-NEXT:    movdqa %xmm1, %xmm0
+; SSE41-NEXT:    pmovsxbw %xmm0, %xmm1
+; SSE41-NEXT:    pmovsxbw {{.*}}(%rip), %xmm2
+; SSE41-NEXT:    pmullw %xmm2, %xmm1
+; SSE41-NEXT:    psrlw $8, %xmm1
+; SSE41-NEXT:    pshufd {{.*#+}} xmm3 = xmm0[2,3,0,1]
+; SSE41-NEXT:    pmovsxbw %xmm3, %xmm3
+; SSE41-NEXT:    pmullw %xmm2, %xmm3
+; SSE41-NEXT:    psrlw $8, %xmm3
+; SSE41-NEXT:    packuswb %xmm3, %xmm1
+; SSE41-NEXT:    paddb %xmm0, %xmm1
+; SSE41-NEXT:    movdqa %xmm1, %xmm2
+; SSE41-NEXT:    psrlw $2, %xmm2
+; SSE41-NEXT:    pand {{.*}}(%rip), %xmm2
+; SSE41-NEXT:    movdqa {{.*#+}} xmm3 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; SSE41-NEXT:    pxor %xmm3, %xmm2
+; SSE41-NEXT:    psubb %xmm3, %xmm2
+; SSE41-NEXT:    psrlw $7, %xmm1
+; SSE41-NEXT:    pand {{.*}}(%rip), %xmm1
+; SSE41-NEXT:    paddb %xmm2, %xmm1
+; SSE41-NEXT:    pmovsxbw %xmm1, %xmm2
+; SSE41-NEXT:    pmovsxbw {{.*}}(%rip), %xmm3
+; SSE41-NEXT:    pmullw %xmm3, %xmm2
+; SSE41-NEXT:    movdqa {{.*#+}} xmm4 = [255,255,255,255,255,255,255,255]
+; SSE41-NEXT:    pand %xmm4, %xmm2
+; SSE41-NEXT:    pshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
+; SSE41-NEXT:    pmovsxbw %xmm1, %xmm1
+; SSE41-NEXT:    pmullw %xmm3, %xmm1
+; SSE41-NEXT:    pand %xmm4, %xmm1
+; SSE41-NEXT:    packuswb %xmm1, %xmm2
+; SSE41-NEXT:    psubb %xmm2, %xmm0
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: test_rem7_16i8:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpextrb $1, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %edx
-; AVX-NEXT:    imull $-109, %edx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb $7, %dil
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %edx
-; AVX-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %esi
-; AVX-NEXT:    imull $-109, %esi, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    movb %al, %cl
-; AVX-NEXT:    shrb $7, %cl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %sil
-; AVX-NEXT:    movzbl %sil, %eax
-; AVX-NEXT:    vmovd %eax, %xmm1
-; AVX-NEXT:    vpinsrb $1, %edx, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $2, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $3, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $3, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $5, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $5, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $6, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $7, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $7, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $8, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $9, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $9, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $10, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $10, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $11, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $11, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $12, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $13, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $13, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $14, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $14, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $15, %xmm0, %eax
-; AVX-NEXT:    movsbl %al, %ecx
-; AVX-NEXT:    imull $-109, %ecx, %eax
-; AVX-NEXT:    shrl $8, %eax
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    movb %al, %dl
-; AVX-NEXT:    shrb $7, %dl
-; AVX-NEXT:    sarb $2, %al
-; AVX-NEXT:    addb %dl, %al
-; AVX-NEXT:    mulb %dil
-; AVX-NEXT:    subb %al, %cl
-; AVX-NEXT:    movzbl %cl, %eax
-; AVX-NEXT:    vpinsrb $15, %eax, %xmm1, %xmm0
-; AVX-NEXT:    retq
+; AVX1-LABEL: test_rem7_16i8:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpmovsxbw %xmm0, %xmm1
+; AVX1-NEXT:    vpmovsxbw {{.*}}(%rip), %xmm2
+; AVX1-NEXT:    vpmullw %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpsrlw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[2,3,0,1]
+; AVX1-NEXT:    vpmovsxbw %xmm3, %xmm3
+; AVX1-NEXT:    vpmullw %xmm2, %xmm3, %xmm2
+; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
+; AVX1-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpaddb %xmm0, %xmm1, %xmm1
+; AVX1-NEXT:    vpsrlw $2, %xmm1, %xmm2
+; AVX1-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm3 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; AVX1-NEXT:    vpxor %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vpsubb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vpsrlw $7, %xmm1, %xmm1
+; AVX1-NEXT:    vpand {{.*}}(%rip), %xmm1, %xmm1
+; AVX1-NEXT:    vpaddb %xmm1, %xmm2, %xmm1
+; AVX1-NEXT:    vpmovsxbw %xmm1, %xmm2
+; AVX1-NEXT:    vpmovsxbw {{.*}}(%rip), %xmm3
+; AVX1-NEXT:    vpmullw %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm4 = [255,255,255,255,255,255,255,255]
+; AVX1-NEXT:    vpand %xmm4, %xmm2, %xmm2
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
+; AVX1-NEXT:    vpmovsxbw %xmm1, %xmm1
+; AVX1-NEXT:    vpmullw %xmm3, %xmm1, %xmm1
+; AVX1-NEXT:    vpand %xmm4, %xmm1, %xmm1
+; AVX1-NEXT:    vpackuswb %xmm1, %xmm2, %xmm1
+; AVX1-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: test_rem7_16i8:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpmovsxbw %xmm0, %ymm1
+; AVX2-NEXT:    vpmovsxbw {{.*}}(%rip), %ymm2
+; AVX2-NEXT:    vpmullw %ymm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpaddb %xmm0, %xmm1, %xmm1
+; AVX2-NEXT:    vpsrlw $2, %xmm1, %xmm2
+; AVX2-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
+; AVX2-NEXT:    vmovdqa {{.*#+}} xmm3 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; AVX2-NEXT:    vpxor %xmm3, %xmm2, %xmm2
+; AVX2-NEXT:    vpsubb %xmm3, %xmm2, %xmm2
+; AVX2-NEXT:    vpsrlw $7, %xmm1, %xmm1
+; AVX2-NEXT:    vpand {{.*}}(%rip), %xmm1, %xmm1
+; AVX2-NEXT:    vpaddb %xmm1, %xmm2, %xmm1
+; AVX2-NEXT:    vpmovsxbw %xmm1, %ymm1
+; AVX2-NEXT:    vpmovsxbw {{.*}}(%rip), %ymm2
+; AVX2-NEXT:    vpmullw %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:    vpsubb %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vzeroupper
+; AVX2-NEXT:    retq
   %res = srem <16 x i8> %a, <i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7>
   ret <16 x i8> %res
 }

llvm/trunk/test/CodeGen/X86/vector-idiv-sdiv-256.ll

; AVX2-NEXT:    retq
   %res = sdiv <16 x i16> %a, <i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7>
   ret <16 x i16> %res
 }
 
