[x86] Lowering Mask Set1 intrinsics to LLVM IR

This patch, together with a matching clang patch (https://reviews.llvm.org/D37668), implements the lowering of X86 mask set1 intrinsics to IR.

Differential Revision: https://reviews.llvm.org/D37669

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@313625 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/include/llvm/IR/IntrinsicsX86.td b/include/llvm/IR/IntrinsicsX86.td
index fe63072275563c012812ca4c88b942478de7cecc..a5f32d23cfca7ae8675f00cdce2d475583706e2e 100644 (file)
--- a/include/llvm/IR/IntrinsicsX86.td
+++ b/include/llvm/IR/IntrinsicsX86.td
@@ -1871,65 +1871,6 @@ let TargetPrefix = "x86" in {  // All intrinsics start with "llvm.x86.".
                          llvm_v32i8_ty], [IntrNoMem]>;
 }
 
-// Vector load with broadcast
-let TargetPrefix = "x86" in {  // All intrinsics start with "llvm.x86.".
-  def int_x86_avx512_mask_pbroadcast_b_gpr_128 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastb128_gpr_mask">,
-          Intrinsic<[llvm_v16i8_ty],
-                    [llvm_i8_ty, llvm_v16i8_ty, llvm_i16_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_pbroadcast_b_gpr_256 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastb256_gpr_mask">,
-          Intrinsic<[llvm_v32i8_ty],
-                    [llvm_i8_ty, llvm_v32i8_ty, llvm_i32_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_pbroadcast_b_gpr_512 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastb512_gpr_mask">,
-          Intrinsic<[llvm_v64i8_ty],
-                    [llvm_i8_ty, llvm_v64i8_ty, llvm_i64_ty], [IntrNoMem]>;
-
-  def int_x86_avx512_mask_pbroadcast_w_gpr_128 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastw128_gpr_mask">,
-          Intrinsic<[llvm_v8i16_ty],
-                    [llvm_i16_ty, llvm_v8i16_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_pbroadcast_w_gpr_256 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastw256_gpr_mask">,
-          Intrinsic<[llvm_v16i16_ty],
-                    [llvm_i16_ty, llvm_v16i16_ty, llvm_i16_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_pbroadcast_w_gpr_512 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastw512_gpr_mask">,
-          Intrinsic<[llvm_v32i16_ty],
-                    [llvm_i16_ty, llvm_v32i16_ty, llvm_i32_ty], [IntrNoMem]>;
-
-  def int_x86_avx512_mask_pbroadcast_d_gpr_128 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastd128_gpr_mask">,
-          Intrinsic<[llvm_v4i32_ty],
-                    [llvm_i32_ty, llvm_v4i32_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_pbroadcast_d_gpr_256 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastd256_gpr_mask">,
-          Intrinsic<[llvm_v8i32_ty],
-                    [llvm_i32_ty, llvm_v8i32_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_pbroadcast_d_gpr_512 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastd512_gpr_mask">,
-          Intrinsic<[llvm_v16i32_ty],
-                    [llvm_i32_ty, llvm_v16i32_ty, llvm_i16_ty], [IntrNoMem]>;
-
-  def int_x86_avx512_mask_pbroadcast_q_gpr_128 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastq128_gpr_mask">,
-          Intrinsic<[llvm_v2i64_ty],
-                    [llvm_i64_ty, llvm_v2i64_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_pbroadcast_q_gpr_256 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastq256_gpr_mask">,
-          Intrinsic<[llvm_v4i64_ty],
-                    [llvm_i64_ty, llvm_v4i64_ty, llvm_i8_ty], [IntrNoMem]>;
-  def int_x86_avx512_mask_pbroadcast_q_gpr_512 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastq512_gpr_mask">,
-          Intrinsic<[llvm_v8i64_ty],
-                    [llvm_i64_ty, llvm_v8i64_ty, llvm_i8_ty], [IntrNoMem]>;
-
-  def int_x86_avx512_mask_pbroadcast_q_mem_512 :
-          GCCBuiltin<"__builtin_ia32_pbroadcastq512_mem_mask">,
-          Intrinsic<[llvm_v8i64_ty],
-                    [llvm_i64_ty, llvm_v8i64_ty, llvm_i8_ty], [IntrNoMem]>;
-}
 
 // Vector permutation
 let TargetPrefix = "x86" in {  // All intrinsics start with "llvm.x86.".

diff --git a/lib/IR/AutoUpgrade.cpp b/lib/IR/AutoUpgrade.cpp
index 2e8a41e06e456847661cfa31a09eda6bf716ff59..b1c8e74084faea6909e3b338cd663783d7e4b9e2 100644 (file)
--- a/lib/IR/AutoUpgrade.cpp
+++ b/lib/IR/AutoUpgrade.cpp
@@ -77,6 +77,7 @@ static bool ShouldUpgradeX86Intrinsic(Function *F, StringRef Name) {
       Name=="ssse3.pabs.d.128" || // Added in 6.0
       Name.startswith("avx2.pabs.") || // Added in 6.0
       Name.startswith("avx512.mask.pabs.") || // Added in 6.0
+      Name.startswith("avx512.mask.pbroadcast") || // Added in 6.0
       Name.startswith("sse2.pcmpeq.") || // Added in 3.1
       Name.startswith("sse2.pcmpgt.") || // Added in 3.1
       Name.startswith("avx2.pcmpeq.") || // Added in 3.1
@@ -1031,6 +1032,12 @@ void llvm::UpgradeIntrinsicCall(CallInst *CI, Function *NewFn) {
       Rep = Builder.CreateICmp(CmpEq ? ICmpInst::ICMP_EQ : ICmpInst::ICMP_SGT,
                                CI->getArgOperand(0), CI->getArgOperand(1));
       Rep = Builder.CreateSExt(Rep, CI->getType(), "");
+    } else if (IsX86 && (Name.startswith("avx512.mask.pbroadcast"))){
+      unsigned NumElts =
+          CI->getArgOperand(1)->getType()->getVectorNumElements();
+      Rep = Builder.CreateVectorSplat(NumElts, CI->getArgOperand(0));
+      Rep = EmitX86Select(Builder, CI->getArgOperand(2), Rep,
+                          CI->getArgOperand(1));
     } else if (IsX86 && (Name == "sse.add.ss" || Name == "sse2.add.sd")) {
       Type *I32Ty = Type::getInt32Ty(C);
       Value *Elt0 = Builder.CreateExtractElement(CI->getArgOperand(0),

diff --git a/lib/Target/X86/X86IntrinsicsInfo.h b/lib/Target/X86/X86IntrinsicsInfo.h
index 4f1b6572875199d22f692a6ac9b3aa7fb69f9b56..d9d0b06c96077a428d6e42397d39d7d08a652f02 100644 (file)
--- a/lib/Target/X86/X86IntrinsicsInfo.h
+++ b/lib/Target/X86/X86IntrinsicsInfo.h
@@ -797,30 +797,6 @@ static const IntrinsicData  IntrinsicsWithoutChain[] = {
   X86_INTRINSIC_DATA(avx512_mask_paddus_w_128, INTR_TYPE_2OP_MASK, X86ISD::ADDUS, 0),
   X86_INTRINSIC_DATA(avx512_mask_paddus_w_256, INTR_TYPE_2OP_MASK, X86ISD::ADDUS, 0),
   X86_INTRINSIC_DATA(avx512_mask_paddus_w_512, INTR_TYPE_2OP_MASK, X86ISD::ADDUS, 0),
-  X86_INTRINSIC_DATA(avx512_mask_pbroadcast_b_gpr_128, INTR_TYPE_1OP_MASK,
-                     X86ISD::VBROADCAST, 0),
-  X86_INTRINSIC_DATA(avx512_mask_pbroadcast_b_gpr_256, INTR_TYPE_1OP_MASK,
-                     X86ISD::VBROADCAST, 0),
-  X86_INTRINSIC_DATA(avx512_mask_pbroadcast_b_gpr_512, INTR_TYPE_1OP_MASK,
-                     X86ISD::VBROADCAST, 0),
-  X86_INTRINSIC_DATA(avx512_mask_pbroadcast_d_gpr_128, INTR_TYPE_1OP_MASK,
-                     X86ISD::VBROADCAST, 0),
-  X86_INTRINSIC_DATA(avx512_mask_pbroadcast_d_gpr_256, INTR_TYPE_1OP_MASK,
-                     X86ISD::VBROADCAST, 0),
-  X86_INTRINSIC_DATA(avx512_mask_pbroadcast_d_gpr_512, INTR_TYPE_1OP_MASK,
-                     X86ISD::VBROADCAST, 0),
-  X86_INTRINSIC_DATA(avx512_mask_pbroadcast_q_gpr_128, INTR_TYPE_1OP_MASK,
-                     X86ISD::VBROADCAST, 0),
-  X86_INTRINSIC_DATA(avx512_mask_pbroadcast_q_gpr_256, INTR_TYPE_1OP_MASK,
-                     X86ISD::VBROADCAST, 0),
-  X86_INTRINSIC_DATA(avx512_mask_pbroadcast_q_gpr_512, INTR_TYPE_1OP_MASK,
-                     X86ISD::VBROADCAST, 0),
-  X86_INTRINSIC_DATA(avx512_mask_pbroadcast_w_gpr_128, INTR_TYPE_1OP_MASK,
-                     X86ISD::VBROADCAST, 0),
-  X86_INTRINSIC_DATA(avx512_mask_pbroadcast_w_gpr_256, INTR_TYPE_1OP_MASK,
-                     X86ISD::VBROADCAST, 0),
-  X86_INTRINSIC_DATA(avx512_mask_pbroadcast_w_gpr_512, INTR_TYPE_1OP_MASK,
-                     X86ISD::VBROADCAST, 0),
   X86_INTRINSIC_DATA(avx512_mask_permvar_df_256, VPERM_2OP_MASK,
                      X86ISD::VPERMV, 0),
   X86_INTRINSIC_DATA(avx512_mask_permvar_df_512, VPERM_2OP_MASK,

diff --git a/test/CodeGen/X86/avx512-intrinsics-fast-isel.ll b/test/CodeGen/X86/avx512-intrinsics-fast-isel.ll
index 65e655434c44853d6c99ed98a818fede370f6fb3..deef8ba80b64700600673e632d05d37e2283d794 100644 (file)
--- a/test/CodeGen/X86/avx512-intrinsics-fast-isel.ll
+++ b/test/CodeGen/X86/avx512-intrinsics-fast-isel.ll
@@ -4,6 +4,110 @@
 
 ; NOTE: This should use IR equivalent to what is generated by clang/test/CodeGen/avx512f-builtins.c
 
+define <8 x i64> @test_mm512_mask_set1_epi32(<8 x i64> %__O, i16 zeroext %__M, i32 %__A) {
+; X32-LABEL: test_mm512_mask_set1_epi32:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movzwl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    kmovw %ecx, %k1
+; X32-NEXT:    vpbroadcastd %eax, %zmm0 {%k1}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm512_mask_set1_epi32:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovw %edi, %k1
+; X64-NEXT:    vpbroadcastd %esi, %zmm0 {%k1}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <16 x i32> undef, i32 %__A, i32 0
+  %vecinit15.i.i = shufflevector <16 x i32> %vecinit.i.i, <16 x i32> undef, <16 x i32> zeroinitializer
+  %0 = bitcast <8 x i64> %__O to <16 x i32>
+  %1 = bitcast i16 %__M to <16 x i1>
+  %2 = select <16 x i1> %1, <16 x i32> %vecinit15.i.i, <16 x i32> %0
+  %3 = bitcast <16 x i32> %2 to <8 x i64>
+  ret <8 x i64> %3
+}
+
+define <8 x i64> @test_mm512_maskz_set1_epi32(i16 zeroext %__M, i32 %__A)  {
+; X32-LABEL: test_mm512_maskz_set1_epi32:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movzwl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    kmovw %ecx, %k1
+; X32-NEXT:    vpbroadcastd %eax, %zmm0 {%k1} {z}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm512_maskz_set1_epi32:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovw %edi, %k1
+; X64-NEXT:    vpbroadcastd %esi, %zmm0 {%k1} {z}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <16 x i32> undef, i32 %__A, i32 0
+  %vecinit15.i.i = shufflevector <16 x i32> %vecinit.i.i, <16 x i32> undef, <16 x i32> zeroinitializer
+  %0 = bitcast i16 %__M to <16 x i1>
+  %1 = select <16 x i1> %0, <16 x i32> %vecinit15.i.i, <16 x i32> zeroinitializer
+  %2 = bitcast <16 x i32> %1 to <8 x i64>
+  ret <8 x i64> %2
+}
+
+define <8 x i64> @test_mm512_mask_set1_epi64(<8 x i64> %__O, i8 zeroext %__M, i64 %__A) {
+; X32-LABEL: test_mm512_mask_set1_epi64:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; X32-NEXT:    vmovd %edx, %xmm1
+; X32-NEXT:    vpinsrd $1, %ecx, %xmm1, %xmm1
+; X32-NEXT:    vpinsrd $2, %edx, %xmm1, %xmm1
+; X32-NEXT:    vpinsrd $3, %ecx, %xmm1, %xmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm1, %ymm1
+; X32-NEXT:    kmovw %eax, %k1
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm1, %zmm0 {%k1}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm512_mask_set1_epi64:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovw %edi, %k1
+; X64-NEXT:    vpbroadcastq %rsi, %zmm0 {%k1}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <8 x i64> undef, i64 %__A, i32 0
+  %vecinit7.i.i = shufflevector <8 x i64> %vecinit.i.i, <8 x i64> undef, <8 x i32> zeroinitializer
+  %0 = bitcast i8 %__M to <8 x i1>
+  %1 = select <8 x i1> %0, <8 x i64> %vecinit7.i.i, <8 x i64> %__O
+  ret <8 x i64> %1
+}
+
+define <8 x i64> @test_mm512_maskz_set1_epi64(i8 zeroext %__M, i64 %__A)  {
+; X32-LABEL: test_mm512_maskz_set1_epi64:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %edx
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; X32-NEXT:    vmovd %edx, %xmm0
+; X32-NEXT:    vpinsrd $1, %ecx, %xmm0, %xmm0
+; X32-NEXT:    vpinsrd $2, %edx, %xmm0, %xmm0
+; X32-NEXT:    vpinsrd $3, %ecx, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    kmovw %eax, %k1
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm0, %zmm0 {%k1} {z}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm512_maskz_set1_epi64:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovw %edi, %k1
+; X64-NEXT:    vpbroadcastq %rsi, %zmm0 {%k1} {z}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <8 x i64> undef, i64 %__A, i32 0
+  %vecinit7.i.i = shufflevector <8 x i64> %vecinit.i.i, <8 x i64> undef, <8 x i32> zeroinitializer
+  %0 = bitcast i8 %__M to <8 x i1>
+  %1 = select <8 x i1> %0, <8 x i64> %vecinit7.i.i, <8 x i64> zeroinitializer
+  ret <8 x i64> %1
+}
+
+
 define <8 x i64> @test_mm512_broadcastd_epi32(<2 x i64> %a0) {
 ; X32-LABEL: test_mm512_broadcastd_epi32:
 ; X32:       # BB#0:

diff --git a/test/CodeGen/X86/avx512-intrinsics-upgrade.ll b/test/CodeGen/X86/avx512-intrinsics-upgrade.ll
index f401cac82da8f4441952202913d424ade33f1c10..8d653c78182cab6c09a6e02f5988c771d1f8a938 100644 (file)
--- a/test/CodeGen/X86/avx512-intrinsics-upgrade.ll
+++ b/test/CodeGen/X86/avx512-intrinsics-upgrade.ll
@@ -1,6 +1,46 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=knl | FileCheck %s
 
+ define <16 x i32>@test_int_x86_avx512_mask_pbroadcastd_gpr_512(i32 %x0, <16 x i32> %x1, i16 %mask) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcastd_gpr_512:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vpbroadcastd %edi, %zmm1
+; CHECK-NEXT:    kmovw %esi, %k1
+; CHECK-NEXT:    vpbroadcastd %edi, %zmm0 {%k1}
+; CHECK-NEXT:    vpaddd %zmm0, %zmm1, %zmm0
+; CHECK-NEXT:    vpbroadcastd %edi, %zmm1 {%k1} {z}
+; CHECK-NEXT:    vpaddd %zmm0, %zmm1, %zmm0
+; CHECK-NEXT:    retq
+    %res = call <16 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.512(i32 %x0, <16 x i32> %x1, i16 -1)
+    %res1 = call <16 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.512(i32 %x0, <16 x i32> %x1, i16 %mask)
+    %res2 = call <16 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.512(i32 %x0, <16 x i32> zeroinitializer, i16 %mask)
+    %res3 = add <16 x i32> %res, %res1
+    %res4 = add <16 x i32> %res2, %res3
+    ret <16 x i32> %res4
+  }
+declare <16 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.512(i32, <16 x i32>, i16)
+
+
+define <8 x i64>@test_int_x86_avx512_mask_pbroadcastq_gpr_512(i64 %x0, <8 x i64> %x1, i8 %mask) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcastq_gpr_512:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vpbroadcastq %rdi, %zmm1
+; CHECK-NEXT:    kmovw %esi, %k1
+; CHECK-NEXT:    vpbroadcastq %rdi, %zmm0 {%k1}
+; CHECK-NEXT:    vpaddq %zmm0, %zmm1, %zmm0
+; CHECK-NEXT:    vpbroadcastq %rdi, %zmm1 {%k1} {z}
+; CHECK-NEXT:    vpaddq %zmm0, %zmm1, %zmm0
+; CHECK-NEXT:    retq
+   %res = call <8 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.512(i64 %x0, <8 x i64> %x1,i8 -1)
+   %res1 = call <8 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.512(i64 %x0, <8 x i64> %x1,i8 %mask)
+   %res2 = call <8 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.512(i64 %x0, <8 x i64> zeroinitializer,i8 %mask)
+   %res3 = add <8 x i64> %res, %res1
+   %res4 = add <8 x i64> %res2, %res3
+   ret <8 x i64> %res4
+}
+declare <8 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.512(i64, <8 x i64>, i8)
+
+
 declare <16 x float> @llvm.x86.avx512.mask.broadcast.ss.ps.512(<4 x float>, <16 x float>, i16) nounwind readonly
 
 define <16 x float> @test_x86_vbroadcast_ss_ps_512(<4 x float> %a0, <16 x float> %a1, i16 %mask ) {

diff --git a/test/CodeGen/X86/avx512-intrinsics.ll b/test/CodeGen/X86/avx512-intrinsics.ll
index ffbe5b29b48d438ff799d0cc07a73ff969998c1a..4816e10de585c0cc2b922e12653ff457ef26a170 100644 (file)
--- a/test/CodeGen/X86/avx512-intrinsics.ll
+++ b/test/CodeGen/X86/avx512-intrinsics.ll
@@ -4171,44 +4171,8 @@ define i8@test_int_x86_avx512_ptestnm_q_512(<8 x i64> %x0, <8 x i64> %x1, i8 %x2
   ret i8 %res2
 }
 
-define <16 x i32>@test_int_x86_avx512_mask_pbroadcastd_gpr_512(i32 %x0, <16 x i32> %x1, i16 %mask) {
-; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcastd_gpr_512:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %esi, %k1
-; CHECK-NEXT:    vpbroadcastd %edi, %zmm1 {%k1} {z}
-; CHECK-NEXT:    vpbroadcastd %edi, %zmm0 {%k1}
-; CHECK-NEXT:    vpbroadcastd %edi, %zmm2
-; CHECK-NEXT:    vpaddd %zmm0, %zmm2, %zmm0
-; CHECK-NEXT:    vpaddd %zmm0, %zmm1, %zmm0
-; CHECK-NEXT:    retq
-  %res = call <16 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.512(i32 %x0, <16 x i32> %x1, i16 -1)
-  %res1 = call <16 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.512(i32 %x0, <16 x i32> %x1, i16 %mask)
-  %res2 = call <16 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.512(i32 %x0, <16 x i32> zeroinitializer, i16 %mask)
-  %res3 = add <16 x i32> %res, %res1
-  %res4 = add <16 x i32> %res2, %res3
-  ret <16 x i32> %res4
-}
 
-declare <16 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.512(i32, <16 x i32>, i16)
 
-define <8 x i64>@test_int_x86_avx512_mask_pbroadcastq_gpr_512(i64 %x0, <8 x i64> %x1, i8 %mask) {
-; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcastq_gpr_512:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %esi, %k1
-; CHECK-NEXT:    vpbroadcastq %rdi, %zmm1 {%k1} {z}
-; CHECK-NEXT:    vpbroadcastq %rdi, %zmm0 {%k1}
-; CHECK-NEXT:    vpbroadcastq %rdi, %zmm2
-; CHECK-NEXT:    vpaddq %zmm0, %zmm2, %zmm0
-; CHECK-NEXT:    vpaddq %zmm0, %zmm1, %zmm0
-; CHECK-NEXT:    retq
-  %res = call <8 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.512(i64 %x0, <8 x i64> %x1,i8 -1)
-  %res1 = call <8 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.512(i64 %x0, <8 x i64> %x1,i8 %mask)
-  %res2 = call <8 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.512(i64 %x0, <8 x i64> zeroinitializer,i8 %mask)
-  %res3 = add <8 x i64> %res, %res1
-  %res4 = add <8 x i64> %res2, %res3
-  ret <8 x i64> %res4
-}
-declare <8 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.512(i64, <8 x i64>, i8)
 
 declare <2 x double> @llvm.x86.avx512.mask.vfmadd.sd(<2 x double>, <2 x double>, <2 x double>, i8, i32)
 

diff --git a/test/CodeGen/X86/avx512bw-intrinsics-fast-isel.ll b/test/CodeGen/X86/avx512bw-intrinsics-fast-isel.ll
index 50a9076163e8073058963ea3b2ced9fa34c76813..184e152a9e4650d748b6f017505bb2386baece68 100644 (file)
--- a/test/CodeGen/X86/avx512bw-intrinsics-fast-isel.ll
+++ b/test/CodeGen/X86/avx512bw-intrinsics-fast-isel.ll
@@ -4,6 +4,1513 @@
 
 ; NOTE: This should use IR equivalent to what is generated by clang/test/CodeGen/avx512bw-builtins.c
 
