Index: include/llvm/IR/IntrinsicsX86.td
===================================================================
--- include/llvm/IR/IntrinsicsX86.td
+++ include/llvm/IR/IntrinsicsX86.td
@@ -1609,6 +1609,25 @@
                   [IntrArgMemOnly]>;
 }
 
+// BITALG bits shuffle
+let TargetPrefix = "x86" in {  // All intrinsics start with "llvm.x86.".
+  def int_x86_avx512_mask_vpshufbitqmb_128 :
+    GCCBuiltin<"__builtin_ia32_vpshufbitqmb128_mask">,
+    Intrinsic<[llvm_i16_ty],
+              [llvm_v16i8_ty, llvm_v16i8_ty, llvm_i16_ty],
+              [IntrNoMem]>;
+  def int_x86_avx512_mask_vpshufbitqmb_256 :
+    GCCBuiltin<"__builtin_ia32_vpshufbitqmb256_mask">,
+    Intrinsic<[llvm_i32_ty],
+              [llvm_v32i8_ty, llvm_v32i8_ty, llvm_i32_ty],
+              [IntrNoMem]>;
+  def int_x86_avx512_mask_vpshufbitqmb_512 :
+    GCCBuiltin<"__builtin_ia32_vpshufbitqmb512_mask">,
+    Intrinsic<[llvm_i64_ty],
+              [llvm_v64i8_ty, llvm_v64i8_ty, llvm_i64_ty],
+              [IntrNoMem]>;
+}
+
 //===----------------------------------------------------------------------===//
 // AVX2
 
Index: lib/Support/Host.cpp
===================================================================
--- lib/Support/Host.cpp
+++ lib/Support/Host.cpp
@@ -1476,6 +1476,9 @@
   // Enable protection keys
   Features["pku"] = HasLeaf7 && ((ECX >> 4) & 1);
 
+  // Enable Bit Algorithms
+  Features["avx512bitalg"] = HasLeaf7 && ((ECX >> 12) & 1) && HasAVX512Save;
+
   bool HasLeafD = MaxLevel >= 0xd &&
                   !getX86CpuIDAndInfoEx(0xd, 0x1, &EAX, &EBX, &ECX, &EDX);
 
Index: lib/Target/X86/X86.td
===================================================================
--- lib/Target/X86/X86.td
+++ lib/Target/X86/X86.td
@@ -151,6 +151,9 @@
                                       [FeatureAVX512]>;
 def FeaturePKU   : SubtargetFeature<"pku", "HasPKU", "true",
                       "Enable protection keys">;
+def FeatureBITALG  : SubtargetFeature<"avx512bitalg", "HasBITALG", "true",
+                       "Enable AVX-512 Bit Algorithms",
+                        [FeatureBWI]>;
 def FeaturePCLMUL  : SubtargetFeature<"pclmul", "HasPCLMUL", "true",
                          "Enable packed carry-less multiplication instructions",
                                [FeatureSSE2]>;
Index: lib/Target/X86/X86ISelLowering.h
===================================================================
--- lib/Target/X86/X86ISelLowering.h
+++ lib/Target/X86/X86ISelLowering.h
@@ -498,6 +498,9 @@
       COMPRESS,
       EXPAND,
 
+      // Bits shuffle
+      VPSHUFBITQMB,
+
       // Convert Unsigned/Integer to Floating-Point Value with rounding mode.
       SINT_TO_FP_RND, UINT_TO_FP_RND,
       SCALAR_SINT_TO_FP_RND, SCALAR_UINT_TO_FP_RND,
Index: lib/Target/X86/X86ISelLowering.cpp
===================================================================
--- lib/Target/X86/X86ISelLowering.cpp
+++ lib/Target/X86/X86ISelLowering.cpp
@@ -1548,6 +1548,11 @@
     }
   }
 
+  if (Subtarget.hasBITALG())
+    for (auto VT : { MVT::v64i8, MVT::v32i16, MVT::v32i8,
+                     MVT::v16i16, MVT::v16i8, MVT::v8i16 })
+      setOperationAction(ISD::CTPOP, VT, Legal);
+
   // We want to custom lower some of our intrinsics.
   setOperationAction(ISD::INTRINSIC_WO_CHAIN, MVT::Other, Custom);
   setOperationAction(ISD::INTRINSIC_W_CHAIN, MVT::Other, Custom);
@@ -24907,6 +24912,7 @@
   case X86ISD::CVTS2UI_RND:        return "X86ISD::CVTS2UI_RND";
   case X86ISD::LWPINS:             return "X86ISD::LWPINS";
   case X86ISD::MGATHER:            return "X86ISD::MGATHER";
+  case X86ISD::VPSHUFBITQMB:       return "X86ISD::VPSHUFBITQMB";
   }
   return nullptr;
 }
Index: lib/Target/X86/X86InstrAVX512.td
===================================================================
--- lib/Target/X86/X86InstrAVX512.td
+++ lib/Target/X86/X86InstrAVX512.td
@@ -9804,3 +9804,42 @@
 defm : AVX512_scalar_math_f64_patterns<fsub, "SUB">;
 defm : AVX512_scalar_math_f64_patterns<fmul, "MUL">;
 defm : AVX512_scalar_math_f64_patterns<fdiv, "DIV">;
+
+//===----------------------------------------------------------------------===//
+// Bit Algorithms
+//===----------------------------------------------------------------------===//
+
+defm VPOPCNTB : avx512_unary_rm_vl<0x54, "vpopcntb", ctpop,
+                                   avx512vl_i8_info, HasBITALG>,
+                avx512_unary_lowering<ctpop, avx512vl_i8_info, HasBITALG>;
+defm VPOPCNTW : avx512_unary_rm_vl<0x54, "vpopcntw", ctpop,
+                                   avx512vl_i16_info, HasBITALG>,
+                avx512_unary_lowering<ctpop, avx512vl_i16_info, HasBITALG>, VEX_W;
+
+multiclass VPSHUFBITQMB_rm<X86VectorVTInfo VTI> {
+  defm rr : AVX512_maskable_cmp<0x8F, MRMSrcReg, VTI, (outs VTI.KRC:$dst),
+                                (ins VTI.RC:$src1, VTI.RC:$src2),
+                                "vpshufbitqmb",
+                                "$src2, $src1", "$src1, $src2",