 define <32 x i8> @test_div7_32i8(<32 x i8> %a) nounwind {
 ; AVX1-LABEL: test_div7_32i8:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
-; AVX1-NEXT:    vpextrb $1, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpextrb $0, %xmm1, %ecx
-; AVX1-NEXT:    movsbl %cl, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %edx
-; AVX1-NEXT:    shrl $8, %edx
-; AVX1-NEXT:    addb %dl, %cl
-; AVX1-NEXT:    movb %cl, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %cl
-; AVX1-NEXT:    addb %dl, %cl
-; AVX1-NEXT:    movzbl %cl, %ecx
-; AVX1-NEXT:    vmovd %ecx, %xmm2
-; AVX1-NEXT:    vpinsrb $1, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $2, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $3, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $4, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $5, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $6, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $7, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $8, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $9, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $10, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $11, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $12, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $13, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $14, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $15, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm1
-; AVX1-NEXT:    vpextrb $1, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpextrb $0, %xmm0, %ecx
-; AVX1-NEXT:    movsbl %cl, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %edx
-; AVX1-NEXT:    shrl $8, %edx
-; AVX1-NEXT:    addb %dl, %cl
-; AVX1-NEXT:    movb %cl, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %cl
-; AVX1-NEXT:    addb %dl, %cl
-; AVX1-NEXT:    movzbl %cl, %ecx
-; AVX1-NEXT:    vmovd %ecx, %xmm2
-; AVX1-NEXT:    vpinsrb $1, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $2, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $3, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $5, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $6, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $7, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $9, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $10, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $11, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $13, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $14, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $15, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %eax
-; AVX1-NEXT:    imull $-109, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm0
+; AVX1-NEXT:    vpmovsxbw %xmm1, %xmm2
+; AVX1-NEXT:    vpmovsxbw {{.*}}(%rip), %xmm3
+; AVX1-NEXT:    vpmullw %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm4 = xmm1[2,3,0,1]
+; AVX1-NEXT:    vpmovsxbw %xmm4, %xmm4
+; AVX1-NEXT:    vpmullw %xmm3, %xmm4, %xmm4
+; AVX1-NEXT:    vpsrlw $8, %xmm4, %xmm4
+; AVX1-NEXT:    vpackuswb %xmm4, %xmm2, %xmm2
+; AVX1-NEXT:    vpaddb %xmm1, %xmm2, %xmm1
+; AVX1-NEXT:    vpsrlw $7, %xmm1, %xmm2
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm4 = [1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1]
+; AVX1-NEXT:    vpand %xmm4, %xmm2, %xmm2
+; AVX1-NEXT:    vpsrlw $2, %xmm1, %xmm1
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm5 = [63,63,63,63,63,63,63,63,63,63,63,63,63,63,63,63]
+; AVX1-NEXT:    vpand %xmm5, %xmm1, %xmm1
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm6 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; AVX1-NEXT:    vpxor %xmm6, %xmm1, %xmm1
+; AVX1-NEXT:    vpsubb %xmm6, %xmm1, %xmm1
+; AVX1-NEXT:    vpaddb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpmovsxbw %xmm0, %xmm2
+; AVX1-NEXT:    vpmullw %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm7 = xmm0[2,3,0,1]
+; AVX1-NEXT:    vpmovsxbw %xmm7, %xmm7
+; AVX1-NEXT:    vpmullw %xmm3, %xmm7, %xmm3
+; AVX1-NEXT:    vpsrlw $8, %xmm3, %xmm3
+; AVX1-NEXT:    vpackuswb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vpaddb %xmm0, %xmm2, %xmm0
+; AVX1-NEXT:    vpsrlw $7, %xmm0, %xmm2
+; AVX1-NEXT:    vpand %xmm4, %xmm2, %xmm2
+; AVX1-NEXT:    vpsrlw $2, %xmm0, %xmm0
+; AVX1-NEXT:    vpand %xmm5, %xmm0, %xmm0
+; AVX1-NEXT:    vpxor %xmm6, %xmm0, %xmm0
+; AVX1-NEXT:    vpsubb %xmm6, %xmm0, %xmm0
+; AVX1-NEXT:    vpaddb %xmm2, %xmm0, %xmm0
 ; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
 ; AVX1-NEXT:    retq
 ;
 ; AVX2-LABEL: test_div7_32i8:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX2-NEXT:    vpextrb $1, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpextrb $0, %xmm1, %ecx
-; AVX2-NEXT:    movsbl %cl, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %edx
-; AVX2-NEXT:    shrl $8, %edx
-; AVX2-NEXT:    addb %dl, %cl
-; AVX2-NEXT:    movb %cl, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %cl
-; AVX2-NEXT:    addb %dl, %cl
-; AVX2-NEXT:    movzbl %cl, %ecx
-; AVX2-NEXT:    vmovd %ecx, %xmm2
-; AVX2-NEXT:    vpinsrb $1, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $2, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $3, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $4, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $5, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $6, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $7, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $8, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $9, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $10, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $11, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $12, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $13, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $14, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $15, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm1
-; AVX2-NEXT:    vpextrb $1, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpextrb $0, %xmm0, %ecx
-; AVX2-NEXT:    movsbl %cl, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %edx
-; AVX2-NEXT:    shrl $8, %edx
-; AVX2-NEXT:    addb %dl, %cl
-; AVX2-NEXT:    movb %cl, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %cl
-; AVX2-NEXT:    addb %dl, %cl
-; AVX2-NEXT:    movzbl %cl, %ecx
-; AVX2-NEXT:    vmovd %ecx, %xmm2
-; AVX2-NEXT:    vpinsrb $1, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $2, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $3, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $5, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $6, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $7, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $9, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $10, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $11, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $13, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $14, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $15, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %eax
-; AVX2-NEXT:    imull $-109, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm0
-; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT:    vmovdqa {{.*#+}} ymm1 = [147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147]
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpmovsxbw %xmm2, %ymm2
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX2-NEXT:    vpmovsxbw %xmm3, %ymm3
+; AVX2-NEXT:    vpmullw %ymm2, %ymm3, %ymm2
+; AVX2-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; AVX2-NEXT:    vpmovsxbw %xmm1, %ymm1
+; AVX2-NEXT:    vpmovsxbw %xmm0, %ymm3
+; AVX2-NEXT:    vpmullw %ymm1, %ymm3, %ymm1
+; AVX2-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm3 = ymm1[2,3],ymm2[2,3]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpackuswb %ymm3, %ymm1, %ymm1
+; AVX2-NEXT:    vpaddb %ymm0, %ymm1, %ymm0
+; AVX2-NEXT:    vpsrlw $2, %ymm0, %ymm1
+; AVX2-NEXT:    vpand {{.*}}(%rip), %ymm1, %ymm1
+; AVX2-NEXT:    vmovdqa {{.*#+}} ymm2 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; AVX2-NEXT:    vpxor %ymm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpsubb %ymm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpsrlw $7, %ymm0, %ymm0
+; AVX2-NEXT:    vpand {{.*}}(%rip), %ymm0, %ymm0
+; AVX2-NEXT:    vpaddb %ymm0, %ymm1, %ymm0
 ; AVX2-NEXT:    retq
   %res = sdiv <32 x i8> %a, <i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7>
   ret <32 x i8> %res
 }
 
 ;
 ; srem by 7
 ;

 ; AVX2-NEXT:    retq
   %res = srem <16 x i16> %a, <i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7>
   ret <16 x i16> %res
 }
 
 define <32 x i8> @test_rem7_32i8(<32 x i8> %a) nounwind {
 ; AVX1-LABEL: test_rem7_32i8:
 ; AVX1:       # BB#0:
-; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
-; AVX1-NEXT:    vpextrb $1, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %edx
-; AVX1-NEXT:    imull $-109, %edx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb $7, %dil
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %edx
-; AVX1-NEXT:    vpextrb $0, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %esi
-; AVX1-NEXT:    imull $-109, %esi, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %sil
-; AVX1-NEXT:    movzbl %sil, %eax
-; AVX1-NEXT:    vmovd %eax, %xmm2
-; AVX1-NEXT:    vpinsrb $1, %edx, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $2, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $3, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $4, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $5, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $6, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $7, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $8, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $9, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $10, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $11, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $12, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $13, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $14, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $15, %xmm1, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm1
