[X86] Lowering PACK*S (pack with saturation) intrinsics to native IR (clang side)

[mike.dvoretsky]

Summary

This patch lowers the X86 vector packing with saturation intrinsics to native LLVM IR. Comes with an LLVM patch (D45721).

[craig.topper]

lib/CodeGen/CGBuiltin.cpp

8420 ↗

(On Diff #142768)

Why can't these just be APInts instead of uint64_t? Is this so that APInt widths don't have to match RTy below? I'd rather you just created the narrow APInt and then called sext/zext on it to get it to the right width.

8420 ↗

(On Diff #142768)

Pre-select the 8 or 16 based on IsDW. Then you don't need to check IsDW 4 times. You just need to pass the right width.

8432 ↗

(On Diff #142768)

Clearing isn't necessary if you just created it.

8433 ↗

(On Diff #142768)

This loop could probably use some comments. The multiple variables make the logic hard to follow

8443 ↗

(On Diff #142768)

Why arent' these unsigned compares for Unsigned?

8446 ↗

(On Diff #142768)

If you have the 8 or 16 selected above, you can use getIntNTy here I think.

[mike.dvoretsky]

lib/CodeGen/CGBuiltin.cpp

8443 ↗

(On Diff #142768)

The compares are signed on purpose. PACKUS assumes that the input elements are signed, then uses unsigned saturation. So, for instance, an 0xffff value must be evaluated as -1 and saturated to 0, rather than to 0xff as it would be with unsigned comparisons.

[mike.dvoretsky]

Updated per comments.

[craig.topper]

LGTM

[mike.dvoretsky]

Changed the shuffle mask emission code to match D45721.

[mike.dvoretsky]

Closing this due to failure of D45721.

clang/lib/CodeGen/CGBuiltin.cpp

 
 static Value *EmitX86SExtMask(CodeGenFunction &CGF, Value *Op, 
                               llvm::Type *DstTy) {
   unsigned NumberOfElements = DstTy->getVectorNumElements();
   Value *Mask = getMaskVecValue(CGF, Op, NumberOfElements);
   return CGF.Builder.CreateSExt(Mask, DstTy, "vpmovm2");
 }
 
+static Value *EmitX86Pack(CodeGenFunction &CGF, ArrayRef<Value *> Ops,
+                          bool IsUnsigned) {
+  Value *A = Ops[0];
+  Value *B = Ops[1];
+
+  llvm::Type *Ty = A->getType();
+  int EltSize = Ty->getScalarSizeInBits();
+  llvm::APInt MinVal, MaxVal;
+  if (IsUnsigned) {
+    MinVal = llvm::APInt::getMinValue(EltSize / 2).zext(EltSize);
+    MaxVal = llvm::APInt::getMaxValue(EltSize / 2).zext(EltSize);
+  } else {
+    MinVal = llvm::APInt::getSignedMinValue(EltSize / 2).sext(EltSize);
+    MaxVal = llvm::APInt::getSignedMaxValue(EltSize / 2).sext(EltSize);
+  }
+
+  SmallVector<uint32_t, 16> ShuffleMask;
+  unsigned NumElts = Ty->getVectorNumElements();
+  unsigned NumLanes = NumElts * Ty->getScalarSizeInBits() / 128;
+  unsigned NumEltsPerLane = 128 / EltSize;
+
+  for (unsigned Lane = 0; Lane != NumLanes; ++Lane) {
+    for (unsigned Elt = 0; Elt != NumEltsPerLane; ++Elt)
+      ShuffleMask.push_back(Elt + (Lane * NumEltsPerLane));
+    for (unsigned Elt = 0; Elt != NumEltsPerLane; ++Elt)
+      ShuffleMask.push_back(Elt + (Lane * NumEltsPerLane) + NumElts);
+  }
+
+  Value *Res = CGF.Builder.CreateShuffleVector(A, B, ShuffleMask);
+  llvm::Type *RTy = Res->getType();
+  Value *MinVec = llvm::ConstantInt::get(RTy, MinVal);
+  Value *MaxVec = llvm::ConstantInt::get(RTy, MaxVal);
+  Res = EmitX86MinMax(CGF, ICmpInst::ICMP_SLT, {Res, MaxVec});
+  Res = EmitX86MinMax(CGF, ICmpInst::ICMP_SGT, {Res, MinVec});
+  llvm::Type *VTy = llvm::VectorType::get(CGF.Builder.getIntNTy(EltSize / 2),
+                                          NumElts * 2);
+  return CGF.Builder.CreateTrunc(Res, VTy);
+}
+
 Value *CodeGenFunction::EmitX86CpuIs(const CallExpr *E) {
   const Expr *CPUExpr = E->getArg(0)->IgnoreParenCasts();
   StringRef CPUStr = cast<clang::StringLiteral>(CPUExpr)->getString();
   return EmitX86CpuIs(CPUStr);
 }
 