+                                (X86Vpshufbitqmb (VTI.VT VTI.RC:$src1),
+                                 (VTI.VT VTI.RC:$src2))>, EVEX_4V, T8PD;
+  defm rm : AVX512_maskable_cmp<0x8F, MRMSrcMem, VTI, (outs VTI.KRC:$dst),
+                                (ins VTI.RC:$src1, VTI.MemOp:$src2),
+                                "vpshufbitqmb",
+                                "$src2, $src1", "$src1, $src2",
+                                (X86Vpshufbitqmb (VTI.VT VTI.RC:$src1),
+                                 (VTI.VT (bitconvert (VTI.LdFrag addr:$src2))))>,
+                                EVEX_4V, EVEX_CD8<8, CD8VF>, T8PD;
+}
+
+multiclass VPSHUFBITQMB_common<AVX512VLVectorVTInfo VTI> {
+  let Predicates = [HasBITALG] in
+  defm Z      : VPSHUFBITQMB_rm<VTI.info512>, EVEX_V512;
+  let Predicates = [HasBITALG, HasVLX] in {
+    defm Z256 : VPSHUFBITQMB_rm<VTI.info256>, EVEX_V256;
+    defm Z128 : VPSHUFBITQMB_rm<VTI.info128>, EVEX_V128;
+  }
+}
+
+defm VPSHUFBITQMB : VPSHUFBITQMB_common<avx512vl_i8_info>;
+
Index: lib/Target/X86/X86InstrFragmentsSIMD.td
===================================================================
--- lib/Target/X86/X86InstrFragmentsSIMD.td
+++ lib/Target/X86/X86InstrFragmentsSIMD.td
@@ -524,6 +524,13 @@
 def X86expand  : SDNode<"X86ISD::EXPAND", SDTypeProfile<1, 1,
                               [SDTCisSameAs<0, 1>, SDTCisVec<1>]>, []>;
 
+// vpshufbitqmb
+def X86Vpshufbitqmb : SDNode<"X86ISD::VPSHUFBITQMB",
+                             SDTypeProfile<1, 2, [SDTCisVec<0>, SDTCisVec<1>,
+                                                  SDTCisSameAs<1,2>,
+                                                  SDTCVecEltisVT<0,i1>,
+                                                  SDTCisSameNumEltsAs<0,1>]>>;
+
 def SDTintToFPRound: SDTypeProfile<1, 3, [SDTCisVec<0>, SDTCisFP<0>,
                                           SDTCisSameAs<0,1>, SDTCisInt<2>,
                                           SDTCisVT<3, i32>]>;
Index: lib/Target/X86/X86InstrInfo.td
===================================================================
--- lib/Target/X86/X86InstrInfo.td
+++ lib/Target/X86/X86InstrInfo.td
@@ -823,6 +823,7 @@
 def NoVLX_Or_NoDQI : Predicate<"!Subtarget->hasVLX() || !Subtarget->hasDQI()">;
 def PKU        : Predicate<"Subtarget->hasPKU()">;
 
+def HasBITALG    : Predicate<"Subtarget->hasBITALG()">;
 def HasPOPCNT    : Predicate<"Subtarget->hasPOPCNT()">;
 def HasAES       : Predicate<"Subtarget->hasAES()">;
 def HasFXSR      : Predicate<"Subtarget->hasFXSR()">;
Index: lib/Target/X86/X86IntrinsicsInfo.h
===================================================================
--- lib/Target/X86/X86IntrinsicsInfo.h
+++ lib/Target/X86/X86IntrinsicsInfo.h
@@ -1211,6 +1211,12 @@
                      X86ISD::VPMADD52L, 0),
   X86_INTRINSIC_DATA(avx512_mask_vpmadd52l_uq_512 , IFMA_OP_MASK,
                      X86ISD::VPMADD52L, 0),
+  X86_INTRINSIC_DATA(avx512_mask_vpshufbitqmb_128, CMP_MASK,
+                     X86ISD::VPSHUFBITQMB, 0),
+  X86_INTRINSIC_DATA(avx512_mask_vpshufbitqmb_256, CMP_MASK,
+                     X86ISD::VPSHUFBITQMB, 0),
+  X86_INTRINSIC_DATA(avx512_mask_vpshufbitqmb_512, CMP_MASK,
+                     X86ISD::VPSHUFBITQMB, 0),
   X86_INTRINSIC_DATA(avx512_mask3_vfmadd_pd_128, FMA_OP_MASK3, ISD::FMA, 0),
   X86_INTRINSIC_DATA(avx512_mask3_vfmadd_pd_256, FMA_OP_MASK3, ISD::FMA, 0),
   X86_INTRINSIC_DATA(avx512_mask3_vfmadd_pd_512, FMA_OP_MASK3, ISD::FMA,
Index: lib/Target/X86/X86Subtarget.h
===================================================================
--- lib/Target/X86/X86Subtarget.h
+++ lib/Target/X86/X86Subtarget.h
@@ -297,6 +297,9 @@
   /// Processor has PKU extenstions
   bool HasPKU;
 
+  /// Processor has AVX-512 Bit Algorithms instructions
+  bool HasBITALG;
+
   /// Processor supports MPX - Memory Protection Extensions
   bool HasMPX;
 
@@ -520,6 +523,7 @@
   bool hasBWI() const { return HasBWI; }
   bool hasVLX() const { return HasVLX; }
   bool hasPKU() const { return HasPKU; }
+  bool hasBITALG() const { return HasBITALG; }
   bool hasMPX() const { return HasMPX; }
   bool hasCLFLUSHOPT() const { return HasCLFLUSHOPT; }
   bool hasCLWB() const { return HasCLWB; }
Index: lib/Target/X86/X86Subtarget.cpp
===================================================================
--- lib/Target/X86/X86Subtarget.cpp
+++ lib/Target/X86/X86Subtarget.cpp
@@ -319,6 +319,7 @@
   HasVLX = false;
   HasADX = false;
   HasPKU = false;
+  HasBITALG = false;
   HasSHA = false;
   HasPRFCHW = false;
   HasRDSEED = false;
Index: test/CodeGen/X86/vector-popcnt-128.ll
===================================================================
--- test/CodeGen/X86/vector-popcnt-128.ll
+++ test/CodeGen/X86/vector-popcnt-128.ll
@@ -6,6 +6,8 @@
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX1
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX2
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512vpopcntdq | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX512VPOPCNTDQ