-; AVX1-NEXT:    vpextrb $1, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %esi
-; AVX1-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %edx
-; AVX1-NEXT:    imull $-109, %edx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    movb %al, %cl
-; AVX1-NEXT:    shrb $7, %cl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vmovd %eax, %xmm2
-; AVX1-NEXT:    vpinsrb $1, %esi, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $2, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $3, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $5, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $6, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $7, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $9, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $10, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $11, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $13, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $14, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $15, %xmm0, %eax
-; AVX1-NEXT:    movsbl %al, %ecx
-; AVX1-NEXT:    imull $-109, %ecx, %eax
-; AVX1-NEXT:    shrl $8, %eax
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    movb %al, %dl
-; AVX1-NEXT:    shrb $7, %dl
-; AVX1-NEXT:    sarb $2, %al
-; AVX1-NEXT:    addb %dl, %al
-; AVX1-NEXT:    mulb %dil
-; AVX1-NEXT:    subb %al, %cl
-; AVX1-NEXT:    movzbl %cl, %eax
-; AVX1-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm0
-; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpmovsxbw %xmm2, %xmm3
+; AVX1-NEXT:    vpmovsxbw {{.*}}(%rip), %xmm1
+; AVX1-NEXT:    vpmullw %xmm1, %xmm3, %xmm3
+; AVX1-NEXT:    vpsrlw $8, %xmm3, %xmm3
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm4 = xmm2[2,3,0,1]
+; AVX1-NEXT:    vpmovsxbw %xmm4, %xmm4
+; AVX1-NEXT:    vpmullw %xmm1, %xmm4, %xmm4
+; AVX1-NEXT:    vpsrlw $8, %xmm4, %xmm4
+; AVX1-NEXT:    vpackuswb %xmm4, %xmm3, %xmm3
+; AVX1-NEXT:    vpaddb %xmm2, %xmm3, %xmm3
+; AVX1-NEXT:    vpsrlw $7, %xmm3, %xmm4
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm8 = [1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1]
+; AVX1-NEXT:    vpand %xmm8, %xmm4, %xmm4
+; AVX1-NEXT:    vpsrlw $2, %xmm3, %xmm3
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm9 = [63,63,63,63,63,63,63,63,63,63,63,63,63,63,63,63]
+; AVX1-NEXT:    vpand %xmm9, %xmm3, %xmm3
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm7 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; AVX1-NEXT:    vpxor %xmm7, %xmm3, %xmm3
+; AVX1-NEXT:    vpsubb %xmm7, %xmm3, %xmm3
+; AVX1-NEXT:    vpaddb %xmm4, %xmm3, %xmm3
+; AVX1-NEXT:    vpmovsxbw %xmm3, %xmm4
+; AVX1-NEXT:    vpmovsxbw {{.*}}(%rip), %xmm5
+; AVX1-NEXT:    vpmullw %xmm5, %xmm4, %xmm4
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm6 = [255,255,255,255,255,255,255,255]
+; AVX1-NEXT:    vpand %xmm6, %xmm4, %xmm4
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm3 = xmm3[2,3,0,1]
+; AVX1-NEXT:    vpmovsxbw %xmm3, %xmm3
+; AVX1-NEXT:    vpmullw %xmm5, %xmm3, %xmm3
+; AVX1-NEXT:    vpand %xmm6, %xmm3, %xmm3
+; AVX1-NEXT:    vpackuswb %xmm3, %xmm4, %xmm3
+; AVX1-NEXT:    vpsubb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vpmovsxbw %xmm0, %xmm3
+; AVX1-NEXT:    vpmullw %xmm1, %xmm3, %xmm3
+; AVX1-NEXT:    vpsrlw $8, %xmm3, %xmm3
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm4 = xmm0[2,3,0,1]
+; AVX1-NEXT:    vpmovsxbw %xmm4, %xmm4
+; AVX1-NEXT:    vpmullw %xmm1, %xmm4, %xmm1
+; AVX1-NEXT:    vpsrlw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpackuswb %xmm1, %xmm3, %xmm1
+; AVX1-NEXT:    vpaddb %xmm0, %xmm1, %xmm1
+; AVX1-NEXT:    vpsrlw $7, %xmm1, %xmm3
+; AVX1-NEXT:    vpand %xmm8, %xmm3, %xmm3
+; AVX1-NEXT:    vpsrlw $2, %xmm1, %xmm1
+; AVX1-NEXT:    vpand %xmm9, %xmm1, %xmm1
+; AVX1-NEXT:    vpxor %xmm7, %xmm1, %xmm1
+; AVX1-NEXT:    vpsubb %xmm7, %xmm1, %xmm1
+; AVX1-NEXT:    vpaddb %xmm3, %xmm1, %xmm1
+; AVX1-NEXT:    vpmovsxbw %xmm1, %xmm3
+; AVX1-NEXT:    vpmullw %xmm5, %xmm3, %xmm3
+; AVX1-NEXT:    vpand %xmm6, %xmm3, %xmm3
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
+; AVX1-NEXT:    vpmovsxbw %xmm1, %xmm1
+; AVX1-NEXT:    vpmullw %xmm5, %xmm1, %xmm1
+; AVX1-NEXT:    vpand %xmm6, %xmm1, %xmm1
+; AVX1-NEXT:    vpackuswb %xmm1, %xmm3, %xmm1
+; AVX1-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
 ; AVX1-NEXT:    retq
 ;
 ; AVX2-LABEL: test_rem7_32i8:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX2-NEXT:    vpextrb $1, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %edx
-; AVX2-NEXT:    imull $-109, %edx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb $7, %dil
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %edx
-; AVX2-NEXT:    vpextrb $0, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %esi
-; AVX2-NEXT:    imull $-109, %esi, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %sil
-; AVX2-NEXT:    movzbl %sil, %eax
-; AVX2-NEXT:    vmovd %eax, %xmm2
-; AVX2-NEXT:    vpinsrb $1, %edx, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $2, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $3, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $4, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $5, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $6, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $7, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $8, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $9, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $10, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $11, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $12, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $13, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $14, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $15, %xmm1, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm1
-; AVX2-NEXT:    vpextrb $1, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %esi
-; AVX2-NEXT:    vpextrb $0, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %edx
-; AVX2-NEXT:    imull $-109, %edx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    movb %al, %cl
-; AVX2-NEXT:    shrb $7, %cl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vmovd %eax, %xmm2
-; AVX2-NEXT:    vpinsrb $1, %esi, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $2, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $3, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $5, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $6, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $7, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $9, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $10, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $11, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $13, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $14, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $15, %xmm0, %eax
-; AVX2-NEXT:    movsbl %al, %ecx
-; AVX2-NEXT:    imull $-109, %ecx, %eax
-; AVX2-NEXT:    shrl $8, %eax
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    movb %al, %dl
-; AVX2-NEXT:    shrb $7, %dl
-; AVX2-NEXT:    sarb $2, %al
-; AVX2-NEXT:    addb %dl, %al
-; AVX2-NEXT:    mulb %dil
-; AVX2-NEXT:    subb %al, %cl
-; AVX2-NEXT:    movzbl %cl, %eax
-; AVX2-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm0
-; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT:    vmovdqa {{.*#+}} ymm1 = [147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147,147]
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpmovsxbw %xmm2, %ymm2
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX2-NEXT:    vpmovsxbw %xmm3, %ymm3
+; AVX2-NEXT:    vpmullw %ymm2, %ymm3, %ymm2
+; AVX2-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; AVX2-NEXT:    vpmovsxbw %xmm1, %ymm1
+; AVX2-NEXT:    vpmovsxbw %xmm0, %ymm3
+; AVX2-NEXT:    vpmullw %ymm1, %ymm3, %ymm1
+; AVX2-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm3 = ymm1[2,3],ymm2[2,3]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpackuswb %ymm3, %ymm1, %ymm1
+; AVX2-NEXT:    vpaddb %ymm0, %ymm1, %ymm1
+; AVX2-NEXT:    vpsrlw $2, %ymm1, %ymm2
+; AVX2-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX2-NEXT:    vmovdqa {{.*#+}} ymm3 = [32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32]
+; AVX2-NEXT:    vpxor %ymm3, %ymm2, %ymm2
+; AVX2-NEXT:    vpsubb %ymm3, %ymm2, %ymm2
+; AVX2-NEXT:    vpsrlw $7, %ymm1, %ymm1
+; AVX2-NEXT:    vpand {{.*}}(%rip), %ymm1, %ymm1
+; AVX2-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpmovsxbw %xmm2, %ymm2
+; AVX2-NEXT:    vpmovsxbw {{.*}}(%rip), %ymm3
+; AVX2-NEXT:    vpmullw %ymm3, %ymm2, %ymm2
+; AVX2-NEXT:    vextracti128 $1, %ymm2, %xmm4
+; AVX2-NEXT:    vmovdqa {{.*#+}} xmm5 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX2-NEXT:    vpshufb %xmm5, %xmm4, %xmm4
+; AVX2-NEXT:    vpshufb %xmm5, %xmm2, %xmm2
+; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm4[0]
+; AVX2-NEXT:    vpmovsxbw %xmm1, %ymm1
+; AVX2-NEXT:    vpmullw %ymm3, %ymm1, %ymm1
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm3
+; AVX2-NEXT:    vpshufb %xmm5, %xmm3, %xmm3
+; AVX2-NEXT:    vpshufb %xmm5, %xmm1, %xmm1
+; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpsubb %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %res = srem <32 x i8> %a, <i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7>
   ret <32 x i8> %res
 }