 Value *CodeGenFunction::EmitX86CpuIs(StringRef CPUStr) {
 
case X86::BI__readgsqword: {
     llvm::Type *IntTy = ConvertType(E->getType());
     Value *Ptr = Builder.CreateIntToPtr(EmitScalarExpr(E->getArg(0)),
                                         llvm::PointerType::get(IntTy, 256));
     LoadInst *Load = Builder.CreateAlignedLoad(
         IntTy, Ptr, getContext().getTypeAlignInChars(E->getType()));
     Load->setVolatile(true);
     return Load;
   }
+
+  // Packing intrinsics.
+  case X86::BI__builtin_ia32_packsswb128:
+  case X86::BI__builtin_ia32_packssdw128:
+  case X86::BI__builtin_ia32_packsswb256:
+  case X86::BI__builtin_ia32_packssdw256:
+  case X86::BI__builtin_ia32_packsswb512:
+  case X86::BI__builtin_ia32_packssdw512:
+    return EmitX86Pack(*this, Ops, false /*IsUnsigned*/);
+  case X86::BI__builtin_ia32_packuswb128:
+  case X86::BI__builtin_ia32_packusdw128:
+  case X86::BI__builtin_ia32_packuswb256:
+  case X86::BI__builtin_ia32_packusdw256:
+  case X86::BI__builtin_ia32_packuswb512:
+  case X86::BI__builtin_ia32_packusdw512:
+    return EmitX86Pack(*this, Ops, true /*IsUnsigned*/);
   }
 }
 
 
 Value *CodeGenFunction::EmitPPCBuiltinExpr(unsigned BuiltinID,
                                            const CallExpr *E) {
   SmallVector<Value*, 4> Ops;
 

clang/test/CodeGen/avx2-builtins.c

 __m256i test_mm256_or_si256(__m256i a, __m256i b) {
   // CHECK-LABEL: test_mm256_or_si256
   // CHECK: or <4 x i64>
   return _mm256_or_si256(a, b);
 }
 
 __m256i test_mm256_packs_epi16(__m256i a, __m256i b) {
   // CHECK-LABEL: test_mm256_packs_epi16
-  // CHECK: call <32 x i8> @llvm.x86.avx2.packsswb(<16 x i16> %{{.*}}, <16 x i16> %{{.*}})
+  // CHECK: %{{.*}} = shufflevector <16 x i16> %{{.*}}, <16 x i16> %{{.*}}, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31>
+  // CHECK: %{{.*}} = icmp slt <32 x i16> %{{.*}}, <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = icmp sgt <32 x i16> %{{.*}}, <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = trunc <32 x i16> %{{.*}} to <32 x i8>
   return _mm256_packs_epi16(a, b);
 }
 
 __m256i test_mm256_packs_epi32(__m256i a, __m256i b) {
   // CHECK-LABEL: test_mm256_packs_epi32
-  // CHECK: call <16 x i16> @llvm.x86.avx2.packssdw(<8 x i32> %{{.*}}, <8 x i32> %{{.*}})
+  // CHECK: %{{.*}} = shufflevector <8 x i32> %{{.*}}, <8 x i32> %{{.*}}, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11, i32 4, i32 5, i32 6, i32 7, i32 12, i32 13, i32 14, i32 15>
+  // CHECK: %{{.*}} = icmp slt <16 x i32> %{{.*}}, <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = icmp sgt <16 x i32> %{{.*}}, <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = trunc <16 x i32> %{{.*}} to <16 x i16>
   return _mm256_packs_epi32(a, b);
 }
 