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512bitalg | FileCheck %s --check-prefix=ALL --check-prefix=BITALG_NOVLX
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512bitalg,+avx512vl | FileCheck %s --check-prefix=ALL --check-prefix=BITALG
 
 define <2 x i64> @testv2i64(<2 x i64> %in) nounwind {
 ; SSE2-LABEL: testv2i64:
@@ -381,6 +383,19 @@
 ; AVX512VPOPCNTDQ-NEXT:    vpmovqw %zmm0, %xmm0
 ; AVX512VPOPCNTDQ-NEXT:    vzeroupper
 ; AVX512VPOPCNTDQ-NEXT:    retq
+;
+; BITALG_NOVLX-LABEL: testv8i16:
+; BITALG_NOVLX:       # BB#0:
+; BITALG_NOVLX-NEXT:    # kill: %XMM0<def> %XMM0<kill> %ZMM0<def>
+; BITALG_NOVLX-NEXT:    vpopcntw %zmm0, %zmm0
+; BITALG_NOVLX-NEXT:    # kill: %XMM0<def> %XMM0<kill> %ZMM0<kill>
+; BITALG_NOVLX-NEXT:    vzeroupper
+; BITALG_NOVLX-NEXT:    retq
+;
+; BITALG-LABEL: testv8i16:
+; BITALG:       # BB#0:
+; BITALG-NEXT:    vpopcntw %xmm0, %xmm0
+; BITALG-NEXT:    retq
   %out = call <8 x i16> @llvm.ctpop.v8i16(<8 x i16> %in)
   ret <8 x i16> %out
 }
@@ -485,6 +500,19 @@
 ; AVX512VPOPCNTDQ-NEXT:    vpmovdb %zmm0, %xmm0
 ; AVX512VPOPCNTDQ-NEXT:    vzeroupper
 ; AVX512VPOPCNTDQ-NEXT:    retq
+;
+; BITALG_NOVLX-LABEL: testv16i8:
+; BITALG_NOVLX:       # BB#0:
+; BITALG_NOVLX-NEXT:    # kill: %XMM0<def> %XMM0<kill> %ZMM0<def>
+; BITALG_NOVLX-NEXT:    vpopcntb %zmm0, %zmm0
+; BITALG_NOVLX-NEXT:    # kill: %XMM0<def> %XMM0<kill> %ZMM0<kill>
+; BITALG_NOVLX-NEXT:    vzeroupper
+; BITALG_NOVLX-NEXT:    retq
+;
+; BITALG-LABEL: testv16i8:
+; BITALG:       # BB#0:
+; BITALG-NEXT:    vpopcntb %xmm0, %xmm0
+; BITALG-NEXT:    retq
   %out = call <16 x i8> @llvm.ctpop.v16i8(<16 x i8> %in)
   ret <16 x i8> %out
 }
Index: test/CodeGen/X86/vector-popcnt-256.ll
===================================================================
--- test/CodeGen/X86/vector-popcnt-256.ll
+++ test/CodeGen/X86/vector-popcnt-256.ll
@@ -2,6 +2,8 @@
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX1
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX2
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512vpopcntdq | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX512VPOPCNTDQ
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512bitalg | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=BITALG_NOVLX
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512bitalg,+avx512vl | FileCheck %s --check-prefix=ALL --check-prefix=BITALG
 
 define <4 x i64> @testv4i64(<4 x i64> %in) nounwind {
 ; AVX1-LABEL: testv4i64:
@@ -159,6 +161,18 @@
 ; AVX512VPOPCNTDQ-NEXT:    vpopcntd %zmm0, %zmm0
 ; AVX512VPOPCNTDQ-NEXT:    vpmovdw %zmm0, %ymm0
 ; AVX512VPOPCNTDQ-NEXT:    retq
+;
+; BITALG_NOVLX-LABEL: testv16i16:
+; BITALG_NOVLX:       # BB#0:
+; BITALG_NOVLX-NEXT:    # kill: %YMM0<def> %YMM0<kill> %ZMM0<def>
+; BITALG_NOVLX-NEXT:    vpopcntw %zmm0, %zmm0
+; BITALG_NOVLX-NEXT:    # kill: %YMM0<def> %YMM0<kill> %ZMM0<kill>
+; BITALG_NOVLX-NEXT:    retq
+;
+; BITALG-LABEL: testv16i16:
+; BITALG:       # BB#0:
+; BITALG-NEXT:    vpopcntw %ymm0, %ymm0
+; BITALG-NEXT:    retq
   %out = call <16 x i16> @llvm.ctpop.v16i16(<16 x i16> %in)
   ret <16 x i16> %out
 }
@@ -207,6 +221,18 @@
 ; AVX512VPOPCNTDQ-NEXT:    vpshufb %ymm0, %ymm3, %ymm0
 ; AVX512VPOPCNTDQ-NEXT:    vpaddb %ymm2, %ymm0, %ymm0
 ; AVX512VPOPCNTDQ-NEXT:    retq
+;
+; BITALG_NOVLX-LABEL: testv32i8:
+; BITALG_NOVLX:       # BB#0:
+; BITALG_NOVLX-NEXT:    # kill: %YMM0<def> %YMM0<kill> %ZMM0<def>
+; BITALG_NOVLX-NEXT:    vpopcntb %zmm0, %zmm0
+; BITALG_NOVLX-NEXT:    # kill: %YMM0<def> %YMM0<kill> %ZMM0<kill>
+; BITALG_NOVLX-NEXT:    retq
+;
+; BITALG-LABEL: testv32i8:
+; BITALG:       # BB#0:
+; BITALG-NEXT:    vpopcntb %ymm0, %ymm0
+; BITALG-NEXT:    retq
   %out = call <32 x i8> @llvm.ctpop.v32i8(<32 x i8> %in)
   ret <32 x i8> %out
 }
Index: test/CodeGen/X86/vector-popcnt-512.ll
===================================================================
--- test/CodeGen/X86/vector-popcnt-512.ll
+++ test/CodeGen/X86/vector-popcnt-512.ll
@@ -3,6 +3,7 @@
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512bw | FileCheck %s --check-prefix=AVX512 --check-prefix=AVX512BW
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512vpopcntdq | FileCheck %s --check-prefix=AVX512 --check-prefix=AVX512VPOPCNTDQ --check-prefix=AVX512VPOPCNTDQ-NOBW
 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512vpopcntdq,+avx512bw | FileCheck %s --check-prefix=AVX512 --check-prefix=AVX512VPOPCNTDQ --check-prefix=AVX512VPOPCNTDQ-BW
+; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx512bitalg | FileCheck %s --check-prefix=AVX512 --check-prefix=BITALG
 
 define <8 x i64> @testv8i64(<8 x i64> %in) nounwind {
 ; AVX512F-LABEL: testv8i64:
@@ -172,6 +173,11 @@
 ; AVX512VPOPCNTDQ-BW-NEXT:    vpaddb %zmm0, %zmm1, %zmm0
 ; AVX512VPOPCNTDQ-BW-NEXT:    vpsrlw $8, %zmm0, %zmm0
 ; AVX512VPOPCNTDQ-BW-NEXT:    retq
+;
+; BITALG-LABEL: testv32i16:
+; BITALG:       # BB#0:
+; BITALG-NEXT:    vpopcntw %zmm0, %zmm0
+; BITALG-NEXT:    retq
   %out = call <32 x i16> @llvm.ctpop.v32i16(<32 x i16> %in)