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

 ; AVX-NEXT:    retq
   %res = udiv <8 x i16> %a, <i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7>
   ret <8 x i16> %res
 }
 
 define <16 x i8> @test_div7_16i8(<16 x i8> %a) nounwind {
 ; SSE2-LABEL: test_div7_16i8:
 ; SSE2:       # BB#0:
-; SSE2-NEXT:    pushq %rbp
-; SSE2-NEXT:    pushq %rbx
-; SSE2-NEXT:    movaps %xmm0, -{{[0-9]+}}(%rsp)
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %eax
-; SSE2-NEXT:    imull $37, %eax, %ecx
-; SSE2-NEXT:    shrl $8, %ecx
-; SSE2-NEXT:    subb %cl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %ebx
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %edx
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %r9d
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %eax
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %r11d
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %ecx
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %r8d
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %esi
-; SSE2-NEXT:    imull $37, %esi, %edi
-; SSE2-NEXT:    shrl $8, %edi
-; SSE2-NEXT:    subb %dil, %sil
-; SSE2-NEXT:    shrb %sil
-; SSE2-NEXT:    addb %dil, %sil
-; SSE2-NEXT:    shrb $2, %sil
-; SSE2-NEXT:    movzbl %sil, %esi
-; SSE2-NEXT:    movd %esi, %xmm1
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
-; SSE2-NEXT:    imull $37, %eax, %esi
-; SSE2-NEXT:    shrl $8, %esi
-; SSE2-NEXT:    subb %sil, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %sil, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %eax
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %esi
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %r10d
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %edi
-; SSE2-NEXT:    imull $37, %edi, %ebp
-; SSE2-NEXT:    shrl $8, %ebp
-; SSE2-NEXT:    subb %bpl, %dil
-; SSE2-NEXT:    shrb %dil
-; SSE2-NEXT:    addb %bpl, %dil
-; SSE2-NEXT:    shrb $2, %dil
-; SSE2-NEXT:    movzbl %dil, %edi
-; SSE2-NEXT:    movd %edi, %xmm0
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; SSE2-NEXT:    imull $37, %edx, %edi
-; SSE2-NEXT:    shrl $8, %edi
-; SSE2-NEXT:    subb %dil, %dl
-; SSE2-NEXT:    shrb %dl
-; SSE2-NEXT:    addb %dil, %dl
-; SSE2-NEXT:    shrb $2, %dl
-; SSE2-NEXT:    movzbl %dl, %edx
-; SSE2-NEXT:    movd %edx, %xmm1
-; SSE2-NEXT:    imull $37, %esi, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    subb %dl, %sil
-; SSE2-NEXT:    shrb %sil
-; SSE2-NEXT:    addb %dl, %sil
-; SSE2-NEXT:    shrb $2, %sil
-; SSE2-NEXT:    movzbl %sil, %edx
-; SSE2-NEXT:    movd %edx, %xmm2
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1],xmm2[2],xmm1[2],xmm2[3],xmm1[3],xmm2[4],xmm1[4],xmm2[5],xmm1[5],xmm2[6],xmm1[6],xmm2[7],xmm1[7]
-; SSE2-NEXT:    imull $37, %ecx, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    subb %dl, %cl
-; SSE2-NEXT:    shrb %cl
-; SSE2-NEXT:    addb %dl, %cl
-; SSE2-NEXT:    shrb $2, %cl
-; SSE2-NEXT:    movzbl %cl, %ecx
-; SSE2-NEXT:    movd %ecx, %xmm3
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %ecx
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %edx
-; SSE2-NEXT:    imull $37, %edx, %esi
-; SSE2-NEXT:    shrl $8, %esi
-; SSE2-NEXT:    subb %sil, %dl
-; SSE2-NEXT:    shrb %dl
-; SSE2-NEXT:    addb %sil, %dl
-; SSE2-NEXT:    shrb $2, %dl
-; SSE2-NEXT:    movzbl %dl, %edx
-; SSE2-NEXT:    movd %edx, %xmm1
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1],xmm1[2],xmm3[2],xmm1[3],xmm3[3],xmm1[4],xmm3[4],xmm1[5],xmm3[5],xmm1[6],xmm3[6],xmm1[7],xmm3[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3],xmm1[4],xmm2[4],xmm1[5],xmm2[5],xmm1[6],xmm2[6],xmm1[7],xmm2[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
-; SSE2-NEXT:    imull $37, %ebx, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    subb %dl, %bl
-; SSE2-NEXT:    shrb %bl
-; SSE2-NEXT:    addb %dl, %bl
-; SSE2-NEXT:    shrb $2, %bl
-; SSE2-NEXT:    movzbl %bl, %edx
-; SSE2-NEXT:    movd %edx, %xmm2
-; SSE2-NEXT:    imull $37, %eax, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    subb %dl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-NEXT:    imull $37, %r11d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    subb %al, %r11b
-; SSE2-NEXT:    shrb %r11b
-; SSE2-NEXT:    addb %al, %r11b
-; SSE2-NEXT:    shrb $2, %r11b
-; SSE2-NEXT:    movzbl %r11b, %eax
-; SSE2-NEXT:    movd %eax, %xmm3
-; SSE2-NEXT:    imull $37, %ecx, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    subb %al, %cl
-; SSE2-NEXT:    shrb %cl
-; SSE2-NEXT:    addb %al, %cl
-; SSE2-NEXT:    shrb $2, %cl
-; SSE2-NEXT:    movzbl %cl, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1],xmm2[2],xmm3[2],xmm2[3],xmm3[3],xmm2[4],xmm3[4],xmm2[5],xmm3[5],xmm2[6],xmm3[6],xmm2[7],xmm3[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3],xmm2[4],xmm0[4],xmm2[5],xmm0[5],xmm2[6],xmm0[6],xmm2[7],xmm0[7]
-; SSE2-NEXT:    imull $37, %r9d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    subb %al, %r9b
-; SSE2-NEXT:    shrb %r9b
-; SSE2-NEXT:    addb %al, %r9b
-; SSE2-NEXT:    shrb $2, %r9b
-; SSE2-NEXT:    movzbl %r9b, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    imull $37, %r10d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    subb %al, %r10b
-; SSE2-NEXT:    shrb %r10b
-; SSE2-NEXT:    addb %al, %r10b
-; SSE2-NEXT:    shrb $2, %r10b
-; SSE2-NEXT:    movzbl %r10b, %eax
-; SSE2-NEXT:    movd %eax, %xmm3
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm3 = xmm3[0],xmm0[0],xmm3[1],xmm0[1],xmm3[2],xmm0[2],xmm3[3],xmm0[3],xmm3[4],xmm0[4],xmm3[5],xmm0[5],xmm3[6],xmm0[6],xmm3[7],xmm0[7]
-; SSE2-NEXT:    imull $37, %r8d, %eax
-; SSE2-NEXT:    shrl $8, %eax
-; SSE2-NEXT:    subb %al, %r8b
-; SSE2-NEXT:    shrb %r8b
-; SSE2-NEXT:    addb %al, %r8b
-; SSE2-NEXT:    shrb $2, %r8b
-; SSE2-NEXT:    movzbl %r8b, %eax
-; SSE2-NEXT:    movd %eax, %xmm4
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %eax
-; SSE2-NEXT:    imull $37, %eax, %ecx
-; SSE2-NEXT:    shrl $8, %ecx
-; SSE2-NEXT:    subb %cl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    movzbl %al, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm4[0],xmm0[1],xmm4[1],xmm0[2],xmm4[2],xmm0[3],xmm4[3],xmm0[4],xmm4[4],xmm0[5],xmm4[5],xmm0[6],xmm4[6],xmm0[7],xmm4[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1],xmm0[2],xmm3[2],xmm0[3],xmm3[3],xmm0[4],xmm3[4],xmm0[5],xmm3[5],xmm0[6],xmm3[6],xmm0[7],xmm3[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; SSE2-NEXT:    popq %rbx
-; SSE2-NEXT:    popq %rbp
+; SSE2-NEXT:    movdqa {{.*#+}} xmm1 = [37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37]
+; SSE2-NEXT:    psrlw $8, %xmm1
+; SSE2-NEXT:    movdqa %xmm0, %xmm2
+; SSE2-NEXT:    punpckhbw {{.*#+}} xmm2 = xmm2[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; SSE2-NEXT:    psrlw $8, %xmm2
+; SSE2-NEXT:    pmullw %xmm1, %xmm2
+; SSE2-NEXT:    psrlw $8, %xmm2
+; SSE2-NEXT:    movdqa %xmm0, %xmm3
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm3 = xmm3[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    psrlw $8, %xmm3
+; SSE2-NEXT:    pmullw %xmm1, %xmm3
+; SSE2-NEXT:    psrlw $8, %xmm3
+; SSE2-NEXT:    packuswb %xmm2, %xmm3
+; SSE2-NEXT:    psubb %xmm3, %xmm0
+; SSE2-NEXT:    psrlw $1, %xmm0
+; SSE2-NEXT:    pand {{.*}}(%rip), %xmm0
+; SSE2-NEXT:    paddb %xmm3, %xmm0
+; SSE2-NEXT:    psrlw $2, %xmm0
+; SSE2-NEXT:    pand {{.*}}(%rip), %xmm0
 ; SSE2-NEXT:    retq
 ;
 ; SSE41-LABEL: test_div7_16i8:
 ; SSE41:       # BB#0:
-; SSE41-NEXT:    pextrb $1, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pextrb $0, %xmm0, %ecx
-; SSE41-NEXT:    imull $37, %ecx, %edx
-; SSE41-NEXT:    shrl $8, %edx
-; SSE41-NEXT:    subb %dl, %cl
-; SSE41-NEXT:    shrb %cl
-; SSE41-NEXT:    addb %dl, %cl
-; SSE41-NEXT:    shrb $2, %cl
-; SSE41-NEXT:    movzbl %cl, %ecx
-; SSE41-NEXT:    movd %ecx, %xmm1
-; SSE41-NEXT:    pinsrb $1, %eax, %xmm1
-; SSE41-NEXT:    pextrb $2, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $2, %eax, %xmm1
-; SSE41-NEXT:    pextrb $3, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $3, %eax, %xmm1
-; SSE41-NEXT:    pextrb $4, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $4, %eax, %xmm1
-; SSE41-NEXT:    pextrb $5, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $5, %eax, %xmm1
-; SSE41-NEXT:    pextrb $6, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $6, %eax, %xmm1
-; SSE41-NEXT:    pextrb $7, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $7, %eax, %xmm1
-; SSE41-NEXT:    pextrb $8, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $8, %eax, %xmm1
-; SSE41-NEXT:    pextrb $9, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $9, %eax, %xmm1
-; SSE41-NEXT:    pextrb $10, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $10, %eax, %xmm1
-; SSE41-NEXT:    pextrb $11, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $11, %eax, %xmm1
-; SSE41-NEXT:    pextrb $12, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $12, %eax, %xmm1
-; SSE41-NEXT:    pextrb $13, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $13, %eax, %xmm1
-; SSE41-NEXT:    pextrb $14, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $14, %eax, %xmm1
-; SSE41-NEXT:    pextrb $15, %xmm0, %eax
-; SSE41-NEXT:    imull $37, %eax, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movzbl %al, %eax
-; SSE41-NEXT:    pinsrb $15, %eax, %xmm1
-; SSE41-NEXT:    movdqa %xmm1, %xmm0
+; SSE41-NEXT:    pmovzxbw {{.*#+}} xmm1 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero
+; SSE41-NEXT:    pmovzxbw {{.*#+}} xmm2 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
+; SSE41-NEXT:    pmullw %xmm2, %xmm1
+; SSE41-NEXT:    psrlw $8, %xmm1
+; SSE41-NEXT:    pshufd {{.*#+}} xmm3 = xmm0[2,3,0,1]
+; SSE41-NEXT:    pmovzxbw {{.*#+}} xmm3 = xmm3[0],zero,xmm3[1],zero,xmm3[2],zero,xmm3[3],zero,xmm3[4],zero,xmm3[5],zero,xmm3[6],zero,xmm3[7],zero
+; SSE41-NEXT:    pmullw %xmm2, %xmm3
+; SSE41-NEXT:    psrlw $8, %xmm3
+; SSE41-NEXT:    packuswb %xmm3, %xmm1
+; SSE41-NEXT:    psubb %xmm1, %xmm0
+; SSE41-NEXT:    psrlw $1, %xmm0
+; SSE41-NEXT:    pand {{.*}}(%rip), %xmm0
+; SSE41-NEXT:    paddb %xmm1, %xmm0
+; SSE41-NEXT:    psrlw $2, %xmm0
+; SSE41-NEXT:    pand {{.*}}(%rip), %xmm0
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: test_div7_16i8:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpextrb $1, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpextrb $0, %xmm0, %ecx
-; AVX-NEXT:    imull $37, %ecx, %edx
-; AVX-NEXT:    shrl $8, %edx
-; AVX-NEXT:    subb %dl, %cl
-; AVX-NEXT:    shrb %cl
-; AVX-NEXT:    addb %dl, %cl
-; AVX-NEXT:    shrb $2, %cl
-; AVX-NEXT:    movzbl %cl, %ecx
-; AVX-NEXT:    vmovd %ecx, %xmm1
-; AVX-NEXT:    vpinsrb $1, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $2, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $3, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $3, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $5, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $5, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $6, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $7, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $7, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $8, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $9, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $9, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $10, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $10, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $11, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $11, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $12, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $13, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $13, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $14, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $14, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $15, %xmm0, %eax
-; AVX-NEXT:    imull $37, %eax, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movzbl %al, %eax
-; AVX-NEXT:    vpinsrb $15, %eax, %xmm1, %xmm0
-; AVX-NEXT:    retq
+; AVX1-LABEL: test_div7_16i8:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm1 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm2 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
+; AVX1-NEXT:    vpmullw %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpsrlw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[2,3,0,1]
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm3 = xmm3[0],zero,xmm3[1],zero,xmm3[2],zero,xmm3[3],zero,xmm3[4],zero,xmm3[5],zero,xmm3[6],zero,xmm3[7],zero
+; AVX1-NEXT:    vpmullw %xmm2, %xmm3, %xmm2
+; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
+; AVX1-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    vpsrlw $1, %xmm0, %xmm0
+; AVX1-NEXT:    vpand {{.*}}(%rip), %xmm0, %xmm0
+; AVX1-NEXT:    vpaddb %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    vpsrlw $2, %xmm0, %xmm0
+; AVX1-NEXT:    vpand {{.*}}(%rip), %xmm0, %xmm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: test_div7_16i8:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm1 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero,xmm0[8],zero,xmm0[9],zero,xmm0[10],zero,xmm0[11],zero,xmm0[12],zero,xmm0[13],zero,xmm0[14],zero,xmm0[15],zero
+; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm2 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero,mem[8],zero,mem[9],zero,mem[10],zero,mem[11],zero,mem[12],zero,mem[13],zero,mem[14],zero,mem[15],zero
+; AVX2-NEXT:    vpmullw %ymm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vpsrlw $1, %xmm0, %xmm0
+; AVX2-NEXT:    vpand {{.*}}(%rip), %xmm0, %xmm0
+; AVX2-NEXT:    vpaddb %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vpsrlw $2, %xmm0, %xmm0
+; AVX2-NEXT:    vpand {{.*}}(%rip), %xmm0, %xmm0
+; AVX2-NEXT:    vzeroupper
+; AVX2-NEXT:    retq
   %res = udiv <16 x i8> %a, <i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7>
   ret <16 x i8> %res
 }
 
 ;
 ; urem by 7
 ;
 