 __m256i test_mm256_packs_epu16(__m256i a, __m256i b) {
   // CHECK-LABEL: test_mm256_packs_epu16
-  // CHECK:  call <32 x i8> @llvm.x86.avx2.packuswb(<16 x i16> %{{.*}}, <16 x i16> %{{.*}})
+  // CHECK: %{{.*}} = shufflevector <16 x i16> %{{.*}}, <16 x i16> %{{.*}}, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31>
+  // CHECK: %{{.*}} = icmp slt <32 x i16> %{{.*}}, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = icmp sgt <32 x i16> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> zeroinitializer
+  // CHECK: %{{.*}} = trunc <32 x i16> %{{.*}} to <32 x i8>
   return _mm256_packus_epi16(a, b);
 }
 
 __m256i test_mm256_packs_epu32(__m256i a, __m256i b) {
   // CHECK-LABEL: test_mm256_packs_epu32
-  // CHECK: call <16 x i16> @llvm.x86.avx2.packusdw(<8 x i32> %{{.*}}, <8 x i32> %{{.*}})
+  // CHECK: %{{.*}} = shufflevector <8 x i32> %{{.*}}, <8 x i32> %{{.*}}, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11, i32 4, i32 5, i32 6, i32 7, i32 12, i32 13, i32 14, i32 15>
+  // CHECK: %{{.*}} = icmp slt <16 x i32> %{{.*}}, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = icmp sgt <16 x i32> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> zeroinitializer
+  // CHECK: %{{.*}} = trunc <16 x i32> %{{.*}} to <16 x i16>
   return _mm256_packus_epi32(a, b);
 }
 
 __m256i test_mm256_permute2x128_si256(__m256i a, __m256i b) {
   // CHECK-LABEL: test_mm256_permute2x128_si256
   // CHECK: shufflevector <4 x i64> zeroinitializer, <4 x i64> %{{.*}}, <4 x i32> <i32 0, i32 1, i32 6, i32 7>
   return _mm256_permute2x128_si256(a, b, 0x38);
 }