   ret <32 x i16> %out
 }
@@ -236,6 +242,11 @@
 ; AVX512VPOPCNTDQ-BW-NEXT:    vpshufb %zmm0, %zmm3, %zmm0
 ; AVX512VPOPCNTDQ-BW-NEXT:    vpaddb %zmm2, %zmm0, %zmm0
 ; AVX512VPOPCNTDQ-BW-NEXT:    retq
+;
+; BITALG-LABEL: testv64i8:
+; BITALG:       # BB#0:
+; BITALG-NEXT:    vpopcntb %zmm0, %zmm0
+; BITALG-NEXT:    retq
   %out = call <64 x i8> @llvm.ctpop.v64i8(<64 x i8> %in)
   ret <64 x i8> %out
 }
Index: test/CodeGen/X86/vpshufbitqmb-intrinsics.ll
===================================================================
--- /dev/null
+++ test/CodeGen/X86/vpshufbitqmb-intrinsics.ll
@@ -0,0 +1,41 @@
+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mattr=+avx512bitalg,+avx512vl | FileCheck %s
+
+declare i16 @llvm.x86.avx512.mask.vpshufbitqmb.128(<16 x i8> %a, <16 x i8> %b, i16 %mask)
+define i16 @test_vpshufbitqmb_128(<16 x i8> %a, <16 x i8> %b, i16 %mask) {
+; CHECK-LABEL: test_vpshufbitqmb_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    vpshufbitqmb %xmm1, %xmm0, %k0 {%k1}
+; CHECK-NEXT:    kmovd %k0, %eax
+; CHECK-NEXT:    ## kill: %AX<def> %AX<kill> %EAX<kill>
+; CHECK-NEXT:    retq
+  %res = call i16 @llvm.x86.avx512.mask.vpshufbitqmb.128(<16 x i8> %a, <16 x i8> %b, i16 %mask)
+  ret i16 %res
+}
+
+declare i32 @llvm.x86.avx512.mask.vpshufbitqmb.256(<32 x i8> %a, <32 x i8> %b, i32 %mask)
+define i32 @test_vpshufbitqmb_256(<32 x i8> %a, <32 x i8> %b, i32 %mask) {
+; CHECK-LABEL: test_vpshufbitqmb_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    vpshufbitqmb %ymm1, %ymm0, %k0 {%k1}
+; CHECK-NEXT:    kmovd %k0, %eax
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %res = call i32 @llvm.x86.avx512.mask.vpshufbitqmb.256(<32 x i8> %a, <32 x i8> %b, i32 %mask)
+  ret i32 %res
+}
+
+declare i64 @llvm.x86.avx512.mask.vpshufbitqmb.512(<64 x i8> %a, <64 x i8> %b, i64 %mask)
+define i64 @test_vpshufbitqmb_512(<64 x i8> %a, <64 x i8> %b, i64 %mask) {
+; CHECK-LABEL: test_vpshufbitqmb_512:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    kmovq %rdi, %k1
+; CHECK-NEXT:    vpshufbitqmb %zmm1, %zmm0, %k0 {%k1}
+; CHECK-NEXT:    kmovq %k0, %rax
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %res = call i64 @llvm.x86.avx512.mask.vpshufbitqmb.512(<64 x i8> %a, <64 x i8> %b, i64 %mask)
+  ret i64 %res
+}
Index: test/MC/X86/avx512bitalg-encoding.s
===================================================================
--- /dev/null
+++ test/MC/X86/avx512bitalg-encoding.s
@@ -0,0 +1,170 @@
+// RUN: llvm-mc -triple x86_64-unknown-unknown -mcpu=knl -mattr=+avx512bitalg --show-encoding < %s | FileCheck %s
+
+// CHECK: vpopcntb %zmm23, %zmm21
+// CHECK: encoding: [0x62,0xa2,0x7d,0x48,0x54,0xef]
+          vpopcntb %zmm23, %zmm21
+
+// CHECK: vpopcntw %zmm23, %zmm21
+// CHECK: encoding: [0x62,0xa2,0xfd,0x48,0x54,0xef]
+          vpopcntw %zmm23, %zmm21
+
+// CHECK: vpopcntb %zmm3, %zmm1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x7d,0x4a,0x54,0xcb]
+          vpopcntb %zmm3, %zmm1 {%k2}
+
+// CHECK: vpopcntw %zmm3, %zmm1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0xfd,0x4a,0x54,0xcb]
+          vpopcntw %zmm3, %zmm1 {%k2}
+
+// CHECK: vpopcntb  (%rcx), %zmm1
+// CHECK: encoding: [0x62,0xf2,0x7d,0x48,0x54,0x09]
+          vpopcntb  (%rcx), %zmm1
+
+// CHECK: vpopcntb  -256(%rsp), %zmm1
+// CHECK: encoding: [0x62,0xf2,0x7d,0x48,0x54,0x4c,0x24,0xfc]
+          vpopcntb  -256(%rsp), %zmm1
+
+// CHECK: vpopcntb  256(%rsp), %zmm1
+// CHECK: encoding: [0x62,0xf2,0x7d,0x48,0x54,0x4c,0x24,0x04]
+          vpopcntb  256(%rsp), %zmm1
+
+// CHECK: vpopcntb  268435456(%rcx,%r14,8), %zmm1
+// CHECK: encoding: [0x62,0xb2,0x7d,0x48,0x54,0x8c,0xf1,0x00,0x00,0x00,0x10]
+          vpopcntb  268435456(%rcx,%r14,8), %zmm1
+
+// CHECK: vpopcntb  -536870912(%rcx,%r14,8), %zmm1
+// CHECK: encoding: [0x62,0xb2,0x7d,0x48,0x54,0x8c,0xf1,0x00,0x00,0x00,0xe0]
+          vpopcntb  -536870912(%rcx,%r14,8), %zmm1
+
+// CHECK: vpopcntb  -536870910(%rcx,%r14,8), %zmm1
+// CHECK: encoding: [0x62,0xb2,0x7d,0x48,0x54,0x8c,0xf1,0x02,0x00,0x00,0xe0]
+          vpopcntb  -536870910(%rcx,%r14,8), %zmm1
+
+// CHECK: vpopcntw  (%rcx), %zmm1
+// CHECK: encoding: [0x62,0xf2,0xfd,0x48,0x54,0x09]
+          vpopcntw  (%rcx), %zmm1
+
+// CHECK: vpopcntw  -256(%rsp), %zmm1
+// CHECK: encoding: [0x62,0xf2,0xfd,0x48,0x54,0x4c,0x24,0xfc]
+          vpopcntw  -256(%rsp), %zmm1
+
+// CHECK: vpopcntw  256(%rsp), %zmm1
+// CHECK: encoding: [0x62,0xf2,0xfd,0x48,0x54,0x4c,0x24,0x04]
+          vpopcntw  256(%rsp), %zmm1
+
+// CHECK: vpopcntw  268435456(%rcx,%r14,8), %zmm1
+// CHECK: encoding: [0x62,0xb2,0xfd,0x48,0x54,0x8c,0xf1,0x00,0x00,0x00,0x10]
+          vpopcntw  268435456(%rcx,%r14,8), %zmm1
+
+// CHECK: vpopcntw  -536870912(%rcx,%r14,8), %zmm1
+// CHECK: encoding: [0x62,0xb2,0xfd,0x48,0x54,0x8c,0xf1,0x00,0x00,0x00,0xe0]
+          vpopcntw  -536870912(%rcx,%r14,8), %zmm1
+
+// CHECK: vpopcntw  -536870910(%rcx,%r14,8), %zmm1
+// CHECK: encoding: [0x62,0xb2,0xfd,0x48,0x54,0x8c,0xf1,0x02,0x00,0x00,0xe0]