+define <8 x i64> @test_mm512_mask_set1_epi8(<8 x i64> %__O, i64 %__M, i8 signext %__A)  {
+; X32-LABEL: test_mm512_mask_set1_epi8:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    pushl %ebx
+; X32-NEXT:  .Lcfi0:
+; X32-NEXT:    .cfi_def_cfa_offset 8
+; X32-NEXT:  .Lcfi1:
+; X32-NEXT:    .cfi_offset %ebx, -8
+; X32-NEXT:    vmovdqa64 %zmm0, %zmm3
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andb $2, %cl
+; X32-NEXT:    shrb %cl
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vpsllw $8, %xmm1, %xmm1
+; X32-NEXT:    kmovd %eax, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andb $15, %cl
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $2, %dl
+; X32-NEXT:    kmovd %edx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vpbroadcastw %xmm2, %xmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm1, %ymm2, %ymm2
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm2[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    shrb $3, %cl
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vpslld $24, %xmm2, %xmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm1, %ymm2, %ymm2
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm2[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $4, %cl
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vpbroadcastd %xmm2, %xmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm1, %ymm2, %ymm2
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm2[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $5, %cl
+; X32-NEXT:    andb $1, %cl
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vpsllq $40, %xmm2, %xmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm1, %ymm2, %ymm2
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm2[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $6, %cl
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vpbroadcastw %xmm2, %xmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm1, %ymm2, %ymm2
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm2[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $7, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpsllq $56, %xmm1, %xmm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movb %ah, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpbroadcastq %xmm1, %xmm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    andb $2, %cl
+; X32-NEXT:    shrb %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpslldq {{.*#+}} xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0,1,2,3,4,5,6]
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movb %ah, %cl
+; X32-NEXT:    andb $15, %cl
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $2, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpbroadcastw %xmm1, %xmm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    shrb $3, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpslldq {{.*#+}} xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0,1,2,3,4]
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andl $61440, %ecx # imm = 0xF000
+; X32-NEXT:    shrl $12, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpbroadcastd %xmm1, %xmm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $13, %ecx
+; X32-NEXT:    andb $1, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpslldq {{.*#+}} xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0,1,2]
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andl $49152, %ecx # imm = 0xC000
+; X32-NEXT:    shrl $14, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpbroadcastw %xmm1, %xmm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andl $32768, %ecx # imm = 0x8000
+; X32-NEXT:    shrl $15, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpslldq {{.*#+}} xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0]
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $16, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    andb $2, %dl
+; X32-NEXT:    shrb %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpsllw $8, %xmm1, %xmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    andb $15, %dl
+; X32-NEXT:    movl %edx, %ebx
+; X32-NEXT:    shrb $2, %bl
+; X32-NEXT:    kmovd %ebx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpbroadcastw %xmm1, %xmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    shrb $3, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpslld $24, %xmm1, %xmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $4, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpbroadcastd %xmm1, %xmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $5, %dl
+; X32-NEXT:    andb $1, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpsllq $40, %xmm1, %xmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $6, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpbroadcastw %xmm1, %xmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    shrb $7, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpsllq $56, %xmm1, %xmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $24, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpbroadcastq %xmm1, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    andb $2, %dl
+; X32-NEXT:    shrb %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpslldq {{.*#+}} xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0,1,2,3,4,5,6]
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm5 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm5, %ymm2, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm2[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    andb $15, %cl
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $2, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpbroadcastw %xmm1, %xmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm4, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm4[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    shrb $3, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpslldq {{.*#+}} xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0,1,2,3,4]
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm4, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm4[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $28, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpbroadcastd %xmm1, %xmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm0 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255]
+; X32-NEXT:    vpblendvb %ymm0, %ymm4, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm1 = zmm1[0,1,2,3],zmm4[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm1, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $29, %ecx
+; X32-NEXT:    andb $1, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm1
+; X32-NEXT:    vpslldq {{.*#+}} xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm1[0,1,2]
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vmovdqa {{.*#+}} ymm4 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255]
+; X32-NEXT:    vpblendvb %ymm4, %ymm0, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm1[0,1,2,3],zmm0[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $30, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vmovdqa {{.*#+}} ymm4 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255]
+; X32-NEXT:    vpblendvb %ymm4, %ymm0, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm1[0,1,2,3],zmm0[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrl $31, %eax
+; X32-NEXT:    kmovd %eax, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0]
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0]
+; X32-NEXT:    vpblendvb %ymm7, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    kmovd %eax, %k1
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
+; X32-NEXT:    vpmovm2b %k1, %zmm7
+; X32-NEXT:    vmovdqa {{.*#+}} ymm6 = [0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm6, %ymm1, %ymm7, %ymm1
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm0, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andb $2, %cl
+; X32-NEXT:    shrb %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllw $8, %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andb $15, %cl
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $2, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrb $3, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslld $24, %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $4, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastd %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $5, %cl
+; X32-NEXT:    andb $1, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllq $40, %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $6, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $7, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllq $56, %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movb %ah, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastq %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    andb $2, %cl
+; X32-NEXT:    shrb %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4,5,6]
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movb %ah, %cl
+; X32-NEXT:    andb $15, %cl
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $2, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrb $3, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4]
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andl $61440, %ecx # imm = 0xF000
+; X32-NEXT:    shrl $12, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastd %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $13, %ecx
+; X32-NEXT:    andb $1, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2]
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andl $49152, %ecx # imm = 0xC000
+; X32-NEXT:    shrl $14, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andl $32768, %ecx # imm = 0x8000
+; X32-NEXT:    shrl $15, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0]
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $16, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    andb $2, %dl
+; X32-NEXT:    shrb %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllw $8, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    andb $15, %dl
+; X32-NEXT:    movl %edx, %ebx
+; X32-NEXT:    shrb $2, %bl
+; X32-NEXT:    kmovd %ebx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrb $3, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslld $24, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $4, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastd %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $5, %dl
+; X32-NEXT:    andb $1, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllq $40, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $6, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrb $7, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllq $56, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $24, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastq %xmm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    andb $2, %dl
+; X32-NEXT:    shrb %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4,5,6]
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm6
+; X32-NEXT:    vpblendvb %ymm5, %ymm6, %ymm0, %ymm0
+; X32-NEXT:    andb $15, %cl
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $2, %dl
+; X32-NEXT:    kmovd %edx, %k0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vpbroadcastw %xmm1, %xmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm1
+; X32-NEXT:    vextracti64x4 $1, %zmm0, %ymm5
+; X32-NEXT:    vpblendvb %ymm2, %ymm5, %ymm1, %ymm2
+; X32-NEXT:    shrb $3, %cl
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $29, %ecx
+; X32-NEXT:    andb $1, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vinserti64x4 $1, %ymm2, %zmm0, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k2
+; X32-NEXT:    vpmovm2b %k2, %zmm0
+; X32-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vpslldq {{.*#+}} xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[0,1,2,3,4]
+; X32-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm5 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm5, %ymm1, %ymm2, %ymm1
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm0, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $28, %ecx
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vpbroadcastd %xmm2, %xmm2
+; X32-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm5 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255]
+; X32-NEXT:    vpblendvb %ymm5, %ymm1, %ymm2, %ymm1
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm0, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
+; X32-NEXT:    vpmovm2b %k1, %zmm2
+; X32-NEXT:    vpslldq {{.*#+}} xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[0,1,2]
+; X32-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm5 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255]
+; X32-NEXT:    vpblendvb %ymm5, %ymm1, %ymm2, %ymm1
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm0, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $30, %ecx
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vpbroadcastw %xmm2, %xmm2
+; X32-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm2
+; X32-NEXT:    vpblendvb %ymm4, %ymm1, %ymm2, %ymm1
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm0, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrl $31, %eax
+; X32-NEXT:    kmovd %eax, %k1
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; X32-NEXT:    kshiftlq $1, %k0, %k0
+; X32-NEXT:    kshiftrq $1, %k0, %k0
+; X32-NEXT:    kshiftlq $63, %k1, %k1
+; X32-NEXT:    korq %k1, %k0, %k1
+; X32-NEXT:    vpbroadcastb %eax, %zmm3 {%k1}
+; X32-NEXT:    vmovdqa64 %zmm3, %zmm0
+; X32-NEXT:    popl %ebx
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm512_mask_set1_epi8:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovq %rdi, %k1
+; X64-NEXT:    vpbroadcastb %esi, %zmm0 {%k1}
+; X64-NEXT:    retq
+  entry:
+  %vecinit.i.i = insertelement <64 x i8> undef, i8 %__A, i32 0
+  %vecinit63.i.i = shufflevector <64 x i8> %vecinit.i.i, <64 x i8> undef, <64 x i32> zeroinitializer
+  %0 = bitcast <8 x i64> %__O to <64 x i8>
+  %1 = bitcast i64 %__M to <64 x i1>
+  %2 = select <64 x i1> %1, <64 x i8> %vecinit63.i.i, <64 x i8> %0
+  %3 = bitcast <64 x i8> %2 to <8 x i64>
+  ret <8 x i64> %3
+}
+
+define <8 x i64> @test_mm512_maskz_set1_epi8(i64 %__M, i8 signext %__A)  {
+; X32-LABEL: test_mm512_maskz_set1_epi8:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    pushl %ebx
+; X32-NEXT:  .Lcfi2:
+; X32-NEXT:    .cfi_def_cfa_offset 8
+; X32-NEXT:  .Lcfi3:
+; X32-NEXT:    .cfi_offset %ebx, -8
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andb $2, %cl
+; X32-NEXT:    shrb %cl
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vpsllw $8, %xmm0, %xmm0
+; X32-NEXT:    kmovd %eax, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andb $15, %cl
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $2, %dl
+; X32-NEXT:    kmovd %edx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vpbroadcastw %xmm1, %xmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm1[0,1,2,3],zmm0[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    shrb $3, %cl
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vpslld $24, %xmm1, %xmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm1[0,1,2,3],zmm0[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $4, %cl
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vpbroadcastd %xmm1, %xmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm1[0,1,2,3],zmm0[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $5, %cl
+; X32-NEXT:    andb $1, %cl
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vpsllq $40, %xmm1, %xmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm1[0,1,2,3],zmm0[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $6, %cl
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vpbroadcastw %xmm1, %xmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm1[0,1,2,3],zmm0[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $7, %cl
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vpsllq $56, %xmm1, %xmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm0, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm1[0,1,2,3],zmm0[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movb %ah, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastq %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    andb $2, %cl
+; X32-NEXT:    shrb %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4,5,6]
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movb %ah, %cl
+; X32-NEXT:    andb $15, %cl
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $2, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrb $3, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4]
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andl $61440, %ecx # imm = 0xF000
+; X32-NEXT:    shrl $12, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastd %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $13, %ecx
+; X32-NEXT:    andb $1, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2]
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andl $49152, %ecx # imm = 0xC000
+; X32-NEXT:    shrl $14, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andl $32768, %ecx # imm = 0x8000
+; X32-NEXT:    shrl $15, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0]
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $16, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    andb $2, %dl
+; X32-NEXT:    shrb %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllw $8, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    andb $15, %dl
+; X32-NEXT:    movl %edx, %ebx
+; X32-NEXT:    shrb $2, %bl
+; X32-NEXT:    kmovd %ebx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrb $3, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslld $24, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $4, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastd %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $5, %dl
+; X32-NEXT:    andb $1, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllq $40, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $6, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrb $7, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllq $56, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm6 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm6, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $24, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastq %xmm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm2 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    andb $2, %dl
+; X32-NEXT:    shrb %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4,5,6]
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm5 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm5, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm4[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    andb $15, %cl
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $2, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm1 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm1, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm4[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrb $3, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4]
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm1 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm1, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm4[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $28, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastd %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm1 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255]
+; X32-NEXT:    vpblendvb %ymm1, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm4[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $29, %ecx
+; X32-NEXT:    andb $1, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2]
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vmovdqa {{.*#+}} ymm3 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255]
+; X32-NEXT:    vpblendvb %ymm3, %ymm0, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm1[0,1,2,3],zmm0[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $30, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vmovdqa {{.*#+}} ymm3 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255]
+; X32-NEXT:    vpblendvb %ymm3, %ymm0, %ymm1, %ymm1
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm1[0,1,2,3],zmm0[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrl $31, %eax
+; X32-NEXT:    kmovd %eax, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0]
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0]
+; X32-NEXT:    vpblendvb %ymm7, %ymm1, %ymm0, %ymm0
+; X32-NEXT:    vshufi64x2 {{.*#+}} zmm0 = zmm0[0,1,2,3],zmm1[4,5,6,7]
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    kmovd %eax, %k1
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
+; X32-NEXT:    vpmovm2b %k1, %zmm7
+; X32-NEXT:    vmovdqa {{.*#+}} ymm4 = [0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm4, %ymm1, %ymm7, %ymm1
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm0, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andb $2, %cl
+; X32-NEXT:    shrb %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllw $8, %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andb $15, %cl
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $2, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrb $3, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslld $24, %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $4, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastd %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $5, %cl
+; X32-NEXT:    andb $1, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllq $40, %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $6, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrb $7, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllq $56, %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movb %ah, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastq %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    andb $2, %cl
+; X32-NEXT:    shrb %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4,5,6]
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movb %ah, %cl
+; X32-NEXT:    andb $15, %cl
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $2, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrb $3, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4]
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andl $61440, %ecx # imm = 0xF000
+; X32-NEXT:    shrl $12, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastd %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $13, %ecx
+; X32-NEXT:    andb $1, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2]
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andl $49152, %ecx # imm = 0xC000
+; X32-NEXT:    shrl $14, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    andl $32768, %ecx # imm = 0x8000
+; X32-NEXT:    shrl $15, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0]
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $16, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    andb $2, %dl
+; X32-NEXT:    shrb %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllw $8, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    andb $15, %dl
+; X32-NEXT:    movl %edx, %ebx
+; X32-NEXT:    shrb $2, %bl
+; X32-NEXT:    kmovd %ebx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrb $3, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpslld $24, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $4, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastd %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $5, %dl
+; X32-NEXT:    andb $1, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllq $40, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $6, %dl
+; X32-NEXT:    kmovd %edx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastw %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vmovdqa {{.*#+}} ymm7 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm7, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    shrb $7, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpsllq $56, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vpblendvb %ymm6, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $24, %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpmovm2b %k1, %zmm0
+; X32-NEXT:    vpbroadcastq %xmm0, %ymm0
+; X32-NEXT:    vpmovm2b %k0, %zmm1
+; X32-NEXT:    vextracti64x4 $1, %zmm1, %ymm4
+; X32-NEXT:    vpblendvb %ymm2, %ymm4, %ymm0, %ymm0
+; X32-NEXT:    vinserti64x4 $1, %ymm0, %zmm1, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    andb $2, %dl
+; X32-NEXT:    shrb %dl
+; X32-NEXT:    kmovd %edx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vpslldq {{.*#+}} xmm0 = zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4,5,6]
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    vextracti64x4 $1, %zmm2, %ymm1
+; X32-NEXT:    vpblendvb %ymm5, %ymm1, %ymm0, %ymm1
+; X32-NEXT:    andb $15, %cl
+; X32-NEXT:    movl %ecx, %edx
+; X32-NEXT:    shrb $2, %dl
+; X32-NEXT:    kmovd %edx, %k0
+; X32-NEXT:    shrb $3, %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $29, %ecx
+; X32-NEXT:    andb $1, %cl
+; X32-NEXT:    kmovd %ecx, %k2
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm2, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k3
+; X32-NEXT:    vpmovm2b %k3, %zmm0
+; X32-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vpbroadcastw %xmm2, %xmm2
+; X32-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm4 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm4, %ymm1, %ymm2, %ymm1
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm0, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
+; X32-NEXT:    vpmovm2b %k1, %zmm2
+; X32-NEXT:    vpslldq {{.*#+}} xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[0,1,2,3,4]
+; X32-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm4 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255,255]
+; X32-NEXT:    vpblendvb %ymm4, %ymm1, %ymm2, %ymm1
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm0, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $28, %ecx
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vpbroadcastd %xmm2, %xmm2
+; X32-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm4 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255,255]
+; X32-NEXT:    vpblendvb %ymm4, %ymm1, %ymm2, %ymm1
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm0, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
+; X32-NEXT:    vpmovm2b %k2, %zmm2
+; X32-NEXT:    vpslldq {{.*#+}} xmm2 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm2[0,1,2]
+; X32-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm2
+; X32-NEXT:    vmovdqa {{.*#+}} ymm4 = [255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,0,255,255]
+; X32-NEXT:    vpblendvb %ymm4, %ymm1, %ymm2, %ymm1
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm0, %zmm0
+; X32-NEXT:    vpmovb2m %zmm0, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm0
+; X32-NEXT:    vextracti64x4 $1, %zmm0, %ymm1
+; X32-NEXT:    movl %eax, %ecx
+; X32-NEXT:    shrl $30, %ecx
+; X32-NEXT:    kmovd %ecx, %k0
+; X32-NEXT:    vpmovm2b %k0, %zmm2
+; X32-NEXT:    vpbroadcastw %xmm2, %xmm2
+; X32-NEXT:    vinserti128 $1, %xmm2, %ymm0, %ymm2
+; X32-NEXT:    vpblendvb %ymm3, %ymm1, %ymm2, %ymm1
+; X32-NEXT:    vinserti64x4 $1, %ymm1, %zmm0, %zmm0
+; X32-NEXT:    shrl $31, %eax
+; X32-NEXT:    kmovd %eax, %k0
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; X32-NEXT:    vpmovb2m %zmm0, %k1
+; X32-NEXT:    kshiftlq $1, %k1, %k1
+; X32-NEXT:    kshiftrq $1, %k1, %k1
+; X32-NEXT:    kshiftlq $63, %k0, %k0
+; X32-NEXT:    korq %k0, %k1, %k1
+; X32-NEXT:    vpbroadcastb %eax, %zmm0 {%k1} {z}
+; X32-NEXT:    popl %ebx
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm512_maskz_set1_epi8:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovq %rdi, %k1
+; X64-NEXT:    vpbroadcastb %esi, %zmm0 {%k1} {z}
+; X64-NEXT:    retq
+  entry:
+  %vecinit.i.i = insertelement <64 x i8> undef, i8 %__A, i32 0
+  %vecinit63.i.i = shufflevector <64 x i8> %vecinit.i.i, <64 x i8> undef, <64 x i32> zeroinitializer
+  %0 = bitcast i64 %__M to <64 x i1>
+  %1 = select <64 x i1> %0, <64 x i8> %vecinit63.i.i, <64 x i8> zeroinitializer
+  %2 = bitcast <64 x i8> %1 to <8 x i64>
+  ret <8 x i64> %2
+}
+
+define <8 x i64> @test_mm512_mask_set1_epi16(<8 x i64> %__O, i32 %__M, i16 signext %__A)  {
+; X32-LABEL: test_mm512_mask_set1_epi16:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movzwl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpbroadcastw %eax, %zmm0 {%k1}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm512_mask_set1_epi16:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovd %edi, %k1
+; X64-NEXT:    vpbroadcastw %esi, %zmm0 {%k1}
+; X64-NEXT:    retq
+  entry:
+  %vecinit.i.i = insertelement <32 x i16> undef, i16 %__A, i32 0
+  %vecinit31.i.i = shufflevector <32 x i16> %vecinit.i.i, <32 x i16> undef, <32 x i32> zeroinitializer
+  %0 = bitcast <8 x i64> %__O to <32 x i16>
+  %1 = bitcast i32 %__M to <32 x i1>
+  %2 = select <32 x i1> %1, <32 x i16> %vecinit31.i.i, <32 x i16> %0
+  %3 = bitcast <32 x i16> %2 to <8 x i64>
+  ret <8 x i64> %3
+}
+
+define <8 x i64> @test_mm512_maskz_set1_epi16(i32 %__M, i16 signext %__A)  {
+; X32-LABEL: test_mm512_maskz_set1_epi16:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movzwl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpbroadcastw %eax, %zmm0 {%k1} {z}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm512_maskz_set1_epi16:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovd %edi, %k1
+; X64-NEXT:    vpbroadcastw %esi, %zmm0 {%k1} {z}
+; X64-NEXT:    retq
+  entry:
+  %vecinit.i.i = insertelement <32 x i16> undef, i16 %__A, i32 0
+  %vecinit31.i.i = shufflevector <32 x i16> %vecinit.i.i, <32 x i16> undef, <32 x i32> zeroinitializer
+  %0 = bitcast i32 %__M to <32 x i1>
+  %1 = select <32 x i1> %0, <32 x i16> %vecinit31.i.i, <32 x i16> zeroinitializer
+  %2 = bitcast <32 x i16> %1 to <8 x i64>
+  ret <8 x i64> %2
+}
+
 define <8 x i64> @test_mm512_broadcastb_epi8(<2 x i64> %a0) {
 ; X32-LABEL: test_mm512_broadcastb_epi8:
 ; X32:       # BB#0:

diff --git a/test/CodeGen/X86/avx512bw-intrinsics-upgrade.ll b/test/CodeGen/X86/avx512bw-intrinsics-upgrade.ll
index cb0061188ae6eb6a8db708b1ff63fb6af4c2ff92..ec9481568b7297457fb083de6ddd3bedd4d39119 100644 (file)
--- a/test/CodeGen/X86/avx512bw-intrinsics-upgrade.ll
+++ b/test/CodeGen/X86/avx512bw-intrinsics-upgrade.ll
@@ -2,6 +2,68 @@
 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=knl -mattr=+avx512bw | FileCheck %s --check-prefix=ALL --check-prefix=AVX512BW
 ; RUN: llc < %s -mtriple=i386-unknown-linux-gnu -mcpu=knl -mattr=+avx512bw | FileCheck %s --check-prefix=ALL --check-prefix=AVX512F-32
 
+
+declare <64 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.512(i8, <64 x i8>, i64)
+
+  define <64 x i8>@test_int_x86_avx512_mask_pbroadcast_b_gpr_512(i8 %x0, <64 x i8> %x1, i64 %mask) {
+; AVX512BW-LABEL: test_int_x86_avx512_mask_pbroadcast_b_gpr_512:
+; AVX512BW:       ## BB#0:
+; AVX512BW-NEXT:    vpbroadcastb %edi, %zmm1
+; AVX512BW-NEXT:    kmovq %rsi, %k1
+; AVX512BW-NEXT:    vpbroadcastb %edi, %zmm0 {%k1}
+; AVX512BW-NEXT:    vpaddb %zmm0, %zmm1, %zmm0
+; AVX512BW-NEXT:    vpbroadcastb %edi, %zmm1 {%k1} {z}
+; AVX512BW-NEXT:    vpaddb %zmm0, %zmm1, %zmm0
+; AVX512BW-NEXT:    retq
+;
+; AVX512F-32-LABEL: test_int_x86_avx512_mask_pbroadcast_b_gpr_512:
+; AVX512F-32:       # BB#0:
+; AVX512F-32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; AVX512F-32-NEXT:    vpbroadcastb %eax, %zmm1
+; AVX512F-32-NEXT:    kmovq {{[0-9]+}}(%esp), %k1
+; AVX512F-32-NEXT:    vpbroadcastb %eax, %zmm0 {%k1}
+; AVX512F-32-NEXT:    vpaddb %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpbroadcastb %eax, %zmm1 {%k1} {z}
+; AVX512F-32-NEXT:    vpaddb %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    retl
+    %res = call <64 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.512(i8 %x0, <64 x i8> %x1, i64 -1)
+    %res1 = call <64 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.512(i8 %x0, <64 x i8> %x1, i64 %mask)
+    %res2 = call <64 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.512(i8 %x0, <64 x i8> zeroinitializer, i64 %mask)
+    %res3 = add <64 x i8> %res, %res1
+    %res4 = add <64 x i8> %res2, %res3
+    ret <64 x i8> %res4
+  }
+
+declare <32 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.512(i16, <32 x i16>, i32)
+  define <32 x i16>@test_int_x86_avx512_mask_pbroadcast_w_gpr_512(i16 %x0, <32 x i16> %x1, i32 %mask) {
+; AVX512BW-LABEL: test_int_x86_avx512_mask_pbroadcast_w_gpr_512:
+; AVX512BW:       ## BB#0:
+; AVX512BW-NEXT:    vpbroadcastw %edi, %zmm1
+; AVX512BW-NEXT:    kmovd %esi, %k1
+; AVX512BW-NEXT:    vpbroadcastw %edi, %zmm0 {%k1}
+; AVX512BW-NEXT:    vpaddw %zmm0, %zmm1, %zmm0
+; AVX512BW-NEXT:    vpbroadcastw %edi, %zmm1 {%k1} {z}
+; AVX512BW-NEXT:    vpaddw %zmm0, %zmm1, %zmm0
+; AVX512BW-NEXT:    retq
+;
+; AVX512F-32-LABEL: test_int_x86_avx512_mask_pbroadcast_w_gpr_512:
+; AVX512F-32:       # BB#0:
+; AVX512F-32-NEXT:    movzwl {{[0-9]+}}(%esp), %eax
+; AVX512F-32-NEXT:    vpbroadcastw %eax, %zmm1
+; AVX512F-32-NEXT:    kmovd {{[0-9]+}}(%esp), %k1
+; AVX512F-32-NEXT:    vpbroadcastw %eax, %zmm0 {%k1}
+; AVX512F-32-NEXT:    vpaddw %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpbroadcastw %eax, %zmm1 {%k1} {z}
+; AVX512F-32-NEXT:    vpaddw %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    retl
+    %res = call <32 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.512(i16 %x0, <32 x i16> %x1, i32 -1)
+    %res1 = call <32 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.512(i16 %x0, <32 x i16> %x1, i32 %mask)
+   %res2 = call <32 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.512(i16 %x0, <32 x i16> zeroinitializer, i32 %mask)
+    %res3 = add <32 x i16> %res, %res1
+   %res4 = add <32 x i16> %res2, %res3
+    ret <32 x i16> %res4
+ }
+
 declare void @llvm.x86.avx512.mask.storeu.b.512(i8*, <64 x i8>, i64)
 
 define void@test_int_x86_avx512_mask_storeu_b_512(i8* %ptr1, i8* %ptr2, <64 x i8> %x1, i64 %x2) {

diff --git a/test/CodeGen/X86/avx512bw-intrinsics.ll b/test/CodeGen/X86/avx512bw-intrinsics.ll
index e9962a370af3e5366ed10884a6f413a0455899f4..f499ae50f54f1f056b4233f64c4d085b365ae1f1 100644 (file)
--- a/test/CodeGen/X86/avx512bw-intrinsics.ll
+++ b/test/CodeGen/X86/avx512bw-intrinsics.ll
@@ -1815,71 +1815,6 @@ define i32@test_int_x86_avx512_ptestnm_w_512(<32 x i16> %x0, <32 x i16> %x1, i32
   ret i32 %res2
 }
 
-declare <64 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.512(i8, <64 x i8>, i64)
-
-define <64 x i8>@test_int_x86_avx512_mask_pbroadcast_b_gpr_512(i8 %x0, <64 x i8> %x1, i64 %mask) {
-; AVX512BW-LABEL: test_int_x86_avx512_mask_pbroadcast_b_gpr_512:
-; AVX512BW:       ## BB#0:
-; AVX512BW-NEXT:    kmovq %rsi, %k1
-; AVX512BW-NEXT:    vpbroadcastb %edi, %zmm1 {%k1} {z}
-; AVX512BW-NEXT:    vpbroadcastb %edi, %zmm0 {%k1}
-; AVX512BW-NEXT:    vpbroadcastb %edi, %zmm2
-; AVX512BW-NEXT:    vpaddb %zmm0, %zmm2, %zmm0
-; AVX512BW-NEXT:    vpaddb %zmm0, %zmm1, %zmm0
-; AVX512BW-NEXT:    retq
-;
-; AVX512F-32-LABEL: test_int_x86_avx512_mask_pbroadcast_b_gpr_512:
-; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    movb {{[0-9]+}}(%esp), %al
-; AVX512F-32-NEXT:    kmovd {{[0-9]+}}(%esp), %k0
-; AVX512F-32-NEXT:    kmovd {{[0-9]+}}(%esp), %k1
-; AVX512F-32-NEXT:    kunpckdq %k0, %k1, %k1
-; AVX512F-32-NEXT:    vpbroadcastb %eax, %zmm1 {%k1} {z}
-; AVX512F-32-NEXT:    vpbroadcastb %eax, %zmm0 {%k1}
-; AVX512F-32-NEXT:    vpbroadcastb %eax, %zmm2
-; AVX512F-32-NEXT:    vpaddb %zmm0, %zmm2, %zmm0
-; AVX512F-32-NEXT:    vpaddb %zmm0, %zmm1, %zmm0
-; AVX512F-32-NEXT:    retl
-  %res = call <64 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.512(i8 %x0, <64 x i8> %x1, i64 -1)
-  %res1 = call <64 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.512(i8 %x0, <64 x i8> %x1, i64 %mask)
-  %res2 = call <64 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.512(i8 %x0, <64 x i8> zeroinitializer, i64 %mask)
-  %res3 = add <64 x i8> %res, %res1
-  %res4 = add <64 x i8> %res2, %res3
-  ret <64 x i8> %res4
-}
-
-declare <32 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.512(i16, <32 x i16>, i32)
-
-define <32 x i16>@test_int_x86_avx512_mask_pbroadcast_w_gpr_512(i16 %x0, <32 x i16> %x1, i32 %mask) {
-; AVX512BW-LABEL: test_int_x86_avx512_mask_pbroadcast_w_gpr_512:
-; AVX512BW:       ## BB#0:
-; AVX512BW-NEXT:    kmovd %esi, %k1
-; AVX512BW-NEXT:    vpbroadcastw %edi, %zmm1 {%k1} {z}
-; AVX512BW-NEXT:    vpbroadcastw %edi, %zmm0 {%k1}
-; AVX512BW-NEXT:    vpbroadcastw %edi, %zmm2
-; AVX512BW-NEXT:    vpaddw %zmm0, %zmm2, %zmm0
-; AVX512BW-NEXT:    vpaddw %zmm0, %zmm1, %zmm0
-; AVX512BW-NEXT:    retq
-;
-; AVX512F-32-LABEL: test_int_x86_avx512_mask_pbroadcast_w_gpr_512:
-; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    kmovd {{[0-9]+}}(%esp), %k1
-; AVX512F-32-NEXT:    movzwl {{[0-9]+}}(%esp), %eax
-; AVX512F-32-NEXT:    vpbroadcastw %eax, %zmm1 {%k1} {z}
-; AVX512F-32-NEXT:    vpbroadcastw %eax, %zmm0 {%k1}
-; AVX512F-32-NEXT:    vpbroadcastw %eax, %zmm2
-; AVX512F-32-NEXT:    vpaddw %zmm0, %zmm2, %zmm0
-; AVX512F-32-NEXT:    vpaddw %zmm0, %zmm1, %zmm0
-; AVX512F-32-NEXT:    retl
-  %res = call <32 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.512(i16 %x0, <32 x i16> %x1, i32 -1)
-  %res1 = call <32 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.512(i16 %x0, <32 x i16> %x1, i32 %mask)
-  %res2 = call <32 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.512(i16 %x0, <32 x i16> zeroinitializer, i32 %mask)
-  %res3 = add <32 x i16> %res, %res1
-  %res4 = add <32 x i16> %res2, %res3
-  ret <32 x i16> %res4
-}
-
-
 define <32 x i16> @test_x86_avx512_psll_w_512(<32 x i16> %a0, <8 x i16> %a1) {
 ; AVX512BW-LABEL: test_x86_avx512_psll_w_512:
 ; AVX512BW:       ## BB#0:

diff --git a/test/CodeGen/X86/avx512bwvl-intrinsics-fast-isel.ll b/test/CodeGen/X86/avx512bwvl-intrinsics-fast-isel.ll
index a911f6465e69f1548a38230fcd67735cd78a8b14..4edfb539839d60ea0bc278857fc428391e4065ba 100644 (file)
--- a/test/CodeGen/X86/avx512bwvl-intrinsics-fast-isel.ll
+++ b/test/CodeGen/X86/avx512bwvl-intrinsics-fast-isel.ll
@@ -4,6 +4,195 @@
 