clang/test/CodeGen/avx512bw-builtins.c

__m512i test_mm512_maskz_abs_epi16(__mmask32 __U, __m512i __A) {
   // CHECK: [[SUB:%.*]] = sub <32 x i16> zeroinitializer, [[A:%.*]]
   // CHECK: [[CMP:%.*]] = icmp sgt <32 x i16> [[A]], zeroinitializer
   // CHECK: [[SEL:%.*]] = select <32 x i1> [[CMP]], <32 x i16> [[A]], <32 x i16> [[SUB]]
   // CHECK: select <32 x i1> %{{.*}}, <32 x i16> [[SEL]], <32 x i16> %{{.*}}
   return _mm512_maskz_abs_epi16(__U,__A); 
 }
 __m512i test_mm512_packs_epi32(__m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_packs_epi32
-  // CHECK: @llvm.x86.avx512.packssdw.512
+  // CHECK: %{{.*}} = shufflevector <16 x i32> %{{.*}}, <16 x i32> %{{.*}}, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 16, i32 17, i32 18, i32 19, i32 4, i32 5, i32 6, i32 7, i32 20, i32 21, i32 22, i32 23, i32 8, i32 9, i32 10, i32 11, i32 24, i32 25, i32 26, i32 27, i32 12, i32 13, i32 14, i32 15, i32 28, i32 29, i32 30, i32 31>
+  // CHECK: %{{.*}} = icmp slt <32 x i32> %{{.*}}, <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i32> %{{.*}}, <32 x i32> <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = icmp sgt <32 x i32> %{{.*}}, <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i32> %{{.*}}, <32 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = trunc <32 x i32> %{{.*}} to <32 x i16>
   return _mm512_packs_epi32(__A,__B); 
 }
 __m512i test_mm512_maskz_packs_epi32(__mmask32 __M, __m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_maskz_packs_epi32
-  // CHECK: @llvm.x86.avx512.packssdw.512
+  // CHECK: %{{.*}} = shufflevector <16 x i32> %{{.*}}, <16 x i32> %{{.*}}, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 16, i32 17, i32 18, i32 19, i32 4, i32 5, i32 6, i32 7, i32 20, i32 21, i32 22, i32 23, i32 8, i32 9, i32 10, i32 11, i32 24, i32 25, i32 26, i32 27, i32 12, i32 13, i32 14, i32 15, i32 28, i32 29, i32 30, i32 31>
+  // CHECK: %{{.*}} = icmp slt <32 x i32> %{{.*}}, <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i32> %{{.*}}, <32 x i32> <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = icmp sgt <32 x i32> %{{.*}}, <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i32> %{{.*}}, <32 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = trunc <32 x i32> %{{.*}} to <32 x i16>
   // CHECK: select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> %{{.*}}
   return _mm512_maskz_packs_epi32(__M,__A,__B); 
 }
 __m512i test_mm512_mask_packs_epi32(__m512i __W, __mmask32 __M, __m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_mask_packs_epi32
-  // CHECK: @llvm.x86.avx512.packssdw.512
+  // CHECK: %{{.*}} = shufflevector <16 x i32> %{{.*}}, <16 x i32> %{{.*}}, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 16, i32 17, i32 18, i32 19, i32 4, i32 5, i32 6, i32 7, i32 20, i32 21, i32 22, i32 23, i32 8, i32 9, i32 10, i32 11, i32 24, i32 25, i32 26, i32 27, i32 12, i32 13, i32 14, i32 15, i32 28, i32 29, i32 30, i32 31>
+  // CHECK: %{{.*}} = icmp slt <32 x i32> %{{.*}}, <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i32> %{{.*}}, <32 x i32> <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = icmp sgt <32 x i32> %{{.*}}, <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i32> %{{.*}}, <32 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = trunc <32 x i32> %{{.*}} to <32 x i16>
   // CHECK: select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> %{{.*}}
   return _mm512_mask_packs_epi32(__W,__M,__A,__B); 
 }
 __m512i test_mm512_packs_epi16(__m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_packs_epi16
-  // CHECK: @llvm.x86.avx512.packsswb.512
+  // CHECK: %{{.*}} = shufflevector <32 x i16> %{{.*}}, <32 x i16> %{{.*}}, <64 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 32, i32 33, i32 34, i32 35, i32 36, i32 37, i32 38, i32 39, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 40, i32 41, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 48, i32 49, i32 50, i32 51, i32 52, i32 53, i32 54, i32 55, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 56, i32 57, i32 58, i32 59, i32 60, i32 61, i32 62, i32 63>
+  // CHECK: %{{.*}} = icmp slt <64 x i16> %{{.*}}, <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = select <64 x i1> %{{.*}}, <64 x i16> %{{.*}}, <64 x i16> <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = icmp sgt <64 x i16> %{{.*}}, <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = select <64 x i1> %{{.*}}, <64 x i16> %{{.*}}, <64 x i16> <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = trunc <64 x i16> %{{.*}} to <64 x i8>
   return _mm512_packs_epi16(__A,__B); 