+          vpopcntw  -536870910(%rcx,%r14,8), %zmm1
+
+// CHECK: vpopcntb  (%rcx), %zmm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0x7d,0x4a,0x54,0x29]
+          vpopcntb  (%rcx), %zmm21 {%k2}
+
+// CHECK: vpopcntb  -256(%rsp), %zmm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0x7d,0x4a,0x54,0x6c,0x24,0xfc]
+          vpopcntb  -256(%rsp), %zmm21 {%k2}
+
+// CHECK: vpopcntb  256(%rsp), %zmm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0x7d,0x4a,0x54,0x6c,0x24,0x04]
+          vpopcntb  256(%rsp), %zmm21 {%k2}
+
+// CHECK: vpopcntb  268435456(%rcx,%r14,8), %zmm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0x7d,0x4a,0x54,0xac,0xf1,0x00,0x00,0x00,0x10]
+          vpopcntb  268435456(%rcx,%r14,8), %zmm21 {%k2}
+
+// CHECK: vpopcntb  -536870912(%rcx,%r14,8), %zmm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0x7d,0x4a,0x54,0xac,0xf1,0x00,0x00,0x00,0xe0]
+          vpopcntb  -536870912(%rcx,%r14,8), %zmm21 {%k2}
+
+// CHECK: vpopcntb  -536870910(%rcx,%r14,8), %zmm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0x7d,0x4a,0x54,0xac,0xf1,0x02,0x00,0x00,0xe0]
+          vpopcntb  -536870910(%rcx,%r14,8), %zmm21 {%k2}
+
+// CHECK: vpopcntw  (%rcx), %zmm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0xfd,0x4a,0x54,0x29]
+          vpopcntw  (%rcx), %zmm21 {%k2}
+
+// CHECK: vpopcntw  -256(%rsp), %zmm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0xfd,0x4a,0x54,0x6c,0x24,0xfc]
+          vpopcntw  -256(%rsp), %zmm21 {%k2}
+
+// CHECK: vpopcntw  256(%rsp), %zmm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0xfd,0x4a,0x54,0x6c,0x24,0x04]
+          vpopcntw  256(%rsp), %zmm21 {%k2}
+
+// CHECK: vpopcntw  268435456(%rcx,%r14,8), %zmm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0xfd,0x4a,0x54,0xac,0xf1,0x00,0x00,0x00,0x10]
+          vpopcntw  268435456(%rcx,%r14,8), %zmm21 {%k2}
+
+// CHECK: vpopcntw  -536870912(%rcx,%r14,8), %zmm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0xfd,0x4a,0x54,0xac,0xf1,0x00,0x00,0x00,0xe0]
+          vpopcntw  -536870912(%rcx,%r14,8), %zmm21 {%k2}
+
+// CHECK: vpopcntw  -536870910(%rcx,%r14,8), %zmm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0xfd,0x4a,0x54,0xac,0xf1,0x02,0x00,0x00,0xe0]
+          vpopcntw  -536870910(%rcx,%r14,8), %zmm21 {%k2}
+
+// CHECK: vpshufbitqmb %zmm2, %zmm23, %k1
+// CHECK: encoding: [0x62,0xf2,0x45,0x40,0x8f,0xca]
+          vpshufbitqmb %zmm2, %zmm23, %k1
+
+// CHECK: vpshufbitqmb %zmm2, %zmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x45,0x42,0x8f,0xca]
+          vpshufbitqmb %zmm2, %zmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  (%rcx), %zmm23, %k1
+// CHECK: encoding: [0x62,0xf2,0x45,0x40,0x8f,0x09]
+          vpshufbitqmb  (%rcx), %zmm23, %k1
+
+// CHECK: vpshufbitqmb  -256(%rsp), %zmm23, %k1
+// CHECK: encoding: [0x62,0xf2,0x45,0x40,0x8f,0x4c,0x24,0xfc]
+          vpshufbitqmb  -256(%rsp), %zmm23, %k1
+
+// CHECK: vpshufbitqmb  256(%rsp), %zmm23, %k1
+// CHECK: encoding: [0x62,0xf2,0x45,0x40,0x8f,0x4c,0x24,0x04]
+          vpshufbitqmb  256(%rsp), %zmm23, %k1
+
+// CHECK: vpshufbitqmb  268435456(%rcx,%r14,8), %zmm23, %k1
+// CHECK: encoding: [0x62,0xb2,0x45,0x40,0x8f,0x8c,0xf1,0x00,0x00,0x00,0x10]
+          vpshufbitqmb  268435456(%rcx,%r14,8), %zmm23, %k1
+
+// CHECK: vpshufbitqmb  -536870912(%rcx,%r14,8), %zmm23, %k1
+// CHECK: encoding: [0x62,0xb2,0x45,0x40,0x8f,0x8c,0xf1,0x00,0x00,0x00,0xe0]
+          vpshufbitqmb  -536870912(%rcx,%r14,8), %zmm23, %k1
+
+// CHECK: vpshufbitqmb  -536870910(%rcx,%r14,8), %zmm23, %k1
+// CHECK: encoding: [0x62,0xb2,0x45,0x40,0x8f,0x8c,0xf1,0x02,0x00,0x00,0xe0]
+          vpshufbitqmb  -536870910(%rcx,%r14,8), %zmm23, %k1
+
+// CHECK: vpshufbitqmb  (%rcx), %zmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x45,0x42,0x8f,0x09]
+          vpshufbitqmb  (%rcx), %zmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  -256(%rsp), %zmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x45,0x42,0x8f,0x4c,0x24,0xfc]
+          vpshufbitqmb  -256(%rsp), %zmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  256(%rsp), %zmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x45,0x42,0x8f,0x4c,0x24,0x04]
+          vpshufbitqmb  256(%rsp), %zmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  268435456(%rcx,%r14,8), %zmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xb2,0x45,0x42,0x8f,0x8c,0xf1,0x00,0x00,0x00,0x10]
+          vpshufbitqmb  268435456(%rcx,%r14,8), %zmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  -536870912(%rcx,%r14,8), %zmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xb2,0x45,0x42,0x8f,0x8c,0xf1,0x00,0x00,0x00,0xe0]
+          vpshufbitqmb  -536870912(%rcx,%r14,8), %zmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  -536870910(%rcx,%r14,8), %zmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xb2,0x45,0x42,0x8f,0x8c,0xf1,0x02,0x00,0x00,0xe0]
+          vpshufbitqmb  -536870910(%rcx,%r14,8), %zmm23, %k1 {%k2}
+
Index: test/MC/X86/avx512vl_bitalg-encoding.s