 ; NOTE: This should use IR equivalent to what is generated by clang/test/CodeGen/avx512vlbw-builtins.c
 
+define <2 x i64> @test_mm_mask_set1_epi8(<2 x i64> %__O, i16 zeroext %__M, i8 signext %__A) local_unnamed_addr #0 {
+; X32-LABEL: test_mm_mask_set1_epi8:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; X32-NEXT:    movzwl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpbroadcastb %eax, %xmm0 {%k1}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm_mask_set1_epi8:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovd %edi, %k1
+; X64-NEXT:    vpbroadcastb %esi, %xmm0 {%k1}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <16 x i8> undef, i8 %__A, i32 0
+  %vecinit15.i.i = shufflevector <16 x i8> %vecinit.i.i, <16 x i8> undef, <16 x i32> zeroinitializer
+  %0 = bitcast <2 x i64> %__O to <16 x i8>
+  %1 = bitcast i16 %__M to <16 x i1>
+  %2 = select <16 x i1> %1, <16 x i8> %vecinit15.i.i, <16 x i8> %0
+  %3 = bitcast <16 x i8> %2 to <2 x i64>
+  ret <2 x i64> %3
+}
+
+define <2 x i64> @test_mm_maskz_set1_epi8(i16 zeroext %__M, i8 signext %__A)  {
+; X32-LABEL: test_mm_maskz_set1_epi8:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; X32-NEXT:    movzwl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpbroadcastb %eax, %xmm0 {%k1} {z}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm_maskz_set1_epi8:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovd %edi, %k1
+; X64-NEXT:    vpbroadcastb %esi, %xmm0 {%k1} {z}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <16 x i8> undef, i8 %__A, i32 0
+  %vecinit15.i.i = shufflevector <16 x i8> %vecinit.i.i, <16 x i8> undef, <16 x i32> zeroinitializer
+  %0 = bitcast i16 %__M to <16 x i1>
+  %1 = select <16 x i1> %0, <16 x i8> %vecinit15.i.i, <16 x i8> zeroinitializer
+  %2 = bitcast <16 x i8> %1 to <2 x i64>
+  ret <2 x i64> %2
+}
+
+define <4 x i64> @test_mm256_mask_set1_epi8(<4 x i64> %__O, i32 %__M, i8 signext %__A){
+; X32-LABEL: test_mm256_mask_set1_epi8:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpbroadcastb %eax, %ymm0 {%k1}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm256_mask_set1_epi8:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovd %edi, %k1
+; X64-NEXT:    vpbroadcastb %esi, %ymm0 {%k1}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <32 x i8> undef, i8 %__A, i32 0
+  %vecinit31.i.i = shufflevector <32 x i8> %vecinit.i.i, <32 x i8> undef, <32 x i32> zeroinitializer
+  %0 = bitcast <4 x i64> %__O to <32 x i8>
+  %1 = bitcast i32 %__M to <32 x i1>
+  %2 = select <32 x i1> %1, <32 x i8> %vecinit31.i.i, <32 x i8> %0
+  %3 = bitcast <32 x i8> %2 to <4 x i64>
+  ret <4 x i64> %3
+}
+
+define <4 x i64> @test_mm256_maskz_set1_epi8(i32 %__M, i8 signext %__A)  {
+; X32-LABEL: test_mm256_maskz_set1_epi8:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpbroadcastb %eax, %ymm0 {%k1} {z}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm256_maskz_set1_epi8:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovd %edi, %k1
+; X64-NEXT:    vpbroadcastb %esi, %ymm0 {%k1} {z}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <32 x i8> undef, i8 %__A, i32 0
+  %vecinit31.i.i = shufflevector <32 x i8> %vecinit.i.i, <32 x i8> undef, <32 x i32> zeroinitializer
+  %0 = bitcast i32 %__M to <32 x i1>
+  %1 = select <32 x i1> %0, <32 x i8> %vecinit31.i.i, <32 x i8> zeroinitializer
+  %2 = bitcast <32 x i8> %1 to <4 x i64>
+  ret <4 x i64> %2
+}
+
+define <4 x i64> @test_mm256_mask_set1_epi16(<4 x i64> %__O, i16 zeroext %__M, i16 signext %__A)  {
+; X32-LABEL: test_mm256_mask_set1_epi16:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movzwl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movzwl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpbroadcastw %eax, %ymm0 {%k1}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm256_mask_set1_epi16:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovd %edi, %k1
+; X64-NEXT:    vpbroadcastw %esi, %ymm0 {%k1}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <16 x i16> undef, i16 %__A, i32 0
+  %vecinit15.i.i = shufflevector <16 x i16> %vecinit.i.i, <16 x i16> undef, <16 x i32> zeroinitializer
+  %0 = bitcast <4 x i64> %__O to <16 x i16>
+  %1 = bitcast i16 %__M to <16 x i1>
+  %2 = select <16 x i1> %1, <16 x i16> %vecinit15.i.i, <16 x i16> %0
+  %3 = bitcast <16 x i16> %2 to <4 x i64>
+  ret <4 x i64> %3
+}
+
+define <4 x i64> @test_mm256_maskz_set1_epi16(i16 zeroext %__M, i16 signext %__A) {
+; X32-LABEL: test_mm256_maskz_set1_epi16:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movzwl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movzwl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpbroadcastw %eax, %ymm0 {%k1} {z}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm256_maskz_set1_epi16:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovd %edi, %k1
+; X64-NEXT:    vpbroadcastw %esi, %ymm0 {%k1} {z}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <16 x i16> undef, i16 %__A, i32 0
+  %vecinit15.i.i = shufflevector <16 x i16> %vecinit.i.i, <16 x i16> undef, <16 x i32> zeroinitializer
+  %0 = bitcast i16 %__M to <16 x i1>
+  %1 = select <16 x i1> %0, <16 x i16> %vecinit15.i.i, <16 x i16> zeroinitializer
+  %2 = bitcast <16 x i16> %1 to <4 x i64>
+  ret <4 x i64> %2
+}
+
+define <2 x i64> @test_mm_mask_set1_epi16(<2 x i64> %__O, i8 zeroext %__M, i16 signext %__A) {
+; X32-LABEL: test_mm_mask_set1_epi16:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movzwl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpbroadcastw %eax, %xmm0 {%k1}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm_mask_set1_epi16:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovd %edi, %k1
+; X64-NEXT:    vpbroadcastw %esi, %xmm0 {%k1}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <8 x i16> undef, i16 %__A, i32 0
+  %vecinit7.i.i = shufflevector <8 x i16> %vecinit.i.i, <8 x i16> undef, <8 x i32> zeroinitializer
+  %0 = bitcast <2 x i64> %__O to <8 x i16>
+  %1 = bitcast i8 %__M to <8 x i1>
+  %2 = select <8 x i1> %1, <8 x i16> %vecinit7.i.i, <8 x i16> %0
+  %3 = bitcast <8 x i16> %2 to <2 x i64>
+  ret <2 x i64> %3
+}
+
+define <2 x i64> @test_mm_maskz_set1_epi16(i8 zeroext %__M, i16 signext %__A) {
+; X32-LABEL: test_mm_maskz_set1_epi16:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movzwl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %cl
+; X32-NEXT:    kmovd %ecx, %k1
+; X32-NEXT:    vpbroadcastw %eax, %xmm0 {%k1} {z}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm_maskz_set1_epi16:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovd %edi, %k1
+; X64-NEXT:    vpbroadcastw %esi, %xmm0 {%k1} {z}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <8 x i16> undef, i16 %__A, i32 0
+  %vecinit7.i.i = shufflevector <8 x i16> %vecinit.i.i, <8 x i16> undef, <8 x i32> zeroinitializer
+  %0 = bitcast i8 %__M to <8 x i1>
+  %1 = select <8 x i1> %0, <8 x i16> %vecinit7.i.i, <8 x i16> zeroinitializer
+  %2 = bitcast <8 x i16> %1 to <2 x i64>
+  ret <2 x i64> %2
+}
+
+
 define <2 x i64> @test_mm_broadcastb_epi8(<2 x i64> %a0) {
 ; X32-LABEL: test_mm_broadcastb_epi8:
 ; X32:       # BB#0:

diff --git a/test/CodeGen/X86/avx512bwvl-intrinsics-upgrade.ll b/test/CodeGen/X86/avx512bwvl-intrinsics-upgrade.ll
index beac8b81bc59d910f1ee386d7d1d5923ed983ed3..16497701c90637bd8a1ba77554a8c14651545cf2 100644 (file)
--- a/test/CodeGen/X86/avx512bwvl-intrinsics-upgrade.ll
+++ b/test/CodeGen/X86/avx512bwvl-intrinsics-upgrade.ll
@@ -1,6 +1,90 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=knl -mattr=+avx512bw -mattr=+avx512vl --show-mc-encoding| FileCheck %s
 