 ; AVX-NEXT:    retq
   %res = urem <8 x i16> %a, <i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7>
   ret <8 x i16> %res
 }
 
 define <16 x i8> @test_rem7_16i8(<16 x i8> %a) nounwind {
 ; SSE2-LABEL: test_rem7_16i8:
 ; SSE2:       # BB#0:
-; SSE2-NEXT:    pushq %rbp
-; SSE2-NEXT:    pushq %r15
-; SSE2-NEXT:    pushq %r14
-; SSE2-NEXT:    pushq %rbx
-; SSE2-NEXT:    movaps %xmm0, -{{[0-9]+}}(%rsp)
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %ecx
-; SSE2-NEXT:    imull $37, %ecx, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    movb %cl, %al
-; SSE2-NEXT:    subb %dl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    movb $7, %r11b
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %cl
-; SSE2-NEXT:    movzbl %cl, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %r15d
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %ebx
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %r9d
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %edi
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %r14d
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %ecx
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %r8d
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %ebp
-; SSE2-NEXT:    imull $37, %ebp, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    movb %bpl, %al
-; SSE2-NEXT:    subb %dl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %bpl
-; SSE2-NEXT:    movzbl %bpl, %eax
-; SSE2-NEXT:    movd %eax, %xmm1
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
-; SSE2-NEXT:    imull $37, %edi, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    movb %dil, %al
-; SSE2-NEXT:    subb %dl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %dil
-; SSE2-NEXT:    movzbl %dil, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %edi
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %ebp
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %r10d
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %edx
-; SSE2-NEXT:    imull $37, %edx, %esi
-; SSE2-NEXT:    shrl $8, %esi
-; SSE2-NEXT:    movb %dl, %al
-; SSE2-NEXT:    subb %sil, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %sil, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %dl
-; SSE2-NEXT:    movzbl %dl, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; SSE2-NEXT:    imull $37, %ebx, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    movb %bl, %al
-; SSE2-NEXT:    subb %dl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %bl
-; SSE2-NEXT:    movzbl %bl, %eax
-; SSE2-NEXT:    movd %eax, %xmm1
-; SSE2-NEXT:    imull $37, %ebp, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    movb %bpl, %al
-; SSE2-NEXT:    subb %dl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %bpl
-; SSE2-NEXT:    movzbl %bpl, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1],xmm2[2],xmm1[2],xmm2[3],xmm1[3],xmm2[4],xmm1[4],xmm2[5],xmm1[5],xmm2[6],xmm1[6],xmm2[7],xmm1[7]
-; SSE2-NEXT:    imull $37, %ecx, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    movb %cl, %al
-; SSE2-NEXT:    subb %dl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %cl
-; SSE2-NEXT:    movzbl %cl, %eax
-; SSE2-NEXT:    movd %eax, %xmm3
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %ecx
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %edx
-; SSE2-NEXT:    imull $37, %edx, %esi
-; SSE2-NEXT:    shrl $8, %esi
-; SSE2-NEXT:    movb %dl, %al
-; SSE2-NEXT:    subb %sil, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %sil, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %dl
-; SSE2-NEXT:    movzbl %dl, %eax
-; SSE2-NEXT:    movd %eax, %xmm1
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1],xmm1[2],xmm3[2],xmm1[3],xmm3[3],xmm1[4],xmm3[4],xmm1[5],xmm3[5],xmm1[6],xmm3[6],xmm1[7],xmm3[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1],xmm1[2],xmm2[2],xmm1[3],xmm2[3],xmm1[4],xmm2[4],xmm1[5],xmm2[5],xmm1[6],xmm2[6],xmm1[7],xmm2[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7]
-; SSE2-NEXT:    imull $37, %r15d, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    movb %r15b, %al
-; SSE2-NEXT:    subb %dl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %r15b
-; SSE2-NEXT:    movzbl %r15b, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    imull $37, %edi, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    movb %dil, %al
-; SSE2-NEXT:    subb %dl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %dil
-; SSE2-NEXT:    movzbl %dil, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-NEXT:    imull $37, %r14d, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    movb %r14b, %al
-; SSE2-NEXT:    subb %dl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %r14b
-; SSE2-NEXT:    movzbl %r14b, %eax
-; SSE2-NEXT:    movd %eax, %xmm3
-; SSE2-NEXT:    imull $37, %ecx, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    movb %cl, %al
-; SSE2-NEXT:    subb %dl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %cl
-; SSE2-NEXT:    movzbl %cl, %eax
-; SSE2-NEXT:    movd %eax, %xmm2
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1],xmm2[2],xmm3[2],xmm2[3],xmm3[3],xmm2[4],xmm3[4],xmm2[5],xmm3[5],xmm2[6],xmm3[6],xmm2[7],xmm3[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm2 = xmm2[0],xmm0[0],xmm2[1],xmm0[1],xmm2[2],xmm0[2],xmm2[3],xmm0[3],xmm2[4],xmm0[4],xmm2[5],xmm0[5],xmm2[6],xmm0[6],xmm2[7],xmm0[7]
-; SSE2-NEXT:    imull $37, %r9d, %ecx
-; SSE2-NEXT:    shrl $8, %ecx
-; SSE2-NEXT:    movb %r9b, %al
-; SSE2-NEXT:    subb %cl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %r9b
-; SSE2-NEXT:    movzbl %r9b, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    imull $37, %r10d, %ecx
-; SSE2-NEXT:    shrl $8, %ecx
-; SSE2-NEXT:    movb %r10b, %al
-; SSE2-NEXT:    subb %cl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %r10b
-; SSE2-NEXT:    movzbl %r10b, %eax
-; SSE2-NEXT:    movd %eax, %xmm3
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm3 = xmm3[0],xmm0[0],xmm3[1],xmm0[1],xmm3[2],xmm0[2],xmm3[3],xmm0[3],xmm3[4],xmm0[4],xmm3[5],xmm0[5],xmm3[6],xmm0[6],xmm3[7],xmm0[7]
-; SSE2-NEXT:    imull $37, %r8d, %ecx
-; SSE2-NEXT:    shrl $8, %ecx
-; SSE2-NEXT:    movb %r8b, %al
-; SSE2-NEXT:    subb %cl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %cl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %r8b
-; SSE2-NEXT:    movzbl %r8b, %eax
-; SSE2-NEXT:    movd %eax, %xmm4
-; SSE2-NEXT:    movzbl -{{[0-9]+}}(%rsp), %ecx
-; SSE2-NEXT:    imull $37, %ecx, %edx
-; SSE2-NEXT:    shrl $8, %edx
-; SSE2-NEXT:    movb %cl, %al
-; SSE2-NEXT:    subb %dl, %al
-; SSE2-NEXT:    shrb %al
-; SSE2-NEXT:    addb %dl, %al
-; SSE2-NEXT:    shrb $2, %al
-; SSE2-NEXT:    mulb %r11b
-; SSE2-NEXT:    subb %al, %cl
-; SSE2-NEXT:    movzbl %cl, %eax
-; SSE2-NEXT:    movd %eax, %xmm0
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm4[0],xmm0[1],xmm4[1],xmm0[2],xmm4[2],xmm0[3],xmm4[3],xmm0[4],xmm4[4],xmm0[5],xmm4[5],xmm0[6],xmm4[6],xmm0[7],xmm4[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1],xmm0[2],xmm3[2],xmm0[3],xmm3[3],xmm0[4],xmm3[4],xmm0[5],xmm3[5],xmm0[6],xmm3[6],xmm0[7],xmm3[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1],xmm0[2],xmm2[2],xmm0[3],xmm2[3],xmm0[4],xmm2[4],xmm0[5],xmm2[5],xmm0[6],xmm2[6],xmm0[7],xmm2[7]
-; SSE2-NEXT:    punpcklbw {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3],xmm0[4],xmm1[4],xmm0[5],xmm1[5],xmm0[6],xmm1[6],xmm0[7],xmm1[7]
-; SSE2-NEXT:    popq %rbx
-; SSE2-NEXT:    popq %r14
-; SSE2-NEXT:    popq %r15
-; SSE2-NEXT:    popq %rbp
+; SSE2-NEXT:    movdqa {{.*#+}} xmm1 = [37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37]
+; SSE2-NEXT:    psrlw $8, %xmm1
+; SSE2-NEXT:    movdqa %xmm0, %xmm2
+; SSE2-NEXT:    punpckhbw {{.*#+}} xmm2 = xmm2[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; SSE2-NEXT:    psrlw $8, %xmm2
+; SSE2-NEXT:    pmullw %xmm1, %xmm2
+; SSE2-NEXT:    psrlw $8, %xmm2
+; SSE2-NEXT:    movdqa %xmm0, %xmm3
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm3 = xmm3[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    psrlw $8, %xmm3
+; SSE2-NEXT:    pmullw %xmm1, %xmm3
+; SSE2-NEXT:    psrlw $8, %xmm3
+; SSE2-NEXT:    packuswb %xmm2, %xmm3
+; SSE2-NEXT:    movdqa %xmm0, %xmm1
+; SSE2-NEXT:    psubb %xmm3, %xmm1
+; SSE2-NEXT:    psrlw $1, %xmm1
+; SSE2-NEXT:    pand {{.*}}(%rip), %xmm1
+; SSE2-NEXT:    paddb %xmm3, %xmm1
+; SSE2-NEXT:    psrlw $2, %xmm1
+; SSE2-NEXT:    pand {{.*}}(%rip), %xmm1
+; SSE2-NEXT:    movdqa %xmm1, %xmm2
+; SSE2-NEXT:    punpckhbw {{.*#+}} xmm2 = xmm2[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15]
+; SSE2-NEXT:    psraw $8, %xmm2
+; SSE2-NEXT:    movdqa {{.*#+}} xmm3 = [7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7]
+; SSE2-NEXT:    psraw $8, %xmm3
+; SSE2-NEXT:    pmullw %xmm3, %xmm2
+; SSE2-NEXT:    movdqa {{.*#+}} xmm4 = [255,255,255,255,255,255,255,255]
+; SSE2-NEXT:    pand %xmm4, %xmm2
+; SSE2-NEXT:    punpcklbw {{.*#+}} xmm1 = xmm1[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7]
+; SSE2-NEXT:    psraw $8, %xmm1
+; SSE2-NEXT:    pmullw %xmm3, %xmm1
+; SSE2-NEXT:    pand %xmm4, %xmm1
+; SSE2-NEXT:    packuswb %xmm2, %xmm1
+; SSE2-NEXT:    psubb %xmm1, %xmm0
 ; SSE2-NEXT:    retq
 ;
 ; SSE41-LABEL: test_rem7_16i8:
 ; SSE41:       # BB#0:
-; SSE41-NEXT:    pextrb $1, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %ecx
-; SSE41-NEXT:    shrl $8, %ecx
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %cl, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %cl, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    movb $7, %cl
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %edx
-; SSE41-NEXT:    pextrb $0, %xmm0, %esi
-; SSE41-NEXT:    imull $37, %esi, %edi
-; SSE41-NEXT:    shrl $8, %edi
-; SSE41-NEXT:    movb %sil, %al
-; SSE41-NEXT:    subb %dil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %dil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %sil
-; SSE41-NEXT:    movzbl %sil, %eax
-; SSE41-NEXT:    movd %eax, %xmm1
-; SSE41-NEXT:    pinsrb $1, %edx, %xmm1
-; SSE41-NEXT:    pextrb $2, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $2, %eax, %xmm1
-; SSE41-NEXT:    pextrb $3, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $3, %eax, %xmm1
-; SSE41-NEXT:    pextrb $4, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $4, %eax, %xmm1
-; SSE41-NEXT:    pextrb $5, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $5, %eax, %xmm1
-; SSE41-NEXT:    pextrb $6, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $6, %eax, %xmm1
-; SSE41-NEXT:    pextrb $7, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $7, %eax, %xmm1
-; SSE41-NEXT:    pextrb $8, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $8, %eax, %xmm1
-; SSE41-NEXT:    pextrb $9, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $9, %eax, %xmm1
-; SSE41-NEXT:    pextrb $10, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $10, %eax, %xmm1
-; SSE41-NEXT:    pextrb $11, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $11, %eax, %xmm1
-; SSE41-NEXT:    pextrb $12, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $12, %eax, %xmm1
-; SSE41-NEXT:    pextrb $13, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $13, %eax, %xmm1
-; SSE41-NEXT:    pextrb $14, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $14, %eax, %xmm1
-; SSE41-NEXT:    pextrb $15, %xmm0, %edx
-; SSE41-NEXT:    imull $37, %edx, %esi
-; SSE41-NEXT:    shrl $8, %esi
-; SSE41-NEXT:    movb %dl, %al
-; SSE41-NEXT:    subb %sil, %al
-; SSE41-NEXT:    shrb %al
-; SSE41-NEXT:    addb %sil, %al
-; SSE41-NEXT:    shrb $2, %al
-; SSE41-NEXT:    mulb %cl
-; SSE41-NEXT:    subb %al, %dl
-; SSE41-NEXT:    movzbl %dl, %eax
-; SSE41-NEXT:    pinsrb $15, %eax, %xmm1
-; SSE41-NEXT:    movdqa %xmm1, %xmm0
+; SSE41-NEXT:    pmovzxbw {{.*#+}} xmm1 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero
+; SSE41-NEXT:    pmovzxbw {{.*#+}} xmm2 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
+; SSE41-NEXT:    pmullw %xmm2, %xmm1
+; SSE41-NEXT:    psrlw $8, %xmm1
+; SSE41-NEXT:    pshufd {{.*#+}} xmm3 = xmm0[2,3,0,1]
+; SSE41-NEXT:    pmovzxbw {{.*#+}} xmm3 = xmm3[0],zero,xmm3[1],zero,xmm3[2],zero,xmm3[3],zero,xmm3[4],zero,xmm3[5],zero,xmm3[6],zero,xmm3[7],zero
+; SSE41-NEXT:    pmullw %xmm2, %xmm3
+; SSE41-NEXT:    psrlw $8, %xmm3
+; SSE41-NEXT:    packuswb %xmm3, %xmm1
+; SSE41-NEXT:    movdqa %xmm0, %xmm2
+; SSE41-NEXT:    psubb %xmm1, %xmm2
+; SSE41-NEXT:    psrlw $1, %xmm2
+; SSE41-NEXT:    pand {{.*}}(%rip), %xmm2
+; SSE41-NEXT:    paddb %xmm1, %xmm2
+; SSE41-NEXT:    psrlw $2, %xmm2
+; SSE41-NEXT:    pand {{.*}}(%rip), %xmm2
+; SSE41-NEXT:    pmovsxbw %xmm2, %xmm1
+; SSE41-NEXT:    pmovsxbw {{.*}}(%rip), %xmm3
+; SSE41-NEXT:    pmullw %xmm3, %xmm1
+; SSE41-NEXT:    movdqa {{.*#+}} xmm4 = [255,255,255,255,255,255,255,255]
+; SSE41-NEXT:    pand %xmm4, %xmm1
+; SSE41-NEXT:    pshufd {{.*#+}} xmm2 = xmm2[2,3,0,1]
+; SSE41-NEXT:    pmovsxbw %xmm2, %xmm2
+; SSE41-NEXT:    pmullw %xmm3, %xmm2
+; SSE41-NEXT:    pand %xmm4, %xmm2
+; SSE41-NEXT:    packuswb %xmm2, %xmm1
+; SSE41-NEXT:    psubb %xmm1, %xmm0
 ; SSE41-NEXT:    retq
 ;
-; AVX-LABEL: test_rem7_16i8:
-; AVX:       # BB#0:
-; AVX-NEXT:    vpextrb $1, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %ecx
-; AVX-NEXT:    shrl $8, %ecx
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %cl, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %cl, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    movb $7, %cl
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %edx
-; AVX-NEXT:    vpextrb $0, %xmm0, %esi
-; AVX-NEXT:    imull $37, %esi, %edi
-; AVX-NEXT:    shrl $8, %edi
-; AVX-NEXT:    movb %sil, %al
-; AVX-NEXT:    subb %dil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %dil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %sil
-; AVX-NEXT:    movzbl %sil, %eax
-; AVX-NEXT:    vmovd %eax, %xmm1
-; AVX-NEXT:    vpinsrb $1, %edx, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $2, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $2, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $3, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $3, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $4, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $4, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $5, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $5, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $6, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $6, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $7, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $7, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $8, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $8, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $9, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $9, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $10, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $10, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $11, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $11, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $12, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $12, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $13, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $13, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $14, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $14, %eax, %xmm1, %xmm1
-; AVX-NEXT:    vpextrb $15, %xmm0, %edx
-; AVX-NEXT:    imull $37, %edx, %esi
-; AVX-NEXT:    shrl $8, %esi
-; AVX-NEXT:    movb %dl, %al
-; AVX-NEXT:    subb %sil, %al
-; AVX-NEXT:    shrb %al
-; AVX-NEXT:    addb %sil, %al
-; AVX-NEXT:    shrb $2, %al
-; AVX-NEXT:    mulb %cl
-; AVX-NEXT:    subb %al, %dl
-; AVX-NEXT:    movzbl %dl, %eax
-; AVX-NEXT:    vpinsrb $15, %eax, %xmm1, %xmm0
-; AVX-NEXT:    retq
+; AVX1-LABEL: test_rem7_16i8:
+; AVX1:       # BB#0:
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm1 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm2 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
+; AVX1-NEXT:    vpmullw %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpsrlw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm3 = xmm0[2,3,0,1]
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm3 = xmm3[0],zero,xmm3[1],zero,xmm3[2],zero,xmm3[3],zero,xmm3[4],zero,xmm3[5],zero,xmm3[6],zero,xmm3[7],zero
+; AVX1-NEXT:    vpmullw %xmm2, %xmm3, %xmm2
+; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
+; AVX1-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpsubb %xmm1, %xmm0, %xmm2
+; AVX1-NEXT:    vpsrlw $1, %xmm2, %xmm2
+; AVX1-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
+; AVX1-NEXT:    vpaddb %xmm1, %xmm2, %xmm1
+; AVX1-NEXT:    vpsrlw $2, %xmm1, %xmm1
+; AVX1-NEXT:    vpand {{.*}}(%rip), %xmm1, %xmm1
+; AVX1-NEXT:    vpmovsxbw %xmm1, %xmm2
+; AVX1-NEXT:    vpmovsxbw {{.*}}(%rip), %xmm3
+; AVX1-NEXT:    vpmullw %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm4 = [255,255,255,255,255,255,255,255]
+; AVX1-NEXT:    vpand %xmm4, %xmm2, %xmm2
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
+; AVX1-NEXT:    vpmovsxbw %xmm1, %xmm1
+; AVX1-NEXT:    vpmullw %xmm3, %xmm1, %xmm1
+; AVX1-NEXT:    vpand %xmm4, %xmm1, %xmm1
+; AVX1-NEXT:    vpackuswb %xmm1, %xmm2, %xmm1
+; AVX1-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: test_rem7_16i8:
+; AVX2:       # BB#0:
+; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm1 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero,xmm0[8],zero,xmm0[9],zero,xmm0[10],zero,xmm0[11],zero,xmm0[12],zero,xmm0[13],zero,xmm0[14],zero,xmm0[15],zero
+; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm2 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero,mem[8],zero,mem[9],zero,mem[10],zero,mem[11],zero,mem[12],zero,mem[13],zero,mem[14],zero,mem[15],zero
+; AVX2-NEXT:    vpmullw %ymm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpackuswb %xmm2, %xmm1, %xmm1
+; AVX2-NEXT:    vpsubb %xmm1, %xmm0, %xmm2
+; AVX2-NEXT:    vpsrlw $1, %xmm2, %xmm2
+; AVX2-NEXT:    vpand {{.*}}(%rip), %xmm2, %xmm2
+; AVX2-NEXT:    vpaddb %xmm1, %xmm2, %xmm1
+; AVX2-NEXT:    vpsrlw $2, %xmm1, %xmm1
+; AVX2-NEXT:    vpand {{.*}}(%rip), %xmm1, %xmm1
+; AVX2-NEXT:    vpmovsxbw %xmm1, %ymm1
+; AVX2-NEXT:    vpmovsxbw {{.*}}(%rip), %ymm2
+; AVX2-NEXT:    vpmullw %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:    vpsubb %xmm1, %xmm0, %xmm0
+; AVX2-NEXT:    vzeroupper
+; AVX2-NEXT:    retq
   %res = urem <16 x i8> %a, <i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7>
   ret <16 x i8> %res
 }