===================================================================
--- /dev/null
+++ test/MC/X86/avx512vl_bitalg-encoding.s
@@ -0,0 +1,338 @@
+// RUN: llvm-mc -triple x86_64-unknown-unknown -mcpu=knl -mattr=+avx512vl,+avx512bitalg --show-encoding < %s | FileCheck %s
+
+// CHECK: vpopcntb %xmm23, %xmm21
+// CHECK: encoding: [0x62,0xa2,0x7d,0x08,0x54,0xef]
+          vpopcntb %xmm23, %xmm21
+
+// CHECK: vpopcntw %xmm23, %xmm21
+// CHECK: encoding: [0x62,0xa2,0xfd,0x08,0x54,0xef]
+          vpopcntw %xmm23, %xmm21
+
+// CHECK: vpopcntb %xmm3, %xmm1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x7d,0x0a,0x54,0xcb]
+          vpopcntb %xmm3, %xmm1 {%k2}
+
+// CHECK: vpopcntw %xmm3, %xmm1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0xfd,0x0a,0x54,0xcb]
+          vpopcntw %xmm3, %xmm1 {%k2}
+
+// CHECK: vpopcntb  (%rcx), %xmm1
+// CHECK: encoding: [0x62,0xf2,0x7d,0x08,0x54,0x09]
+          vpopcntb  (%rcx), %xmm1
+
+// CHECK: vpopcntb  -64(%rsp), %xmm1
+// CHECK: encoding: [0x62,0xf2,0x7d,0x08,0x54,0x4c,0x24,0xfc]
+          vpopcntb  -64(%rsp), %xmm1
+
+// CHECK: vpopcntb  64(%rsp), %xmm1
+// CHECK: encoding: [0x62,0xf2,0x7d,0x08,0x54,0x4c,0x24,0x04]
+          vpopcntb  64(%rsp), %xmm1
+
+// CHECK: vpopcntb  268435456(%rcx,%r14,8), %xmm1
+// CHECK: encoding: [0x62,0xb2,0x7d,0x08,0x54,0x8c,0xf1,0x00,0x00,0x00,0x10]
+          vpopcntb  268435456(%rcx,%r14,8), %xmm1
+
+// CHECK: vpopcntb  -536870912(%rcx,%r14,8), %xmm1
+// CHECK: encoding: [0x62,0xb2,0x7d,0x08,0x54,0x8c,0xf1,0x00,0x00,0x00,0xe0]
+          vpopcntb  -536870912(%rcx,%r14,8), %xmm1
+
+// CHECK: vpopcntb  -536870910(%rcx,%r14,8), %xmm1
+// CHECK: encoding: [0x62,0xb2,0x7d,0x08,0x54,0x8c,0xf1,0x02,0x00,0x00,0xe0]
+          vpopcntb  -536870910(%rcx,%r14,8), %xmm1
+
+// CHECK: vpopcntw  (%rcx), %xmm1
+// CHECK: encoding: [0x62,0xf2,0xfd,0x08,0x54,0x09]
+          vpopcntw  (%rcx), %xmm1
+
+// CHECK: vpopcntw  -64(%rsp), %xmm1
+// CHECK: encoding: [0x62,0xf2,0xfd,0x08,0x54,0x4c,0x24,0xfc]
+          vpopcntw  -64(%rsp), %xmm1
+
+// CHECK: vpopcntw  64(%rsp), %xmm1
+// CHECK: encoding: [0x62,0xf2,0xfd,0x08,0x54,0x4c,0x24,0x04]
+          vpopcntw  64(%rsp), %xmm1
+
+// CHECK: vpopcntw  268435456(%rcx,%r14,8), %xmm1
+// CHECK: encoding: [0x62,0xb2,0xfd,0x08,0x54,0x8c,0xf1,0x00,0x00,0x00,0x10]
+          vpopcntw  268435456(%rcx,%r14,8), %xmm1
+
+// CHECK: vpopcntw  -536870912(%rcx,%r14,8), %xmm1
+// CHECK: encoding: [0x62,0xb2,0xfd,0x08,0x54,0x8c,0xf1,0x00,0x00,0x00,0xe0]
+          vpopcntw  -536870912(%rcx,%r14,8), %xmm1
+
+// CHECK: vpopcntw  -536870910(%rcx,%r14,8), %xmm1
+// CHECK: encoding: [0x62,0xb2,0xfd,0x08,0x54,0x8c,0xf1,0x02,0x00,0x00,0xe0]
+          vpopcntw  -536870910(%rcx,%r14,8), %xmm1
+
+// CHECK: vpopcntb  (%rcx), %xmm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0x7d,0x0a,0x54,0x29]
+          vpopcntb  (%rcx), %xmm21 {%k2}
+
+// CHECK: vpopcntb  -64(%rsp), %xmm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0x7d,0x0a,0x54,0x6c,0x24,0xfc]
+          vpopcntb  -64(%rsp), %xmm21 {%k2}
+
+// CHECK: vpopcntb  64(%rsp), %xmm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0x7d,0x0a,0x54,0x6c,0x24,0x04]
+          vpopcntb  64(%rsp), %xmm21 {%k2}
+
+// CHECK: vpopcntb  268435456(%rcx,%r14,8), %xmm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0x7d,0x0a,0x54,0xac,0xf1,0x00,0x00,0x00,0x10]
+          vpopcntb  268435456(%rcx,%r14,8), %xmm21 {%k2}
+
+// CHECK: vpopcntb  -536870912(%rcx,%r14,8), %xmm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0x7d,0x0a,0x54,0xac,0xf1,0x00,0x00,0x00,0xe0]
+          vpopcntb  -536870912(%rcx,%r14,8), %xmm21 {%k2}
+
+// CHECK: vpopcntb  -536870910(%rcx,%r14,8), %xmm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0x7d,0x0a,0x54,0xac,0xf1,0x02,0x00,0x00,0xe0]
+          vpopcntb  -536870910(%rcx,%r14,8), %xmm21 {%k2}
+
+// CHECK: vpopcntw  (%rcx), %xmm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0xfd,0x0a,0x54,0x29]
+          vpopcntw  (%rcx), %xmm21 {%k2}
+
+// CHECK: vpopcntw  -64(%rsp), %xmm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0xfd,0x0a,0x54,0x6c,0x24,0xfc]
+          vpopcntw  -64(%rsp), %xmm21 {%k2}
+
+// CHECK: vpopcntw  64(%rsp), %xmm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0xfd,0x0a,0x54,0x6c,0x24,0x04]
+          vpopcntw  64(%rsp), %xmm21 {%k2}
+
+// CHECK: vpopcntw  268435456(%rcx,%r14,8), %xmm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0xfd,0x0a,0x54,0xac,0xf1,0x00,0x00,0x00,0x10]
+          vpopcntw  268435456(%rcx,%r14,8), %xmm21 {%k2}
+
+// CHECK: vpopcntw  -536870912(%rcx,%r14,8), %xmm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0xfd,0x0a,0x54,0xac,0xf1,0x00,0x00,0x00,0xe0]
+          vpopcntw  -536870912(%rcx,%r14,8), %xmm21 {%k2}
+
+// CHECK: vpopcntw  -536870910(%rcx,%r14,8), %xmm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0xfd,0x0a,0x54,0xac,0xf1,0x02,0x00,0x00,0xe0]
+          vpopcntw  -536870910(%rcx,%r14,8), %xmm21 {%k2}
+
+// CHECK: vpopcntb %ymm23, %ymm21