+declare <16 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.128(i8, <16 x i8>, i16)
+
+define <16 x i8>@test_int_x86_avx512_mask_pbroadcast_b_gpr_128(i8 %x0, <16 x i8> %x1, i16 %mask) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_b_gpr_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vpbroadcastb %edi, %xmm1 ## encoding: [0x62,0xf2,0x7d,0x08,0x7a,0xcf]
+; CHECK-NEXT:    kmovd %esi, %k1 ## encoding: [0xc5,0xfb,0x92,0xce]
+; CHECK-NEXT:    vpbroadcastb %edi, %xmm0 {%k1} ## encoding: [0x62,0xf2,0x7d,0x09,0x7a,0xc7]
+; CHECK-NEXT:    vpaddb %xmm0, %xmm1, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0xfc,0xc0]
+; CHECK-NEXT:    vpbroadcastb %edi, %xmm1 {%k1} {z} ## encoding: [0x62,0xf2,0x7d,0x89,0x7a,0xcf]
+; CHECK-NEXT:    vpaddb %xmm0, %xmm1, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0xfc,0xc0]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <16 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.128(i8 %x0, <16 x i8> %x1, i16 -1)
+  %res1 = call <16 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.128(i8 %x0, <16 x i8> %x1, i16 %mask)
+  %res2 = call <16 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.128(i8 %x0, <16 x i8> zeroinitializer, i16 %mask)
+  %res3 = add <16 x i8> %res, %res1
+  %res4 = add <16 x i8> %res2, %res3
+  ret <16 x i8> %res4
+}
+
+
+declare <8 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.128(i16, <8 x i16>, i8)
+
+define <8 x i16>@test_int_x86_avx512_mask_pbroadcast_w_gpr_128(i16 %x0, <8 x i16> %x1, i8 %mask) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_w_gpr_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vpbroadcastw %edi, %xmm1 ## encoding: [0x62,0xf2,0x7d,0x08,0x7b,0xcf]
+; CHECK-NEXT:    kmovd %esi, %k1 ## encoding: [0xc5,0xfb,0x92,0xce]
+; CHECK-NEXT:    vpbroadcastw %edi, %xmm0 {%k1} ## encoding: [0x62,0xf2,0x7d,0x09,0x7b,0xc7]
+; CHECK-NEXT:    vpaddw %xmm0, %xmm1, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0xfd,0xc0]
+; CHECK-NEXT:    vpbroadcastw %edi, %xmm1 {%k1} {z} ## encoding: [0x62,0xf2,0x7d,0x89,0x7b,0xcf]
+; CHECK-NEXT:    vpaddw %xmm0, %xmm1, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0xfd,0xc0]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <8 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.128(i16 %x0, <8 x i16> %x1, i8 -1)
+  %res1 = call <8 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.128(i16 %x0, <8 x i16> %x1, i8 %mask)
+  %res2 = call <8 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.128(i16 %x0, <8 x i16> zeroinitializer, i8 %mask)
+  %res3 = add <8 x i16> %res, %res1
+  %res4 = add <8 x i16> %res2, %res3
+  ret <8 x i16> %res4
+}
+
+
+ declare <32 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.256(i8, <32 x i8>, i32)
+
+  define <32 x i8>@test_int_x86_avx512_mask_pbroadcast_b_gpr_256(i8 %x0, <32 x i8> %x1, i32 %mask) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_b_gpr_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vpbroadcastb %edi, %ymm1 ## encoding: [0x62,0xf2,0x7d,0x28,0x7a,0xcf]
+; CHECK-NEXT:    kmovd %esi, %k1 ## encoding: [0xc5,0xfb,0x92,0xce]
+; CHECK-NEXT:    vpbroadcastb %edi, %ymm0 {%k1} ## encoding: [0x62,0xf2,0x7d,0x29,0x7a,0xc7]
+; CHECK-NEXT:    vpaddb %ymm0, %ymm1, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf5,0xfc,0xc0]
+; CHECK-NEXT:    vpbroadcastb %edi, %ymm1 {%k1} {z} ## encoding: [0x62,0xf2,0x7d,0xa9,0x7a,0xcf]
+; CHECK-NEXT:    vpaddb %ymm0, %ymm1, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf5,0xfc,0xc0]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+    %res = call <32 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.256(i8 %x0, <32 x i8> %x1, i32 -1)
+    %res1 = call <32 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.256(i8 %x0, <32 x i8> %x1, i32 %mask)
+    %res2 = call <32 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.256(i8 %x0, <32 x i8> zeroinitializer, i32 %mask)
+    %res3 = add <32 x i8> %res, %res1
+    %res4 = add <32 x i8> %res2, %res3
+    ret <32 x i8> %res4
+  }
+
+
+
+declare <16 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.256(i16, <16 x i16>, i16)
+
+  define <16 x i16>@test_int_x86_avx512_mask_pbroadcast_w_gpr_256(i16 %x0, <16 x i16> %x1, i16 %mask) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_w_gpr_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vpbroadcastw %edi, %ymm1 ## encoding: [0x62,0xf2,0x7d,0x28,0x7b,0xcf]
+; CHECK-NEXT:    kmovd %esi, %k1 ## encoding: [0xc5,0xfb,0x92,0xce]
+; CHECK-NEXT:    vpbroadcastw %edi, %ymm0 {%k1} ## encoding: [0x62,0xf2,0x7d,0x29,0x7b,0xc7]
+; CHECK-NEXT:    vpaddw %ymm0, %ymm1, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf5,0xfd,0xc0]
+; CHECK-NEXT:    vpbroadcastw %edi, %ymm1 {%k1} {z} ## encoding: [0x62,0xf2,0x7d,0xa9,0x7b,0xcf]
+; CHECK-NEXT:    vpaddw %ymm0, %ymm1, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf5,0xfd,0xc0]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+    %res = call <16 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.256(i16 %x0, <16 x i16> %x1, i16 -1)
+    %res1 = call <16 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.256(i16 %x0, <16 x i16> %x1, i16 %mask)
+    %res2 = call <16 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.256(i16 %x0, <16 x i16> zeroinitializer, i16 %mask)
+    %res3 = add <16 x i16> %res, %res1
+    %res4 = add <16 x i16> %res2, %res3
+    ret <16 x i16> %res4
+  }
+
 declare <32 x i8> @llvm.x86.avx512.pbroadcastb.256(<16 x i8>, <32 x i8>, i32)
 
 define <32 x i8>@test_int_x86_avx512_pbroadcastb_256(<16 x i8> %x0, <32 x i8> %x1, i32 %mask) {

diff --git a/test/CodeGen/X86/avx512bwvl-intrinsics.ll b/test/CodeGen/X86/avx512bwvl-intrinsics.ll
index 68fb257eaca6a58cd26859e14fed2486a62fe4cf..e2472bc443301ef2d7595e1fd632a4a4b24d943a 100644 (file)
--- a/test/CodeGen/X86/avx512bwvl-intrinsics.ll
+++ b/test/CodeGen/X86/avx512bwvl-intrinsics.ll
@@ -2665,82 +2665,4 @@ define i16@test_int_x86_avx512_ptestnm_w_256(<16 x i16> %x0, <16 x i16> %x1, i16
   ret i16 %res2
 }
 
-declare <32 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.256(i8, <32 x i8>, i32)
 
-define <32 x i8>@test_int_x86_avx512_mask_pbroadcast_b_gpr_256(i8 %x0, <32 x i8> %x1, i32 %mask) {
-; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_b_gpr_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovd %esi, %k1 ## encoding: [0xc5,0xfb,0x92,0xce]
-; CHECK-NEXT:    vpbroadcastb %edi, %ymm1 {%k1} {z} ## encoding: [0x62,0xf2,0x7d,0xa9,0x7a,0xcf]
-; CHECK-NEXT:    vpbroadcastb %edi, %ymm0 {%k1} ## encoding: [0x62,0xf2,0x7d,0x29,0x7a,0xc7]
-; CHECK-NEXT:    vpbroadcastb %edi, %ymm2 ## encoding: [0x62,0xf2,0x7d,0x28,0x7a,0xd7]
-; CHECK-NEXT:    vpaddb %ymm0, %ymm2, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xed,0xfc,0xc0]
-; CHECK-NEXT:    vpaddb %ymm0, %ymm1, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf5,0xfc,0xc0]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <32 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.256(i8 %x0, <32 x i8> %x1, i32 -1)
-  %res1 = call <32 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.256(i8 %x0, <32 x i8> %x1, i32 %mask)
-  %res2 = call <32 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.256(i8 %x0, <32 x i8> zeroinitializer, i32 %mask)
-  %res3 = add <32 x i8> %res, %res1
-  %res4 = add <32 x i8> %res2, %res3
-  ret <32 x i8> %res4
-}
-
-declare <16 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.128(i8, <16 x i8>, i16)
-
-define <16 x i8>@test_int_x86_avx512_mask_pbroadcast_b_gpr_128(i8 %x0, <16 x i8> %x1, i16 %mask) {
-; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_b_gpr_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovd %esi, %k1 ## encoding: [0xc5,0xfb,0x92,0xce]
-; CHECK-NEXT:    vpbroadcastb %edi, %xmm1 {%k1} {z} ## encoding: [0x62,0xf2,0x7d,0x89,0x7a,0xcf]
-; CHECK-NEXT:    vpbroadcastb %edi, %xmm0 {%k1} ## encoding: [0x62,0xf2,0x7d,0x09,0x7a,0xc7]
-; CHECK-NEXT:    vpbroadcastb %edi, %xmm2 ## encoding: [0x62,0xf2,0x7d,0x08,0x7a,0xd7]
-; CHECK-NEXT:    vpaddb %xmm0, %xmm2, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xe9,0xfc,0xc0]
-; CHECK-NEXT:    vpaddb %xmm0, %xmm1, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0xfc,0xc0]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <16 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.128(i8 %x0, <16 x i8> %x1, i16 -1)
-  %res1 = call <16 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.128(i8 %x0, <16 x i8> %x1, i16 %mask)
-  %res2 = call <16 x i8> @llvm.x86.avx512.mask.pbroadcast.b.gpr.128(i8 %x0, <16 x i8> zeroinitializer, i16 %mask)
-  %res3 = add <16 x i8> %res, %res1
-  %res4 = add <16 x i8> %res2, %res3
-  ret <16 x i8> %res4
-}
-
-declare <16 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.256(i16, <16 x i16>, i16)
-
-define <16 x i16>@test_int_x86_avx512_mask_pbroadcast_w_gpr_256(i16 %x0, <16 x i16> %x1, i16 %mask) {
-; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_w_gpr_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovd %esi, %k1 ## encoding: [0xc5,0xfb,0x92,0xce]
-; CHECK-NEXT:    vpbroadcastw %edi, %ymm1 {%k1} {z} ## encoding: [0x62,0xf2,0x7d,0xa9,0x7b,0xcf]
-; CHECK-NEXT:    vpbroadcastw %edi, %ymm0 {%k1} ## encoding: [0x62,0xf2,0x7d,0x29,0x7b,0xc7]
-; CHECK-NEXT:    vpbroadcastw %edi, %ymm2 ## encoding: [0x62,0xf2,0x7d,0x28,0x7b,0xd7]
-; CHECK-NEXT:    vpaddw %ymm0, %ymm2, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xed,0xfd,0xc0]
-; CHECK-NEXT:    vpaddw %ymm0, %ymm1, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf5,0xfd,0xc0]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <16 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.256(i16 %x0, <16 x i16> %x1, i16 -1)
-  %res1 = call <16 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.256(i16 %x0, <16 x i16> %x1, i16 %mask)
-  %res2 = call <16 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.256(i16 %x0, <16 x i16> zeroinitializer, i16 %mask)
-  %res3 = add <16 x i16> %res, %res1
-  %res4 = add <16 x i16> %res2, %res3
-  ret <16 x i16> %res4
-}
-
-declare <8 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.128(i16, <8 x i16>, i8)
-
-define <8 x i16>@test_int_x86_avx512_mask_pbroadcast_w_gpr_128(i16 %x0, <8 x i16> %x1, i8 %mask) {
-; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_w_gpr_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovd %esi, %k1 ## encoding: [0xc5,0xfb,0x92,0xce]
-; CHECK-NEXT:    vpbroadcastw %edi, %xmm1 {%k1} {z} ## encoding: [0x62,0xf2,0x7d,0x89,0x7b,0xcf]
-; CHECK-NEXT:    vpbroadcastw %edi, %xmm0 {%k1} ## encoding: [0x62,0xf2,0x7d,0x09,0x7b,0xc7]
-; CHECK-NEXT:    vpbroadcastw %edi, %xmm2 ## encoding: [0x62,0xf2,0x7d,0x08,0x7b,0xd7]
-; CHECK-NEXT:    vpaddw %xmm0, %xmm2, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xe9,0xfd,0xc0]
-; CHECK-NEXT:    vpaddw %xmm0, %xmm1, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0xfd,0xc0]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.128(i16 %x0, <8 x i16> %x1, i8 -1)
-  %res1 = call <8 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.128(i16 %x0, <8 x i16> %x1, i8 %mask)
-  %res2 = call <8 x i16> @llvm.x86.avx512.mask.pbroadcast.w.gpr.128(i16 %x0, <8 x i16> zeroinitializer, i8 %mask)
-  %res3 = add <8 x i16> %res, %res1
-  %res4 = add <8 x i16> %res2, %res3
-  ret <8 x i16> %res4
-}

diff --git a/test/CodeGen/X86/avx512vl-intrinsics-fast-isel.ll b/test/CodeGen/X86/avx512vl-intrinsics-fast-isel.ll
index f517e3b0d5d4a1c8b8524db47f3760ca10086d29..7d7964d7bc9f6b6e921198d0808f88afeffdd670 100644 (file)
--- a/test/CodeGen/X86/avx512vl-intrinsics-fast-isel.ll
+++ b/test/CodeGen/X86/avx512vl-intrinsics-fast-isel.ll
@@ -4,6 +4,207 @@
 
 ; NOTE: This should use IR equivalent to what is generated by clang/test/CodeGen/avx512vl-builtins.c
 