llvm/trunk/test/CodeGen/X86/vector-idiv-udiv-256.ll

; AVX2-NEXT:    retq
   %res = udiv <16 x i16> %a, <i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7>
   ret <16 x i16> %res
 }
 
 define <32 x i8> @test_div7_32i8(<32 x i8> %a) nounwind {
 ; AVX1-LABEL: test_div7_32i8:
 ; AVX1:       # BB#0:
 ; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
-; AVX1-NEXT:    vpextrb $1, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpextrb $0, %xmm1, %ecx
-; AVX1-NEXT:    imull $37, %ecx, %edx
-; AVX1-NEXT:    shrl $8, %edx
-; AVX1-NEXT:    subb %dl, %cl
-; AVX1-NEXT:    shrb %cl
-; AVX1-NEXT:    addb %dl, %cl
-; AVX1-NEXT:    shrb $2, %cl
-; AVX1-NEXT:    movzbl %cl, %ecx
-; AVX1-NEXT:    vmovd %ecx, %xmm2
-; AVX1-NEXT:    vpinsrb $1, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $2, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $3, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $4, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $5, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $6, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $7, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $8, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $9, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $10, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $11, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $12, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $13, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $14, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $15, %xmm1, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm1
-; AVX1-NEXT:    vpextrb $1, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpextrb $0, %xmm0, %ecx
-; AVX1-NEXT:    imull $37, %ecx, %edx
-; AVX1-NEXT:    shrl $8, %edx
-; AVX1-NEXT:    subb %dl, %cl
-; AVX1-NEXT:    shrb %cl
-; AVX1-NEXT:    addb %dl, %cl
-; AVX1-NEXT:    shrb $2, %cl
-; AVX1-NEXT:    movzbl %cl, %ecx
-; AVX1-NEXT:    vmovd %ecx, %xmm2
-; AVX1-NEXT:    vpinsrb $1, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $2, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $3, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $5, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $6, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $7, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $9, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $10, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $11, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $13, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $14, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $15, %xmm0, %eax
-; AVX1-NEXT:    imull $37, %eax, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movzbl %al, %eax
-; AVX1-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm0
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm2 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm3 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
+; AVX1-NEXT:    vpmullw %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vpsrlw $8, %xmm2, %xmm2
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm4 = xmm1[2,3,0,1]
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm4 = xmm4[0],zero,xmm4[1],zero,xmm4[2],zero,xmm4[3],zero,xmm4[4],zero,xmm4[5],zero,xmm4[6],zero,xmm4[7],zero
+; AVX1-NEXT:    vpmullw %xmm3, %xmm4, %xmm4
+; AVX1-NEXT:    vpsrlw $8, %xmm4, %xmm4
+; AVX1-NEXT:    vpackuswb %xmm4, %xmm2, %xmm2
+; AVX1-NEXT:    vpsubb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpsrlw $1, %xmm1, %xmm1
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm4 = [127,127,127,127,127,127,127,127,127,127,127,127,127,127,127,127]
+; AVX1-NEXT:    vpand %xmm4, %xmm1, %xmm1
+; AVX1-NEXT:    vpaddb %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpsrlw $2, %xmm1, %xmm1
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm2 = [63,63,63,63,63,63,63,63,63,63,63,63,63,63,63,63]
+; AVX1-NEXT:    vpand %xmm2, %xmm1, %xmm1
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm5 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero
+; AVX1-NEXT:    vpmullw %xmm3, %xmm5, %xmm5
+; AVX1-NEXT:    vpsrlw $8, %xmm5, %xmm5
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm6 = xmm0[2,3,0,1]
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm6 = xmm6[0],zero,xmm6[1],zero,xmm6[2],zero,xmm6[3],zero,xmm6[4],zero,xmm6[5],zero,xmm6[6],zero,xmm6[7],zero
+; AVX1-NEXT:    vpmullw %xmm3, %xmm6, %xmm3
+; AVX1-NEXT:    vpsrlw $8, %xmm3, %xmm3
+; AVX1-NEXT:    vpackuswb %xmm3, %xmm5, %xmm3
+; AVX1-NEXT:    vpsubb %xmm3, %xmm0, %xmm0
+; AVX1-NEXT:    vpsrlw $1, %xmm0, %xmm0
+; AVX1-NEXT:    vpand %xmm4, %xmm0, %xmm0
+; AVX1-NEXT:    vpaddb %xmm3, %xmm0, %xmm0
+; AVX1-NEXT:    vpsrlw $2, %xmm0, %xmm0
+; AVX1-NEXT:    vpand %xmm2, %xmm0, %xmm0
 ; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
 ; AVX1-NEXT:    retq
 ;
 ; AVX2-LABEL: test_div7_32i8:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX2-NEXT:    vpextrb $1, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpextrb $0, %xmm1, %ecx
-; AVX2-NEXT:    imull $37, %ecx, %edx
-; AVX2-NEXT:    shrl $8, %edx
-; AVX2-NEXT:    subb %dl, %cl
-; AVX2-NEXT:    shrb %cl
-; AVX2-NEXT:    addb %dl, %cl
-; AVX2-NEXT:    shrb $2, %cl
-; AVX2-NEXT:    movzbl %cl, %ecx
-; AVX2-NEXT:    vmovd %ecx, %xmm2
-; AVX2-NEXT:    vpinsrb $1, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $2, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $3, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $4, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $5, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $6, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $7, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $8, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $9, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $10, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $11, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $12, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $13, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $14, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $15, %xmm1, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm1
-; AVX2-NEXT:    vpextrb $1, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpextrb $0, %xmm0, %ecx
-; AVX2-NEXT:    imull $37, %ecx, %edx
-; AVX2-NEXT:    shrl $8, %edx
-; AVX2-NEXT:    subb %dl, %cl
-; AVX2-NEXT:    shrb %cl
-; AVX2-NEXT:    addb %dl, %cl
-; AVX2-NEXT:    shrb $2, %cl
-; AVX2-NEXT:    movzbl %cl, %ecx
-; AVX2-NEXT:    vmovd %ecx, %xmm2
-; AVX2-NEXT:    vpinsrb $1, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $2, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $3, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $4, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $5, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $6, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $7, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $8, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $9, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $10, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $11, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $12, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $13, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $14, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $15, %xmm0, %eax
-; AVX2-NEXT:    imull $37, %eax, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movzbl %al, %eax
-; AVX2-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm0
-; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT:    vmovdqa {{.*#+}} ymm1 = [37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37]
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm2 = xmm2[0],zero,xmm2[1],zero,xmm2[2],zero,xmm2[3],zero,xmm2[4],zero,xmm2[5],zero,xmm2[6],zero,xmm2[7],zero,xmm2[8],zero,xmm2[9],zero,xmm2[10],zero,xmm2[11],zero,xmm2[12],zero,xmm2[13],zero,xmm2[14],zero,xmm2[15],zero
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm3 = xmm3[0],zero,xmm3[1],zero,xmm3[2],zero,xmm3[3],zero,xmm3[4],zero,xmm3[5],zero,xmm3[6],zero,xmm3[7],zero,xmm3[8],zero,xmm3[9],zero,xmm3[10],zero,xmm3[11],zero,xmm3[12],zero,xmm3[13],zero,xmm3[14],zero,xmm3[15],zero
+; AVX2-NEXT:    vpmullw %ymm2, %ymm3, %ymm2
+; AVX2-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm1 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero,xmm1[8],zero,xmm1[9],zero,xmm1[10],zero,xmm1[11],zero,xmm1[12],zero,xmm1[13],zero,xmm1[14],zero,xmm1[15],zero
+; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm3 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero,xmm0[8],zero,xmm0[9],zero,xmm0[10],zero,xmm0[11],zero,xmm0[12],zero,xmm0[13],zero,xmm0[14],zero,xmm0[15],zero
+; AVX2-NEXT:    vpmullw %ymm1, %ymm3, %ymm1
+; AVX2-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm3 = ymm1[2,3],ymm2[2,3]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpackuswb %ymm3, %ymm1, %ymm1
+; AVX2-NEXT:    vpsubb %ymm1, %ymm0, %ymm0
+; AVX2-NEXT:    vpsrlw $1, %ymm0, %ymm0
+; AVX2-NEXT:    vpand {{.*}}(%rip), %ymm0, %ymm0
+; AVX2-NEXT:    vpaddb %ymm1, %ymm0, %ymm0
+; AVX2-NEXT:    vpsrlw $2, %ymm0, %ymm0
+; AVX2-NEXT:    vpand {{.*}}(%rip), %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %res = udiv <32 x i8> %a, <i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7>
   ret <32 x i8> %res
 }
 
 ;
 ; urem by 7
 ;

 ; AVX2-NEXT:    retq
   %res = urem <16 x i16> %a, <i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7, i16 7>
   ret <16 x i16> %res
 }
 