+// CHECK: encoding: [0x62,0xa2,0x7d,0x28,0x54,0xef]
+          vpopcntb %ymm23, %ymm21
+
+// CHECK: vpopcntw %ymm23, %ymm21
+// CHECK: encoding: [0x62,0xa2,0xfd,0x28,0x54,0xef]
+          vpopcntw %ymm23, %ymm21
+
+// CHECK: vpopcntb %ymm3, %ymm1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x7d,0x2a,0x54,0xcb]
+          vpopcntb %ymm3, %ymm1 {%k2}
+
+// CHECK: vpopcntw %ymm3, %ymm1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0xfd,0x2a,0x54,0xcb]
+          vpopcntw %ymm3, %ymm1 {%k2}
+
+// CHECK: vpopcntb  (%rcx), %ymm1
+// CHECK: encoding: [0x62,0xf2,0x7d,0x28,0x54,0x09]
+          vpopcntb  (%rcx), %ymm1
+
+// CHECK: vpopcntb  -128(%rsp), %ymm1
+// CHECK: encoding: [0x62,0xf2,0x7d,0x28,0x54,0x4c,0x24,0xfc]
+          vpopcntb  -128(%rsp), %ymm1
+
+// CHECK: vpopcntb  128(%rsp), %ymm1
+// CHECK: encoding: [0x62,0xf2,0x7d,0x28,0x54,0x4c,0x24,0x04]
+          vpopcntb  128(%rsp), %ymm1
+
+// CHECK: vpopcntb  268435456(%rcx,%r14,8), %ymm1
+// CHECK: encoding: [0x62,0xb2,0x7d,0x28,0x54,0x8c,0xf1,0x00,0x00,0x00,0x10]
+          vpopcntb  268435456(%rcx,%r14,8), %ymm1
+
+// CHECK: vpopcntb  -536870912(%rcx,%r14,8), %ymm1
+// CHECK: encoding: [0x62,0xb2,0x7d,0x28,0x54,0x8c,0xf1,0x00,0x00,0x00,0xe0]
+          vpopcntb  -536870912(%rcx,%r14,8), %ymm1
+
+// CHECK: vpopcntb  -536870910(%rcx,%r14,8), %ymm1
+// CHECK: encoding: [0x62,0xb2,0x7d,0x28,0x54,0x8c,0xf1,0x02,0x00,0x00,0xe0]
+          vpopcntb  -536870910(%rcx,%r14,8), %ymm1
+
+// CHECK: vpopcntw  (%rcx), %ymm1
+// CHECK: encoding: [0x62,0xf2,0xfd,0x28,0x54,0x09]
+          vpopcntw  (%rcx), %ymm1
+
+// CHECK: vpopcntw  -128(%rsp), %ymm1
+// CHECK: encoding: [0x62,0xf2,0xfd,0x28,0x54,0x4c,0x24,0xfc]
+          vpopcntw  -128(%rsp), %ymm1
+
+// CHECK: vpopcntw  128(%rsp), %ymm1
+// CHECK: encoding: [0x62,0xf2,0xfd,0x28,0x54,0x4c,0x24,0x04]
+          vpopcntw  128(%rsp), %ymm1
+
+// CHECK: vpopcntw  268435456(%rcx,%r14,8), %ymm1
+// CHECK: encoding: [0x62,0xb2,0xfd,0x28,0x54,0x8c,0xf1,0x00,0x00,0x00,0x10]
+          vpopcntw  268435456(%rcx,%r14,8), %ymm1
+
+// CHECK: vpopcntw  -536870912(%rcx,%r14,8), %ymm1
+// CHECK: encoding: [0x62,0xb2,0xfd,0x28,0x54,0x8c,0xf1,0x00,0x00,0x00,0xe0]
+          vpopcntw  -536870912(%rcx,%r14,8), %ymm1
+
+// CHECK: vpopcntw  -536870910(%rcx,%r14,8), %ymm1
+// CHECK: encoding: [0x62,0xb2,0xfd,0x28,0x54,0x8c,0xf1,0x02,0x00,0x00,0xe0]
+          vpopcntw  -536870910(%rcx,%r14,8), %ymm1
+
+// CHECK: vpopcntb  (%rcx), %ymm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0x7d,0x2a,0x54,0x29]
+          vpopcntb  (%rcx), %ymm21 {%k2}
+
+// CHECK: vpopcntb  -128(%rsp), %ymm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0x7d,0x2a,0x54,0x6c,0x24,0xfc]
+          vpopcntb  -128(%rsp), %ymm21 {%k2}
+
+// CHECK: vpopcntb  128(%rsp), %ymm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0x7d,0x2a,0x54,0x6c,0x24,0x04]
+          vpopcntb  128(%rsp), %ymm21 {%k2}
+
+// CHECK: vpopcntb  268435456(%rcx,%r14,8), %ymm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0x7d,0x2a,0x54,0xac,0xf1,0x00,0x00,0x00,0x10]
+          vpopcntb  268435456(%rcx,%r14,8), %ymm21 {%k2}
+
+// CHECK: vpopcntb  -536870912(%rcx,%r14,8), %ymm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0x7d,0x2a,0x54,0xac,0xf1,0x00,0x00,0x00,0xe0]
+          vpopcntb  -536870912(%rcx,%r14,8), %ymm21 {%k2}
+
+// CHECK: vpopcntb  -536870910(%rcx,%r14,8), %ymm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0x7d,0x2a,0x54,0xac,0xf1,0x02,0x00,0x00,0xe0]
+          vpopcntb  -536870910(%rcx,%r14,8), %ymm21 {%k2}
+
+// CHECK: vpopcntw  (%rcx), %ymm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0xfd,0x2a,0x54,0x29]
+          vpopcntw  (%rcx), %ymm21 {%k2}
+
+// CHECK: vpopcntw  -128(%rsp), %ymm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0xfd,0x2a,0x54,0x6c,0x24,0xfc]
+          vpopcntw  -128(%rsp), %ymm21 {%k2}
+
+// CHECK: vpopcntw  128(%rsp), %ymm21 {%k2}
+// CHECK: encoding: [0x62,0xe2,0xfd,0x2a,0x54,0x6c,0x24,0x04]
+          vpopcntw  128(%rsp), %ymm21 {%k2}
+
+// CHECK: vpopcntw  268435456(%rcx,%r14,8), %ymm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0xfd,0x2a,0x54,0xac,0xf1,0x00,0x00,0x00,0x10]
+          vpopcntw  268435456(%rcx,%r14,8), %ymm21 {%k2}
+
+// CHECK: vpopcntw  -536870912(%rcx,%r14,8), %ymm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0xfd,0x2a,0x54,0xac,0xf1,0x00,0x00,0x00,0xe0]
+          vpopcntw  -536870912(%rcx,%r14,8), %ymm21 {%k2}
+
+// CHECK: vpopcntw  -536870910(%rcx,%r14,8), %ymm21 {%k2}
+// CHECK: encoding: [0x62,0xa2,0xfd,0x2a,0x54,0xac,0xf1,0x02,0x00,0x00,0xe0]
+          vpopcntw  -536870910(%rcx,%r14,8), %ymm21 {%k2}
+
+// CHECK: vpshufbitqmb %xmm2, %xmm23, %k1
+// CHECK: encoding: [0x62,0xf2,0x45,0x00,0x8f,0xca]
+          vpshufbitqmb %xmm2, %xmm23, %k1
+
+// CHECK: vpshufbitqmb %xmm2, %xmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x45,0x02,0x8f,0xca]
+          vpshufbitqmb %xmm2, %xmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  (%rcx), %xmm23, %k1