+define <2 x i64> @test_mm_mask_set1_epi32(<2 x i64> %__O, i8 zeroext %__M)  {
+; X32-LABEL: test_mm_mask_set1_epi32:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; X32-NEXT:    kmovw %eax, %k1
+; X32-NEXT:    vpbroadcastd {{\.LCPI.*}}, %xmm0 {%k1}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm_mask_set1_epi32:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovw %edi, %k1
+; X64-NEXT:    vpbroadcastd {{.*}}(%rip), %xmm0 {%k1}
+; X64-NEXT:    retq
+entry:
+  %0 = bitcast <2 x i64> %__O to <4 x i32>
+  %1 = bitcast i8 %__M to <8 x i1>
+  %extract.i = shufflevector <8 x i1> %1, <8 x i1> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %2 = select <4 x i1> %extract.i, <4 x i32> <i32 5, i32 5, i32 5, i32 5>, <4 x i32> %0
+  %3 = bitcast <4 x i32> %2 to <2 x i64>
+  ret <2 x i64> %3
+}
+
+define <2 x i64> @test_mm_maskz_set1_epi32(i8 zeroext %__M) {
+; X32-LABEL: test_mm_maskz_set1_epi32:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; X32-NEXT:    kmovw %eax, %k1
+; X32-NEXT:    vpbroadcastd {{\.LCPI.*}}, %xmm0 {%k1} {z}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm_maskz_set1_epi32:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovw %edi, %k1
+; X64-NEXT:    vpbroadcastd {{.*}}(%rip), %xmm0 {%k1} {z}
+; X64-NEXT:    retq
+entry:
+  %0 = bitcast i8 %__M to <8 x i1>
+  %extract.i = shufflevector <8 x i1> %0, <8 x i1> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %1 = select <4 x i1> %extract.i, <4 x i32> <i32 5, i32 5, i32 5, i32 5>, <4 x i32> zeroinitializer
+  %2 = bitcast <4 x i32> %1 to <2 x i64>
+  ret <2 x i64> %2
+}
+
+define <4 x i64> @test_mm256_mask_set1_epi32(<4 x i64> %__O, i8 zeroext %__M)  {
+; X32-LABEL: test_mm256_mask_set1_epi32:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; X32-NEXT:    kmovw %eax, %k1
+; X32-NEXT:    vpbroadcastd {{\.LCPI.*}}, %ymm0 {%k1}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm256_mask_set1_epi32:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovw %edi, %k1
+; X64-NEXT:    vpbroadcastd {{.*}}(%rip), %ymm0 {%k1}
+; X64-NEXT:    retq
+entry:
+  %0 = bitcast <4 x i64> %__O to <8 x i32>
+  %1 = bitcast i8 %__M to <8 x i1>
+  %2 = select <8 x i1> %1, <8 x i32> <i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>, <8 x i32> %0
+  %3 = bitcast <8 x i32> %2 to <4 x i64>
+  ret <4 x i64> %3
+}
+
+define <4 x i64> @test_mm256_maskz_set1_epi32(i8 zeroext %__M)  {
+; X32-LABEL: test_mm256_maskz_set1_epi32:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %al
+; X32-NEXT:    kmovw %eax, %k1
+; X32-NEXT:    vpbroadcastd {{\.LCPI.*}}, %ymm0 {%k1} {z}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm256_maskz_set1_epi32:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovw %edi, %k1
+; X64-NEXT:    vpbroadcastd {{.*}}(%rip), %ymm0 {%k1} {z}
+; X64-NEXT:    retq
+entry:
+  %0 = bitcast i8 %__M to <8 x i1>
+  %1 = select <8 x i1> %0, <8 x i32> <i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5, i32 5>, <8 x i32> zeroinitializer
+  %2 = bitcast <8 x i32> %1 to <4 x i64>
+  ret <4 x i64> %2
+}
+
+define <2 x i64> @test_mm_mask_set1_epi64(<2 x i64> %__O, i8 zeroext %__M, i64 %__A)  {
+; X32-LABEL: test_mm_mask_set1_epi64:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %cl
+; X32-NEXT:    vmovd %eax, %xmm1
+; X32-NEXT:    vpbroadcastb %xmm1, %xmm1
+; X32-NEXT:    kmovw %ecx, %k1
+; X32-NEXT:    vmovdqa64 %xmm1, %xmm0 {%k1}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm_mask_set1_epi64:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    vmovd %esi, %xmm1
+; X64-NEXT:    vpbroadcastb %xmm1, %xmm1
+; X64-NEXT:    kmovw %edi, %k1
+; X64-NEXT:    vmovdqa64 %xmm1, %xmm0 {%k1}
+; X64-NEXT:    retq
+entry:
+  %conv.i = trunc i64 %__A to i8
+  %vecinit.i.i = insertelement <16 x i8> undef, i8 %conv.i, i32 0
+  %vecinit15.i.i = shufflevector <16 x i8> %vecinit.i.i, <16 x i8> undef, <16 x i32> zeroinitializer
+  %0 = bitcast <16 x i8> %vecinit15.i.i to <2 x i64>
+  %1 = bitcast i8 %__M to <8 x i1>
+  %extract.i = shufflevector <8 x i1> %1, <8 x i1> undef, <2 x i32> <i32 0, i32 1>
+  %2 = select <2 x i1> %extract.i, <2 x i64> %0, <2 x i64> %__O
+  ret <2 x i64> %2
+}
+
+define <2 x i64> @test_mm_maskz_set1_epi64(i8 zeroext %__M, i64 %__A)  {
+; X32-LABEL: test_mm_maskz_set1_epi64:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %cl
+; X32-NEXT:    vmovd %eax, %xmm0
+; X32-NEXT:    vpbroadcastb %xmm0, %xmm0
+; X32-NEXT:    kmovw %ecx, %k1
+; X32-NEXT:    vmovdqa64 %xmm0, %xmm0 {%k1} {z}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm_maskz_set1_epi64:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    vmovd %esi, %xmm0
+; X64-NEXT:    vpbroadcastb %xmm0, %xmm0
+; X64-NEXT:    kmovw %edi, %k1
+; X64-NEXT:    vmovdqa64 %xmm0, %xmm0 {%k1} {z}
+; X64-NEXT:    retq
+entry:
+  %conv.i = trunc i64 %__A to i8
+  %vecinit.i.i = insertelement <16 x i8> undef, i8 %conv.i, i32 0
+  %vecinit15.i.i = shufflevector <16 x i8> %vecinit.i.i, <16 x i8> undef, <16 x i32> zeroinitializer
+  %0 = bitcast <16 x i8> %vecinit15.i.i to <2 x i64>
+  %1 = bitcast i8 %__M to <8 x i1>
+  %extract.i = shufflevector <8 x i1> %1, <8 x i1> undef, <2 x i32> <i32 0, i32 1>
+  %2 = select <2 x i1> %extract.i, <2 x i64> %0, <2 x i64> zeroinitializer
+  ret <2 x i64> %2
+}
+
+
+define <4 x i64> @test_mm256_mask_set1_epi64(<4 x i64> %__O, i8 zeroext %__M, i64 %__A) {
+; X32-LABEL: test_mm256_mask_set1_epi64:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %dl
+; X32-NEXT:    vmovd %ecx, %xmm1
+; X32-NEXT:    vpinsrd $1, %eax, %xmm1, %xmm1
+; X32-NEXT:    vpinsrd $2, %ecx, %xmm1, %xmm1
+; X32-NEXT:    vpinsrd $3, %eax, %xmm1, %xmm1
+; X32-NEXT:    vinserti128 $1, %xmm1, %ymm1, %ymm1
+; X32-NEXT:    kmovw %edx, %k1
+; X32-NEXT:    vmovdqa64 %ymm1, %ymm0 {%k1}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm256_mask_set1_epi64:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovw %edi, %k1
+; X64-NEXT:    vpbroadcastq %rsi, %ymm0 {%k1}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <4 x i64> undef, i64 %__A, i32 0
+  %vecinit3.i.i = shufflevector <4 x i64> %vecinit.i.i, <4 x i64> undef, <4 x i32> zeroinitializer
+  %0 = bitcast i8 %__M to <8 x i1>
+  %extract.i = shufflevector <8 x i1> %0, <8 x i1> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %1 = select <4 x i1> %extract.i, <4 x i64> %vecinit3.i.i, <4 x i64> %__O
+  ret <4 x i64> %1
+}
+
+define <4 x i64> @test_mm256_maskz_set1_epi64(i8 zeroext %__M, i64 %__A)  {
+; X32-LABEL: test_mm256_maskz_set1_epi64:
+; X32:       # BB#0: # %entry
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %eax
+; X32-NEXT:    movl {{[0-9]+}}(%esp), %ecx
+; X32-NEXT:    movb {{[0-9]+}}(%esp), %dl
+; X32-NEXT:    vmovd %ecx, %xmm0
+; X32-NEXT:    vpinsrd $1, %eax, %xmm0, %xmm0
+; X32-NEXT:    vpinsrd $2, %ecx, %xmm0, %xmm0
+; X32-NEXT:    vpinsrd $3, %eax, %xmm0, %xmm0
+; X32-NEXT:    vinserti128 $1, %xmm0, %ymm0, %ymm0
+; X32-NEXT:    kmovw %edx, %k1
+; X32-NEXT:    vmovdqa64 %ymm0, %ymm0 {%k1} {z}
+; X32-NEXT:    retl
+;
+; X64-LABEL: test_mm256_maskz_set1_epi64:
+; X64:       # BB#0: # %entry
+; X64-NEXT:    kmovw %edi, %k1
+; X64-NEXT:    vpbroadcastq %rsi, %ymm0 {%k1} {z}
+; X64-NEXT:    retq
+entry:
+  %vecinit.i.i = insertelement <4 x i64> undef, i64 %__A, i32 0
+  %vecinit3.i.i = shufflevector <4 x i64> %vecinit.i.i, <4 x i64> undef, <4 x i32> zeroinitializer
+  %0 = bitcast i8 %__M to <8 x i1>
+  %extract.i = shufflevector <8 x i1> %0, <8 x i1> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %1 = select <4 x i1> %extract.i, <4 x i64> %vecinit3.i.i, <4 x i64> zeroinitializer
+  ret <4 x i64> %1
+}
+
 define <2 x i64> @test_mm_broadcastd_epi32(<2 x i64> %a0) {
 ; X32-LABEL: test_mm_broadcastd_epi32:
 ; X32:       # BB#0:

diff --git a/test/CodeGen/X86/avx512vl-intrinsics-upgrade.ll b/test/CodeGen/X86/avx512vl-intrinsics-upgrade.ll
index 948142d6fb8baea0d86416e30dcc16c78d7d7b71..992c4a10887bf8ced330f26995411325618013e1 100644 (file)
--- a/test/CodeGen/X86/avx512vl-intrinsics-upgrade.ll
+++ b/test/CodeGen/X86/avx512vl-intrinsics-upgrade.ll
@@ -1,6 +1,90 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=knl -mattr=+avx512vl --show-mc-encoding| FileCheck %s
 
+declare <4 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.128(i32, <4 x i32>, i8)
+
+define <4 x i32>@test_int_x86_avx512_mask_pbroadcast_d_gpr_128(i32 %x0, <4 x i32> %x1, i8 %mask) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_d_gpr_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vpbroadcastd %edi, %xmm1 ## encoding: [0x62,0xf2,0x7d,0x08,0x7c,0xcf]
+; CHECK-NEXT:    kmovw %esi, %k1 ## encoding: [0xc5,0xf8,0x92,0xce]
+; CHECK-NEXT:    vpbroadcastd %edi, %xmm0 {%k1} ## encoding: [0x62,0xf2,0x7d,0x09,0x7c,0xc7]
+; CHECK-NEXT:    vpaddd %xmm0, %xmm1, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0xfe,0xc0]
+; CHECK-NEXT:    vpbroadcastd %edi, %xmm1 {%k1} {z} ## encoding: [0x62,0xf2,0x7d,0x89,0x7c,0xcf]
+; CHECK-NEXT:    vpaddd %xmm0, %xmm1, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0xfe,0xc0]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <4 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.128(i32 %x0, <4 x i32> %x1, i8 -1)
+  %res1 = call <4 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.128(i32 %x0, <4 x i32> %x1, i8 %mask)
+  %res2 = call <4 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.128(i32 %x0, <4 x i32> zeroinitializer, i8 %mask)
+  %res3 = add <4 x i32> %res, %res1
+  %res4 = add <4 x i32> %res2, %res3
+  ret <4 x i32> %res4
+}
+
+
+declare <2 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.128(i64, <2 x i64>, i8)
+
+define <2 x i64>@test_int_x86_avx512_mask_pbroadcast_q_gpr_128(i64 %x0, <2 x i64> %x1, i8 %mask) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_q_gpr_128:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vpbroadcastq %rdi, %xmm1 ## encoding: [0x62,0xf2,0xfd,0x08,0x7c,0xcf]
+; CHECK-NEXT:    kmovw %esi, %k1 ## encoding: [0xc5,0xf8,0x92,0xce]
+; CHECK-NEXT:    vpbroadcastq %rdi, %xmm0 {%k1} ## encoding: [0x62,0xf2,0xfd,0x09,0x7c,0xc7]
+; CHECK-NEXT:    vpaddq %xmm0, %xmm1, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0xd4,0xc0]
+; CHECK-NEXT:    vpbroadcastq %rdi, %xmm1 {%k1} {z} ## encoding: [0x62,0xf2,0xfd,0x89,0x7c,0xcf]
+; CHECK-NEXT:    vpaddq %xmm0, %xmm1, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0xd4,0xc0]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+  %res = call <2 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.128(i64 %x0, <2 x i64> %x1,i8 -1)
+  %res1 = call <2 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.128(i64 %x0, <2 x i64> %x1,i8 %mask)
+  %res2 = call <2 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.128(i64 %x0, <2 x i64> zeroinitializer,i8 %mask)
+  %res3 = add <2 x i64> %res, %res1
+  %res4 = add <2 x i64> %res2, %res3
+  ret <2 x i64> %res4
+}
+
+
+ declare <8 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.256(i32, <8 x i32>, i8)
+
+  define <8 x i32>@test_int_x86_avx512_mask_pbroadcast_d_gpr_256(i32 %x0, <8 x i32> %x1, i8 %mask) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_d_gpr_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vpbroadcastd %edi, %ymm1 ## encoding: [0x62,0xf2,0x7d,0x28,0x7c,0xcf]
+; CHECK-NEXT:    kmovw %esi, %k1 ## encoding: [0xc5,0xf8,0x92,0xce]
+; CHECK-NEXT:    vpbroadcastd %edi, %ymm0 {%k1} ## encoding: [0x62,0xf2,0x7d,0x29,0x7c,0xc7]
+; CHECK-NEXT:    vpaddd %ymm0, %ymm1, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf5,0xfe,0xc0]
+; CHECK-NEXT:    vpbroadcastd %edi, %ymm1 {%k1} {z} ## encoding: [0x62,0xf2,0x7d,0xa9,0x7c,0xcf]
+; CHECK-NEXT:    vpaddd %ymm0, %ymm1, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf5,0xfe,0xc0]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+    %res = call <8 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.256(i32 %x0, <8 x i32> %x1, i8 -1)
+    %res1 = call <8 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.256(i32 %x0, <8 x i32> %x1, i8 %mask)
+    %res2 = call <8 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.256(i32 %x0, <8 x i32> zeroinitializer, i8 %mask)
+    %res3 = add <8 x i32> %res, %res1
+    %res4 = add <8 x i32> %res2, %res3
+    ret <8 x i32> %res4
+  }
+
+ declare <4 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.256(i64, <4 x i64>, i8)
+
+  define <4 x i64>@test_int_x86_avx512_mask_pbroadcast_q_gpr_256(i64 %x0, <4 x i64> %x1, i8 %mask) {
+; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_q_gpr_256:
+; CHECK:       ## BB#0:
+; CHECK-NEXT:    vpbroadcastq %rdi, %ymm1 ## encoding: [0x62,0xf2,0xfd,0x28,0x7c,0xcf]
+; CHECK-NEXT:    kmovw %esi, %k1 ## encoding: [0xc5,0xf8,0x92,0xce]
+; CHECK-NEXT:    vpbroadcastq %rdi, %ymm0 {%k1} ## encoding: [0x62,0xf2,0xfd,0x29,0x7c,0xc7]
+; CHECK-NEXT:    vpaddq %ymm0, %ymm1, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf5,0xd4,0xc0]
+; CHECK-NEXT:    vpbroadcastq %rdi, %ymm1 {%k1} {z} ## encoding: [0x62,0xf2,0xfd,0xa9,0x7c,0xcf]
+; CHECK-NEXT:    vpaddq %ymm0, %ymm1, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf5,0xd4,0xc0]
+; CHECK-NEXT:    retq ## encoding: [0xc3]
+    %res = call <4 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.256(i64 %x0, <4 x i64> %x1,i8 -1)
+    %res1 = call <4 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.256(i64 %x0, <4 x i64> %x1,i8 %mask)
+    %res2 = call <4 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.256(i64 %x0, <4 x i64> zeroinitializer,i8 %mask)
+    %res3 = add <4 x i64> %res, %res1
+    %res4 = add <4 x i64> %res2, %res3
+    ret <4 x i64> %res4
+  }
+
+
+
 declare <8 x i32> @llvm.x86.avx512.pbroadcastd.256(<4 x i32>, <8 x i32>, i8)
 