 }
 __m512i test_mm512_mask_packs_epi16(__m512i __W, __mmask64 __M, __m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_mask_packs_epi16
-  // CHECK: @llvm.x86.avx512.packsswb.512
+  // CHECK: %{{.*}} = shufflevector <32 x i16> %{{.*}}, <32 x i16> %{{.*}}, <64 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 32, i32 33, i32 34, i32 35, i32 36, i32 37, i32 38, i32 39, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 40, i32 41, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 48, i32 49, i32 50, i32 51, i32 52, i32 53, i32 54, i32 55, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 56, i32 57, i32 58, i32 59, i32 60, i32 61, i32 62, i32 63>
+  // CHECK: %{{.*}} = icmp slt <64 x i16> %{{.*}}, <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = select <64 x i1> %{{.*}}, <64 x i16> %{{.*}}, <64 x i16> <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = icmp sgt <64 x i16> %{{.*}}, <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = select <64 x i1> %{{.*}}, <64 x i16> %{{.*}}, <64 x i16> <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = trunc <64 x i16> %{{.*}} to <64 x i8>
   // CHECK: select <64 x i1> %{{.*}}, <64 x i8> %{{.*}}, <64 x i8> %{{.*}}
   return _mm512_mask_packs_epi16(__W,__M,__A,__B); 
 }
 __m512i test_mm512_maskz_packs_epi16(__mmask64 __M, __m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_maskz_packs_epi16
-  // CHECK: @llvm.x86.avx512.packsswb.512
+  // CHECK: %{{.*}} = shufflevector <32 x i16> %{{.*}}, <32 x i16> %{{.*}}, <64 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 32, i32 33, i32 34, i32 35, i32 36, i32 37, i32 38, i32 39, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 40, i32 41, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 48, i32 49, i32 50, i32 51, i32 52, i32 53, i32 54, i32 55, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 56, i32 57, i32 58, i32 59, i32 60, i32 61, i32 62, i32 63>
+  // CHECK: %{{.*}} = icmp slt <64 x i16> %{{.*}}, <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = select <64 x i1> %{{.*}}, <64 x i16> %{{.*}}, <64 x i16> <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = icmp sgt <64 x i16> %{{.*}}, <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = select <64 x i1> %{{.*}}, <64 x i16> %{{.*}}, <64 x i16> <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = trunc <64 x i16> %{{.*}} to <64 x i8>
   // CHECK: select <64 x i1> %{{.*}}, <64 x i8> %{{.*}}, <64 x i8> %{{.*}}
   return _mm512_maskz_packs_epi16(__M,__A,__B); 
 }
 __m512i test_mm512_packus_epi32(__m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_packus_epi32
-  // CHECK: @llvm.x86.avx512.packusdw.512
+  // CHECK: %{{.*}} = shufflevector <16 x i32> %{{.*}}, <16 x i32> %{{.*}}, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 16, i32 17, i32 18, i32 19, i32 4, i32 5, i32 6, i32 7, i32 20, i32 21, i32 22, i32 23, i32 8, i32 9, i32 10, i32 11, i32 24, i32 25, i32 26, i32 27, i32 12, i32 13, i32 14, i32 15, i32 28, i32 29, i32 30, i32 31>
+  // CHECK: %{{.*}} = icmp slt <32 x i32> %{{.*}}, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i32> %{{.*}}, <32 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = icmp sgt <32 x i32> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i32> %{{.*}}, <32 x i32> zeroinitializer
+  // CHECK: %{{.*}} = trunc <32 x i32> %{{.*}} to <32 x i16>
   return _mm512_packus_epi32(__A,__B); 
 }
 __m512i test_mm512_maskz_packus_epi32(__mmask32 __M, __m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_maskz_packus_epi32
-  // CHECK: @llvm.x86.avx512.packusdw.512
+  // CHECK: %{{.*}} = shufflevector <16 x i32> %{{.*}}, <16 x i32> %{{.*}}, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 16, i32 17, i32 18, i32 19, i32 4, i32 5, i32 6, i32 7, i32 20, i32 21, i32 22, i32 23, i32 8, i32 9, i32 10, i32 11, i32 24, i32 25, i32 26, i32 27, i32 12, i32 13, i32 14, i32 15, i32 28, i32 29, i32 30, i32 31>
+  // CHECK: %{{.*}} = icmp slt <32 x i32> %{{.*}}, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i32> %{{.*}}, <32 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = icmp sgt <32 x i32> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i32> %{{.*}}, <32 x i32> zeroinitializer
+  // CHECK: %{{.*}} = trunc <32 x i32> %{{.*}} to <32 x i16>
   // CHECK: select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> %{{.*}}
   return _mm512_maskz_packus_epi32(__M,__A,__B); 
 }
 __m512i test_mm512_mask_packus_epi32(__m512i __W, __mmask32 __M, __m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_mask_packus_epi32
-  // CHECK: @llvm.x86.avx512.packusdw.512