+// CHECK: encoding: [0x62,0xf2,0x45,0x00,0x8f,0x09]
+          vpshufbitqmb  (%rcx), %xmm23, %k1
+
+// CHECK: vpshufbitqmb  -64(%rsp), %xmm23, %k1
+// CHECK: encoding: [0x62,0xf2,0x45,0x00,0x8f,0x4c,0x24,0xfc]
+          vpshufbitqmb  -64(%rsp), %xmm23, %k1
+
+// CHECK: vpshufbitqmb  64(%rsp), %xmm23, %k1
+// CHECK: encoding: [0x62,0xf2,0x45,0x00,0x8f,0x4c,0x24,0x04]
+          vpshufbitqmb  64(%rsp), %xmm23, %k1
+
+// CHECK: vpshufbitqmb  268435456(%rcx,%r14,8), %xmm23, %k1
+// CHECK: encoding: [0x62,0xb2,0x45,0x00,0x8f,0x8c,0xf1,0x00,0x00,0x00,0x10]
+          vpshufbitqmb  268435456(%rcx,%r14,8), %xmm23, %k1
+
+// CHECK: vpshufbitqmb  -536870912(%rcx,%r14,8), %xmm23, %k1
+// CHECK: encoding: [0x62,0xb2,0x45,0x00,0x8f,0x8c,0xf1,0x00,0x00,0x00,0xe0]
+          vpshufbitqmb  -536870912(%rcx,%r14,8), %xmm23, %k1
+
+// CHECK: vpshufbitqmb  -536870910(%rcx,%r14,8), %xmm23, %k1
+// CHECK: encoding: [0x62,0xb2,0x45,0x00,0x8f,0x8c,0xf1,0x02,0x00,0x00,0xe0]
+          vpshufbitqmb  -536870910(%rcx,%r14,8), %xmm23, %k1
+
+// CHECK: vpshufbitqmb  (%rcx), %xmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x45,0x02,0x8f,0x09]
+          vpshufbitqmb  (%rcx), %xmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  -64(%rsp), %xmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x45,0x02,0x8f,0x4c,0x24,0xfc]
+          vpshufbitqmb  -64(%rsp), %xmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  64(%rsp), %xmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x45,0x02,0x8f,0x4c,0x24,0x04]
+          vpshufbitqmb  64(%rsp), %xmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  268435456(%rcx,%r14,8), %xmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xb2,0x45,0x02,0x8f,0x8c,0xf1,0x00,0x00,0x00,0x10]
+          vpshufbitqmb  268435456(%rcx,%r14,8), %xmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  -536870912(%rcx,%r14,8), %xmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xb2,0x45,0x02,0x8f,0x8c,0xf1,0x00,0x00,0x00,0xe0]
+          vpshufbitqmb  -536870912(%rcx,%r14,8), %xmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  -536870910(%rcx,%r14,8), %xmm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xb2,0x45,0x02,0x8f,0x8c,0xf1,0x02,0x00,0x00,0xe0]
+          vpshufbitqmb  -536870910(%rcx,%r14,8), %xmm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb %ymm2, %ymm23, %k1
+// CHECK: encoding: [0x62,0xf2,0x45,0x20,0x8f,0xca]
+          vpshufbitqmb %ymm2, %ymm23, %k1
+
+// CHECK: vpshufbitqmb %ymm2, %ymm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x45,0x22,0x8f,0xca]
+          vpshufbitqmb %ymm2, %ymm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  (%rcx), %ymm23, %k1
+// CHECK: encoding: [0x62,0xf2,0x45,0x20,0x8f,0x09]
+          vpshufbitqmb  (%rcx), %ymm23, %k1
+
+// CHECK: vpshufbitqmb  -128(%rsp), %ymm23, %k1
+// CHECK: encoding: [0x62,0xf2,0x45,0x20,0x8f,0x4c,0x24,0xfc]
+          vpshufbitqmb  -128(%rsp), %ymm23, %k1
+
+// CHECK: vpshufbitqmb  128(%rsp), %ymm23, %k1
+// CHECK: encoding: [0x62,0xf2,0x45,0x20,0x8f,0x4c,0x24,0x04]
+          vpshufbitqmb  128(%rsp), %ymm23, %k1
+
+// CHECK: vpshufbitqmb  268435456(%rcx,%r14,8), %ymm23, %k1
+// CHECK: encoding: [0x62,0xb2,0x45,0x20,0x8f,0x8c,0xf1,0x00,0x00,0x00,0x10]
+          vpshufbitqmb  268435456(%rcx,%r14,8), %ymm23, %k1
+
+// CHECK: vpshufbitqmb  -536870912(%rcx,%r14,8), %ymm23, %k1
+// CHECK: encoding: [0x62,0xb2,0x45,0x20,0x8f,0x8c,0xf1,0x00,0x00,0x00,0xe0]
+          vpshufbitqmb  -536870912(%rcx,%r14,8), %ymm23, %k1
+
+// CHECK: vpshufbitqmb  -536870910(%rcx,%r14,8), %ymm23, %k1
+// CHECK: encoding: [0x62,0xb2,0x45,0x20,0x8f,0x8c,0xf1,0x02,0x00,0x00,0xe0]
+          vpshufbitqmb  -536870910(%rcx,%r14,8), %ymm23, %k1
+
+// CHECK: vpshufbitqmb  (%rcx), %ymm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x45,0x22,0x8f,0x09]
+          vpshufbitqmb  (%rcx), %ymm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  -128(%rsp), %ymm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x45,0x22,0x8f,0x4c,0x24,0xfc]
+          vpshufbitqmb  -128(%rsp), %ymm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  128(%rsp), %ymm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xf2,0x45,0x22,0x8f,0x4c,0x24,0x04]
+          vpshufbitqmb  128(%rsp), %ymm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  268435456(%rcx,%r14,8), %ymm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xb2,0x45,0x22,0x8f,0x8c,0xf1,0x00,0x00,0x00,0x10]
+          vpshufbitqmb  268435456(%rcx,%r14,8), %ymm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  -536870912(%rcx,%r14,8), %ymm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xb2,0x45,0x22,0x8f,0x8c,0xf1,0x00,0x00,0x00,0xe0]
+          vpshufbitqmb  -536870912(%rcx,%r14,8), %ymm23, %k1 {%k2}
+
+// CHECK: vpshufbitqmb  -536870910(%rcx,%r14,8), %ymm23, %k1 {%k2}
+// CHECK: encoding: [0x62,0xb2,0x45,0x22,0x8f,0x8c,0xf1,0x02,0x00,0x00,0xe0]
+          vpshufbitqmb  -536870910(%rcx,%r14,8), %ymm23, %k1 {%k2}
+