 define <8 x i32>@test_int_x86_avx512_pbroadcastd_256(<4 x i32> %x0, <8 x i32> %x1, i8 %mask, i32 * %y_ptr) {
@@ -3925,9 +4009,9 @@ define <8 x i32>@test_int_x86_avx512_mask_psrav8_si_const() {
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    vmovdqa {{.*}}(%rip), %ymm0 ## EVEX TO VEX Compression ymm0 = [2,9,4294967284,23,4294967270,37,4294967256,51]
 ; CHECK-NEXT:    ## encoding: [0xc5,0xfd,0x6f,0x05,A,A,A,A]
-; CHECK-NEXT:    ## fixup A - offset: 4, value: LCPI276_0-4, kind: reloc_riprel_4byte
+; CHECK-NEXT:    ## fixup A - offset: 4, value: LCPI280_0-4, kind: reloc_riprel_4byte
 ; CHECK-NEXT:    vpsravd {{.*}}(%rip), %ymm0, %ymm0 ## EVEX TO VEX Compression encoding: [0xc4,0xe2,0x7d,0x46,0x05,A,A,A,A]
-; CHECK-NEXT:    ## fixup A - offset: 5, value: LCPI276_1-4, kind: reloc_riprel_4byte
+; CHECK-NEXT:    ## fixup A - offset: 5, value: LCPI280_1-4, kind: reloc_riprel_4byte
 ; CHECK-NEXT:    retq ## encoding: [0xc3]
   %res = call <8 x i32> @llvm.x86.avx512.mask.psrav8.si(<8 x i32> <i32 2, i32 9, i32 -12, i32 23, i32 -26, i32 37, i32 -40, i32 51>, <8 x i32> <i32 1, i32 18, i32 35, i32 52, i32 69, i32 15, i32 32, i32 49>, <8 x i32> zeroinitializer, i8 -1)
   ret <8 x i32> %res
@@ -4508,9 +4592,9 @@ define <2 x i64>@test_int_x86_avx512_mask_psrav_q_128_const(i8 %x3) {
 ; CHECK:       ## BB#0:
 ; CHECK-NEXT:    vmovdqa {{.*}}(%rip), %xmm0 ## EVEX TO VEX Compression xmm0 = [2,18446744073709551607]
 ; CHECK-NEXT:    ## encoding: [0xc5,0xf9,0x6f,0x05,A,A,A,A]
-; CHECK-NEXT:    ## fixup A - offset: 4, value: LCPI304_0-4, kind: reloc_riprel_4byte
+; CHECK-NEXT:    ## fixup A - offset: 4, value: LCPI308_0-4, kind: reloc_riprel_4byte
 ; CHECK-NEXT:    vpsravq {{.*}}(%rip), %xmm0, %xmm0 ## encoding: [0x62,0xf2,0xfd,0x08,0x46,0x05,A,A,A,A]
-; CHECK-NEXT:    ## fixup A - offset: 6, value: LCPI304_1-4, kind: reloc_riprel_4byte
+; CHECK-NEXT:    ## fixup A - offset: 6, value: LCPI308_1-4, kind: reloc_riprel_4byte
 ; CHECK-NEXT:    retq ## encoding: [0xc3]
   %res = call <2 x i64> @llvm.x86.avx512.mask.psrav.q.128(<2 x i64> <i64 2, i64 -9>, <2 x i64> <i64 1, i64 90>, <2 x i64> zeroinitializer, i8 -1)
   ret <2 x i64> %res

diff --git a/test/CodeGen/X86/avx512vl-intrinsics.ll b/test/CodeGen/X86/avx512vl-intrinsics.ll
index bc10be199d5153772a275fec64e088c85ca56dc5..9098ca308971d03cb2f952849c4d92c8f1392bac 100644 (file)
--- a/test/CodeGen/X86/avx512vl-intrinsics.ll
+++ b/test/CodeGen/X86/avx512vl-intrinsics.ll
@@ -4093,85 +4093,6 @@ define i8@test_int_x86_avx512_ptestnm_q_256(<4 x i64> %x0, <4 x i64> %x1, i8 %x2
   ret i8 %res2
 }
 
-declare <8 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.256(i32, <8 x i32>, i8)
-
-define <8 x i32>@test_int_x86_avx512_mask_pbroadcast_d_gpr_256(i32 %x0, <8 x i32> %x1, i8 %mask) {
-; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_d_gpr_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %esi, %k1 ## encoding: [0xc5,0xf8,0x92,0xce]
-; CHECK-NEXT:    vpbroadcastd %edi, %ymm1 {%k1} {z} ## encoding: [0x62,0xf2,0x7d,0xa9,0x7c,0xcf]
-; CHECK-NEXT:    vpbroadcastd %edi, %ymm0 {%k1} ## encoding: [0x62,0xf2,0x7d,0x29,0x7c,0xc7]
-; CHECK-NEXT:    vpbroadcastd %edi, %ymm2 ## encoding: [0x62,0xf2,0x7d,0x28,0x7c,0xd7]
-; CHECK-NEXT:    vpaddd %ymm0, %ymm2, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xed,0xfe,0xc0]
-; CHECK-NEXT:    vpaddd %ymm0, %ymm1, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf5,0xfe,0xc0]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <8 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.256(i32 %x0, <8 x i32> %x1, i8 -1)
-  %res1 = call <8 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.256(i32 %x0, <8 x i32> %x1, i8 %mask)
-  %res2 = call <8 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.256(i32 %x0, <8 x i32> zeroinitializer, i8 %mask)
-  %res3 = add <8 x i32> %res, %res1
-  %res4 = add <8 x i32> %res2, %res3
-  ret <8 x i32> %res4
-}
-
-declare <4 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.128(i32, <4 x i32>, i8)
-
-define <4 x i32>@test_int_x86_avx512_mask_pbroadcast_d_gpr_128(i32 %x0, <4 x i32> %x1, i8 %mask) {
-; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_d_gpr_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %esi, %k1 ## encoding: [0xc5,0xf8,0x92,0xce]
-; CHECK-NEXT:    vpbroadcastd %edi, %xmm1 {%k1} {z} ## encoding: [0x62,0xf2,0x7d,0x89,0x7c,0xcf]
-; CHECK-NEXT:    vpbroadcastd %edi, %xmm0 {%k1} ## encoding: [0x62,0xf2,0x7d,0x09,0x7c,0xc7]
-; CHECK-NEXT:    vpbroadcastd %edi, %xmm2 ## encoding: [0x62,0xf2,0x7d,0x08,0x7c,0xd7]
-; CHECK-NEXT:    vpaddd %xmm0, %xmm2, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xe9,0xfe,0xc0]
-; CHECK-NEXT:    vpaddd %xmm0, %xmm1, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0xfe,0xc0]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.128(i32 %x0, <4 x i32> %x1, i8 -1)
-  %res1 = call <4 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.128(i32 %x0, <4 x i32> %x1, i8 %mask)
-  %res2 = call <4 x i32> @llvm.x86.avx512.mask.pbroadcast.d.gpr.128(i32 %x0, <4 x i32> zeroinitializer, i8 %mask)
-  %res3 = add <4 x i32> %res, %res1
-  %res4 = add <4 x i32> %res2, %res3
-  ret <4 x i32> %res4
-}
-
-declare <4 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.256(i64, <4 x i64>, i8)
-
-define <4 x i64>@test_int_x86_avx512_mask_pbroadcast_q_gpr_256(i64 %x0, <4 x i64> %x1, i8 %mask) {
-; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_q_gpr_256:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %esi, %k1 ## encoding: [0xc5,0xf8,0x92,0xce]
-; CHECK-NEXT:    vpbroadcastq %rdi, %ymm1 {%k1} {z} ## encoding: [0x62,0xf2,0xfd,0xa9,0x7c,0xcf]
-; CHECK-NEXT:    vpbroadcastq %rdi, %ymm0 {%k1} ## encoding: [0x62,0xf2,0xfd,0x29,0x7c,0xc7]
-; CHECK-NEXT:    vpbroadcastq %rdi, %ymm2 ## encoding: [0x62,0xf2,0xfd,0x28,0x7c,0xd7]
-; CHECK-NEXT:    vpaddq %ymm0, %ymm2, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xed,0xd4,0xc0]
-; CHECK-NEXT:    vpaddq %ymm0, %ymm1, %ymm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf5,0xd4,0xc0]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <4 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.256(i64 %x0, <4 x i64> %x1,i8 -1)
-  %res1 = call <4 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.256(i64 %x0, <4 x i64> %x1,i8 %mask)
-  %res2 = call <4 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.256(i64 %x0, <4 x i64> zeroinitializer,i8 %mask)
-  %res3 = add <4 x i64> %res, %res1
-  %res4 = add <4 x i64> %res2, %res3
-  ret <4 x i64> %res4
-}
-
-declare <2 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.128(i64, <2 x i64>, i8)
-
-define <2 x i64>@test_int_x86_avx512_mask_pbroadcast_q_gpr_128(i64 %x0, <2 x i64> %x1, i8 %mask) {
-; CHECK-LABEL: test_int_x86_avx512_mask_pbroadcast_q_gpr_128:
-; CHECK:       ## BB#0:
-; CHECK-NEXT:    kmovw %esi, %k1 ## encoding: [0xc5,0xf8,0x92,0xce]
-; CHECK-NEXT:    vpbroadcastq %rdi, %xmm1 {%k1} {z} ## encoding: [0x62,0xf2,0xfd,0x89,0x7c,0xcf]
-; CHECK-NEXT:    vpbroadcastq %rdi, %xmm0 {%k1} ## encoding: [0x62,0xf2,0xfd,0x09,0x7c,0xc7]
-; CHECK-NEXT:    vpbroadcastq %rdi, %xmm2 ## encoding: [0x62,0xf2,0xfd,0x08,0x7c,0xd7]
-; CHECK-NEXT:    vpaddq %xmm0, %xmm2, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xe9,0xd4,0xc0]
-; CHECK-NEXT:    vpaddq %xmm0, %xmm1, %xmm0 ## EVEX TO VEX Compression encoding: [0xc5,0xf1,0xd4,0xc0]
-; CHECK-NEXT:    retq ## encoding: [0xc3]
-  %res = call <2 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.128(i64 %x0, <2 x i64> %x1,i8 -1)
-  %res1 = call <2 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.128(i64 %x0, <2 x i64> %x1,i8 %mask)
-  %res2 = call <2 x i64> @llvm.x86.avx512.mask.pbroadcast.q.gpr.128(i64 %x0, <2 x i64> zeroinitializer,i8 %mask)
-  %res3 = add <2 x i64> %res, %res1
-  %res4 = add <2 x i64> %res2, %res3
-  ret <2 x i64> %res4
-}
 
 
 define <2 x i64> @test_x86_avx512_psra_q_128(<2 x i64> %a0, <2 x i64> %a1) {