+  // CHECK: %{{.*}} = shufflevector <16 x i32> %{{.*}}, <16 x i32> %{{.*}}, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 16, i32 17, i32 18, i32 19, i32 4, i32 5, i32 6, i32 7, i32 20, i32 21, i32 22, i32 23, i32 8, i32 9, i32 10, i32 11, i32 24, i32 25, i32 26, i32 27, i32 12, i32 13, i32 14, i32 15, i32 28, i32 29, i32 30, i32 31>
+  // CHECK: %{{.*}} = icmp slt <32 x i32> %{{.*}}, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i32> %{{.*}}, <32 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = icmp sgt <32 x i32> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i32> %{{.*}}, <32 x i32> zeroinitializer
+  // CHECK: %{{.*}} = trunc <32 x i32> %{{.*}} to <32 x i16>
   // CHECK: select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> %{{.*}}
   return _mm512_mask_packus_epi32(__W,__M,__A,__B); 
 }
 __m512i test_mm512_packus_epi16(__m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_packus_epi16
-  // CHECK: @llvm.x86.avx512.packuswb.512
+  // CHECK: %{{.*}} = shufflevector <32 x i16> %{{.*}}, <32 x i16> %{{.*}}, <64 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 32, i32 33, i32 34, i32 35, i32 36, i32 37, i32 38, i32 39, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 40, i32 41, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 48, i32 49, i32 50, i32 51, i32 52, i32 53, i32 54, i32 55, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 56, i32 57, i32 58, i32 59, i32 60, i32 61, i32 62, i32 63>
+  // CHECK: %{{.*}} = icmp slt <64 x i16> %{{.*}}, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = select <64 x i1> %{{.*}}, <64 x i16> %{{.*}}, <64 x i16> <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = icmp sgt <64 x i16> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <64 x i1> %{{.*}}, <64 x i16> %{{.*}}, <64 x i16> zeroinitializer
+  // CHECK: %{{.*}} = trunc <64 x i16> %{{.*}} to <64 x i8>
   return _mm512_packus_epi16(__A,__B); 
 }
 __m512i test_mm512_mask_packus_epi16(__m512i __W, __mmask64 __M, __m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_mask_packus_epi16
-  // CHECK: @llvm.x86.avx512.packuswb.512
+  // CHECK: %{{.*}} = shufflevector <32 x i16> %{{.*}}, <32 x i16> %{{.*}}, <64 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 32, i32 33, i32 34, i32 35, i32 36, i32 37, i32 38, i32 39, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 40, i32 41, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 48, i32 49, i32 50, i32 51, i32 52, i32 53, i32 54, i32 55, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 56, i32 57, i32 58, i32 59, i32 60, i32 61, i32 62, i32 63>
+  // CHECK: %{{.*}} = icmp slt <64 x i16> %{{.*}}, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = select <64 x i1> %{{.*}}, <64 x i16> %{{.*}}, <64 x i16> <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = icmp sgt <64 x i16> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <64 x i1> %{{.*}}, <64 x i16> %{{.*}}, <64 x i16> zeroinitializer
+  // CHECK: %{{.*}} = trunc <64 x i16> %{{.*}} to <64 x i8>
   // CHECK: select <64 x i1> %{{.*}}, <64 x i8> %{{.*}}, <64 x i8> %{{.*}}
   return _mm512_mask_packus_epi16(__W,__M,__A,__B); 
 }
 __m512i test_mm512_maskz_packus_epi16(__mmask64 __M, __m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_maskz_packus_epi16
-  // CHECK: @llvm.x86.avx512.packuswb.512
+  // CHECK: %{{.*}} = shufflevector <32 x i16> %{{.*}}, <32 x i16> %{{.*}}, <64 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 32, i32 33, i32 34, i32 35, i32 36, i32 37, i32 38, i32 39, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 40, i32 41, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 48, i32 49, i32 50, i32 51, i32 52, i32 53, i32 54, i32 55, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 56, i32 57, i32 58, i32 59, i32 60, i32 61, i32 62, i32 63>
+  // CHECK: %{{.*}} = icmp slt <64 x i16> %{{.*}}, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = select <64 x i1> %{{.*}}, <64 x i16> %{{.*}}, <64 x i16> <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = icmp sgt <64 x i16> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <64 x i1> %{{.*}}, <64 x i16> %{{.*}}, <64 x i16> zeroinitializer
+  // CHECK: %{{.*}} = trunc <64 x i16> %{{.*}} to <64 x i8>
   // CHECK: select <64 x i1> %{{.*}}, <64 x i8> %{{.*}}, <64 x i8> %{{.*}}
   return _mm512_maskz_packus_epi16(__M,__A,__B); 
 }
 __m512i test_mm512_adds_epi8(__m512i __A, __m512i __B) {
   // CHECK-LABEL: @test_mm512_adds_epi8
   // CHECK: @llvm.x86.avx512.mask.padds.b.512
   return _mm512_adds_epi8(__A,__B); 
 }