 define <32 x i8> @test_rem7_32i8(<32 x i8> %a) nounwind {
 ; AVX1-LABEL: test_rem7_32i8:
 ; AVX1:       # BB#0:
-; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm1
-; AVX1-NEXT:    vpextrb $1, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %ecx
-; AVX1-NEXT:    shrl $8, %ecx
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %cl, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %cl, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    movb $7, %cl
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %edx
-; AVX1-NEXT:    vpextrb $0, %xmm1, %esi
-; AVX1-NEXT:    imull $37, %esi, %edi
-; AVX1-NEXT:    shrl $8, %edi
-; AVX1-NEXT:    movb %sil, %al
-; AVX1-NEXT:    subb %dil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %dil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %sil
-; AVX1-NEXT:    movzbl %sil, %eax
-; AVX1-NEXT:    vmovd %eax, %xmm2
-; AVX1-NEXT:    vpinsrb $1, %edx, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $2, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $3, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $4, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $5, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $6, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $7, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $8, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $9, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $10, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $11, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $12, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $13, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $14, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $15, %xmm1, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm1
-; AVX1-NEXT:    vpextrb $1, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %edx
-; AVX1-NEXT:    vpextrb $0, %xmm0, %esi
-; AVX1-NEXT:    imull $37, %esi, %edi
-; AVX1-NEXT:    shrl $8, %edi
-; AVX1-NEXT:    movb %sil, %al
-; AVX1-NEXT:    subb %dil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %dil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %sil
-; AVX1-NEXT:    movzbl %sil, %eax
-; AVX1-NEXT:    vmovd %eax, %xmm2
-; AVX1-NEXT:    vpinsrb $1, %edx, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $2, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $3, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $4, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $5, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $6, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $7, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $8, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $9, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $10, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $11, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $12, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $13, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $14, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX1-NEXT:    vpextrb $15, %xmm0, %edx
-; AVX1-NEXT:    imull $37, %edx, %esi
-; AVX1-NEXT:    shrl $8, %esi
-; AVX1-NEXT:    movb %dl, %al
-; AVX1-NEXT:    subb %sil, %al
-; AVX1-NEXT:    shrb %al
-; AVX1-NEXT:    addb %sil, %al
-; AVX1-NEXT:    shrb $2, %al
-; AVX1-NEXT:    mulb %cl
-; AVX1-NEXT:    subb %al, %dl
-; AVX1-NEXT:    movzbl %dl, %eax
-; AVX1-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm0
-; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    vextractf128 $1, %ymm0, %xmm2
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm3 = xmm2[0],zero,xmm2[1],zero,xmm2[2],zero,xmm2[3],zero,xmm2[4],zero,xmm2[5],zero,xmm2[6],zero,xmm2[7],zero
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm1 = mem[0],zero,mem[1],zero,mem[2],zero,mem[3],zero,mem[4],zero,mem[5],zero,mem[6],zero,mem[7],zero
+; AVX1-NEXT:    vpmullw %xmm1, %xmm3, %xmm3
+; AVX1-NEXT:    vpsrlw $8, %xmm3, %xmm3
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm4 = xmm2[2,3,0,1]
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm4 = xmm4[0],zero,xmm4[1],zero,xmm4[2],zero,xmm4[3],zero,xmm4[4],zero,xmm4[5],zero,xmm4[6],zero,xmm4[7],zero
+; AVX1-NEXT:    vpmullw %xmm1, %xmm4, %xmm4
+; AVX1-NEXT:    vpsrlw $8, %xmm4, %xmm4
+; AVX1-NEXT:    vpackuswb %xmm4, %xmm3, %xmm3
+; AVX1-NEXT:    vpsubb %xmm3, %xmm2, %xmm4
+; AVX1-NEXT:    vpsrlw $1, %xmm4, %xmm4
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm8 = [127,127,127,127,127,127,127,127,127,127,127,127,127,127,127,127]
+; AVX1-NEXT:    vpand %xmm8, %xmm4, %xmm4
+; AVX1-NEXT:    vpaddb %xmm3, %xmm4, %xmm3
+; AVX1-NEXT:    vpsrlw $2, %xmm3, %xmm3
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm4 = [63,63,63,63,63,63,63,63,63,63,63,63,63,63,63,63]
+; AVX1-NEXT:    vpand %xmm4, %xmm3, %xmm3
+; AVX1-NEXT:    vpmovsxbw %xmm3, %xmm6
+; AVX1-NEXT:    vpmovsxbw {{.*}}(%rip), %xmm7
+; AVX1-NEXT:    vpmullw %xmm7, %xmm6, %xmm6
+; AVX1-NEXT:    vmovdqa {{.*#+}} xmm5 = [255,255,255,255,255,255,255,255]
+; AVX1-NEXT:    vpand %xmm5, %xmm6, %xmm6
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm3 = xmm3[2,3,0,1]
+; AVX1-NEXT:    vpmovsxbw %xmm3, %xmm3
+; AVX1-NEXT:    vpmullw %xmm7, %xmm3, %xmm3
+; AVX1-NEXT:    vpand %xmm5, %xmm3, %xmm3
+; AVX1-NEXT:    vpackuswb %xmm3, %xmm6, %xmm3
+; AVX1-NEXT:    vpsubb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm3 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero
+; AVX1-NEXT:    vpmullw %xmm1, %xmm3, %xmm3
+; AVX1-NEXT:    vpsrlw $8, %xmm3, %xmm3
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm6 = xmm0[2,3,0,1]
+; AVX1-NEXT:    vpmovzxbw {{.*#+}} xmm6 = xmm6[0],zero,xmm6[1],zero,xmm6[2],zero,xmm6[3],zero,xmm6[4],zero,xmm6[5],zero,xmm6[6],zero,xmm6[7],zero
+; AVX1-NEXT:    vpmullw %xmm1, %xmm6, %xmm1
+; AVX1-NEXT:    vpsrlw $8, %xmm1, %xmm1
+; AVX1-NEXT:    vpackuswb %xmm1, %xmm3, %xmm1
+; AVX1-NEXT:    vpsubb %xmm1, %xmm0, %xmm3
+; AVX1-NEXT:    vpsrlw $1, %xmm3, %xmm3
+; AVX1-NEXT:    vpand %xmm8, %xmm3, %xmm3
+; AVX1-NEXT:    vpaddb %xmm1, %xmm3, %xmm1
+; AVX1-NEXT:    vpsrlw $2, %xmm1, %xmm1
+; AVX1-NEXT:    vpand %xmm4, %xmm1, %xmm1
+; AVX1-NEXT:    vpmovsxbw %xmm1, %xmm3
+; AVX1-NEXT:    vpmullw %xmm7, %xmm3, %xmm3
+; AVX1-NEXT:    vpand %xmm5, %xmm3, %xmm3
+; AVX1-NEXT:    vpshufd {{.*#+}} xmm1 = xmm1[2,3,0,1]
+; AVX1-NEXT:    vpmovsxbw %xmm1, %xmm1
+; AVX1-NEXT:    vpmullw %xmm7, %xmm1, %xmm1
+; AVX1-NEXT:    vpand %xmm5, %xmm1, %xmm1
+; AVX1-NEXT:    vpackuswb %xmm1, %xmm3, %xmm1
+; AVX1-NEXT:    vpsubb %xmm1, %xmm0, %xmm0
+; AVX1-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
 ; AVX1-NEXT:    retq
 ;
 ; AVX2-LABEL: test_rem7_32i8:
 ; AVX2:       # BB#0:
-; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm1
-; AVX2-NEXT:    vpextrb $1, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %ecx
-; AVX2-NEXT:    shrl $8, %ecx
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %cl, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %cl, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    movb $7, %cl
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %edx
-; AVX2-NEXT:    vpextrb $0, %xmm1, %esi
-; AVX2-NEXT:    imull $37, %esi, %edi
-; AVX2-NEXT:    shrl $8, %edi
-; AVX2-NEXT:    movb %sil, %al
-; AVX2-NEXT:    subb %dil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %dil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %sil
-; AVX2-NEXT:    movzbl %sil, %eax
-; AVX2-NEXT:    vmovd %eax, %xmm2
-; AVX2-NEXT:    vpinsrb $1, %edx, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $2, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $3, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $4, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $5, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $6, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $7, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $8, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $9, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $10, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $11, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $12, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $13, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $14, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $15, %xmm1, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm1
-; AVX2-NEXT:    vpextrb $1, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %edx
-; AVX2-NEXT:    vpextrb $0, %xmm0, %esi
-; AVX2-NEXT:    imull $37, %esi, %edi
-; AVX2-NEXT:    shrl $8, %edi
-; AVX2-NEXT:    movb %sil, %al
-; AVX2-NEXT:    subb %dil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %dil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %sil
-; AVX2-NEXT:    movzbl %sil, %eax
-; AVX2-NEXT:    vmovd %eax, %xmm2
-; AVX2-NEXT:    vpinsrb $1, %edx, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $2, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $2, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $3, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $3, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $4, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $4, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $5, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $5, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $6, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $6, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $7, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $7, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $8, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $8, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $9, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $9, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $10, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $10, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $11, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $11, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $12, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $12, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $13, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $13, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $14, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $14, %eax, %xmm2, %xmm2
-; AVX2-NEXT:    vpextrb $15, %xmm0, %edx
-; AVX2-NEXT:    imull $37, %edx, %esi
-; AVX2-NEXT:    shrl $8, %esi
-; AVX2-NEXT:    movb %dl, %al
-; AVX2-NEXT:    subb %sil, %al
-; AVX2-NEXT:    shrb %al
-; AVX2-NEXT:    addb %sil, %al
-; AVX2-NEXT:    shrb $2, %al
-; AVX2-NEXT:    mulb %cl
-; AVX2-NEXT:    subb %al, %dl
-; AVX2-NEXT:    movzbl %dl, %eax
-; AVX2-NEXT:    vpinsrb $15, %eax, %xmm2, %xmm0
-; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT:    vmovdqa {{.*#+}} ymm1 = [37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37,37]
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm2 = xmm2[0],zero,xmm2[1],zero,xmm2[2],zero,xmm2[3],zero,xmm2[4],zero,xmm2[5],zero,xmm2[6],zero,xmm2[7],zero,xmm2[8],zero,xmm2[9],zero,xmm2[10],zero,xmm2[11],zero,xmm2[12],zero,xmm2[13],zero,xmm2[14],zero,xmm2[15],zero
+; AVX2-NEXT:    vextracti128 $1, %ymm0, %xmm3
+; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm3 = xmm3[0],zero,xmm3[1],zero,xmm3[2],zero,xmm3[3],zero,xmm3[4],zero,xmm3[5],zero,xmm3[6],zero,xmm3[7],zero,xmm3[8],zero,xmm3[9],zero,xmm3[10],zero,xmm3[11],zero,xmm3[12],zero,xmm3[13],zero,xmm3[14],zero,xmm3[15],zero
+; AVX2-NEXT:    vpmullw %ymm2, %ymm3, %ymm2
+; AVX2-NEXT:    vpsrlw $8, %ymm2, %ymm2
+; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm1 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero,xmm1[8],zero,xmm1[9],zero,xmm1[10],zero,xmm1[11],zero,xmm1[12],zero,xmm1[13],zero,xmm1[14],zero,xmm1[15],zero
+; AVX2-NEXT:    vpmovzxbw {{.*#+}} ymm3 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero,xmm0[8],zero,xmm0[9],zero,xmm0[10],zero,xmm0[11],zero,xmm0[12],zero,xmm0[13],zero,xmm0[14],zero,xmm0[15],zero
+; AVX2-NEXT:    vpmullw %ymm1, %ymm3, %ymm1
+; AVX2-NEXT:    vpsrlw $8, %ymm1, %ymm1
+; AVX2-NEXT:    vperm2i128 {{.*#+}} ymm3 = ymm1[2,3],ymm2[2,3]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpackuswb %ymm3, %ymm1, %ymm1
+; AVX2-NEXT:    vpsubb %ymm1, %ymm0, %ymm2
+; AVX2-NEXT:    vpsrlw $1, %ymm2, %ymm2
+; AVX2-NEXT:    vpand {{.*}}(%rip), %ymm2, %ymm2
+; AVX2-NEXT:    vpaddb %ymm1, %ymm2, %ymm1
+; AVX2-NEXT:    vpsrlw $2, %ymm1, %ymm1
+; AVX2-NEXT:    vpand {{.*}}(%rip), %ymm1, %ymm1
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm2
+; AVX2-NEXT:    vpmovsxbw %xmm2, %ymm2
+; AVX2-NEXT:    vpmovsxbw {{.*}}(%rip), %ymm3
+; AVX2-NEXT:    vpmullw %ymm3, %ymm2, %ymm2
+; AVX2-NEXT:    vextracti128 $1, %ymm2, %xmm4
+; AVX2-NEXT:    vmovdqa {{.*#+}} xmm5 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u>
+; AVX2-NEXT:    vpshufb %xmm5, %xmm4, %xmm4
+; AVX2-NEXT:    vpshufb %xmm5, %xmm2, %xmm2
+; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm4[0]
+; AVX2-NEXT:    vpmovsxbw %xmm1, %ymm1
+; AVX2-NEXT:    vpmullw %ymm3, %ymm1, %ymm1
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm3
+; AVX2-NEXT:    vpshufb %xmm5, %xmm3, %xmm3
+; AVX2-NEXT:    vpshufb %xmm5, %xmm1, %xmm1
+; AVX2-NEXT:    vpunpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
+; AVX2-NEXT:    vinserti128 $1, %xmm2, %ymm1, %ymm1
+; AVX2-NEXT:    vpsubb %ymm1, %ymm0, %ymm0
 ; AVX2-NEXT:    retq
   %res = urem <32 x i8> %a, <i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7, i8 7,i8 7, i8 7, i8 7, i8 7>
   ret <32 x i8> %res
 }