clang/test/CodeGen/avx512vlbw-builtins.c

__m256i test_mm256_maskz_abs_epi16(__mmask16 __U, __m256i __A) {
   // CHECK: [[CMP:%.*]] = icmp sgt <16 x i16> [[A]], zeroinitializer
   // CHECK: [[SEL:%.*]] = select <16 x i1> [[CMP]], <16 x i16> [[A]], <16 x i16> [[SUB]]
   // CHECK: select <16 x i1> %{{.*}}, <16 x i16> [[SEL]], <16 x i16> %{{.*}}
   return _mm256_maskz_abs_epi16(__U,__A); 
 }
 
 __m128i test_mm_maskz_packs_epi32(__mmask8 __M, __m128i __A, __m128i __B) {
   // CHECK-LABEL: @test_mm_maskz_packs_epi32
-  // CHECK: @llvm.x86.sse2.packssdw
+  // CHECK: %{{.*}} = shufflevector <4 x i32> %{{.*}}, <4 x i32> %{{.*}}, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+  // CHECK: %{{.*}} = icmp slt <8 x i32> %{{.*}}, <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = select <8 x i1> %{{.*}}, <8 x i32> %{{.*}}, <8 x i32> <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = icmp sgt <8 x i32> %{{.*}}, <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = select <8 x i1> %{{.*}}, <8 x i32> %{{.*}}, <8 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = trunc <8 x i32> %{{.*}} to <8 x i16>
   // CHECK: select <8 x i1> %{{.*}}, <8 x i16> %{{.*}}, <8 x i16> %{{.*}}
   return _mm_maskz_packs_epi32(__M,__A,__B); 
 }
 __m128i test_mm_mask_packs_epi32(__m128i __W, __mmask16 __M, __m128i __A,          __m128i __B) {
   // CHECK-LABEL: @test_mm_mask_packs_epi32
-  // CHECK: @llvm.x86.sse2.packssdw
+  // CHECK: %{{.*}} = shufflevector <4 x i32> %{{.*}}, <4 x i32> %{{.*}}, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+  // CHECK: %{{.*}} = icmp slt <8 x i32> %{{.*}}, <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = select <8 x i1> %{{.*}}, <8 x i32> %{{.*}}, <8 x i32> <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = icmp sgt <8 x i32> %{{.*}}, <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = select <8 x i1> %{{.*}}, <8 x i32> %{{.*}}, <8 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = trunc <8 x i32> %{{.*}} to <8 x i16>
   // CHECK: select <8 x i1> %{{.*}}, <8 x i16> %{{.*}}, <8 x i16> %{{.*}}
   return _mm_mask_packs_epi32(__W,__M,__A,__B); 
 }
 __m256i test_mm256_maskz_packs_epi32(__mmask16 __M, __m256i __A, __m256i __B) {
   // CHECK-LABEL: @test_mm256_maskz_packs_epi32
-  // CHECK: @llvm.x86.avx2.packssdw
+  // CHECK: %{{.*}} = shufflevector <8 x i32> %{{.*}}, <8 x i32> %{{.*}}, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11, i32 4, i32 5, i32 6, i32 7, i32 12, i32 13, i32 14, i32 15>
+  // CHECK: %{{.*}} = icmp slt <16 x i32> %{{.*}}, <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = icmp sgt <16 x i32> %{{.*}}, <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = trunc <16 x i32> %{{.*}} to <16 x i16>
   // CHECK: select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> %{{.*}}
   return _mm256_maskz_packs_epi32(__M,__A,__B); 
 }
 __m256i test_mm256_mask_packs_epi32(__m256i __W, __mmask16 __M, __m256i __A,       __m256i __B) {
   // CHECK-LABEL: @test_mm256_mask_packs_epi32
-  // CHECK: @llvm.x86.avx2.packssdw
+  // CHECK: %{{.*}} = shufflevector <8 x i32> %{{.*}}, <8 x i32> %{{.*}}, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11, i32 4, i32 5, i32 6, i32 7, i32 12, i32 13, i32 14, i32 15>
+  // CHECK: %{{.*}} = icmp slt <16 x i32> %{{.*}}, <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = icmp sgt <16 x i32> %{{.*}}, <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = trunc <16 x i32> %{{.*}} to <16 x i16>
   // CHECK: select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> %{{.*}}
   return _mm256_mask_packs_epi32(__W,__M,__A,__B); 
 }
 __m128i test_mm_maskz_packs_epi16(__mmask16 __M, __m128i __A, __m128i __B) {
   // CHECK-LABEL: @test_mm_maskz_packs_epi16
-  // CHECK: @llvm.x86.sse2.packsswb
+  // CHECK: %{{.*}} = shufflevector <8 x i16> %{{.*}}, <8 x i16> %{{.*}}, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
+  // CHECK: %{{.*}} = icmp slt <16 x i16> %{{.*}}, <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = icmp sgt <16 x i16> %{{.*}}, <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = trunc <16 x i16> %{{.*}} to <16 x i8>
   // CHECK: select <16 x i1> %{{.*}}, <16 x i8> %{{.*}}, <16 x i8> %{{.*}}
   return _mm_maskz_packs_epi16(__M,__A,__B); 
 }
 __m128i test_mm_mask_packs_epi16(__m128i __W, __mmask16 __M, __m128i __A,          __m128i __B) {
   // CHECK-LABEL: @test_mm_mask_packs_epi16
-  // CHECK: @llvm.x86.sse2.packsswb
+  // CHECK: %{{.*}} = shufflevector <8 x i16> %{{.*}}, <8 x i16> %{{.*}}, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
+  // CHECK: %{{.*}} = icmp slt <16 x i16> %{{.*}}, <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = icmp sgt <16 x i16> %{{.*}}, <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = trunc <16 x i16> %{{.*}} to <16 x i8>
   // CHECK: select <16 x i1> %{{.*}}, <16 x i8> %{{.*}}, <16 x i8> %{{.*}}
   return _mm_mask_packs_epi16(__W,__M,__A,__B); 
 }
 __m256i test_mm256_maskz_packs_epi16(__mmask32 __M, __m256i __A, __m256i __B) {
   // CHECK-LABEL: @test_mm256_maskz_packs_epi16
-  // CHECK: @llvm.x86.avx2.packsswb
+  // CHECK: %{{.*}} = shufflevector <16 x i16> %{{.*}}, <16 x i16> %{{.*}}, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31>
+  // CHECK: %{{.*}} = icmp slt <32 x i16> %{{.*}}, <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = icmp sgt <32 x i16> %{{.*}}, <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = trunc <32 x i16> %{{.*}} to <32 x i8>
   // CHECK: select <32 x i1> %{{.*}}, <32 x i8> %{{.*}}, <32 x i8> %{{.*}}
   return _mm256_maskz_packs_epi16(__M,__A,__B); 
 }
 __m256i test_mm256_mask_packs_epi16(__m256i __W, __mmask32 __M, __m256i __A,       __m256i __B) {
   // CHECK-LABEL: @test_mm256_mask_packs_epi16
-  // CHECK: @llvm.x86.avx2.packsswb
+  // CHECK: %{{.*}} = shufflevector <16 x i16> %{{.*}}, <16 x i16> %{{.*}}, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31>
+  // CHECK: %{{.*}} = icmp slt <32 x i16> %{{.*}}, <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = icmp sgt <32 x i16> %{{.*}}, <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = trunc <32 x i16> %{{.*}} to <32 x i8>
   // CHECK: select <32 x i1> %{{.*}}, <32 x i8> %{{.*}}, <32 x i8> %{{.*}}
   return _mm256_mask_packs_epi16(__W,__M,__A,__B); 
 }
 
 __m128i test_mm_mask_packus_epi32(__m128i __W, __mmask16 __M, __m128i __A,           __m128i __B) {
   // CHECK-LABEL: @test_mm_mask_packus_epi32
-  // CHECK: @llvm.x86.sse41.packusdw
+  // CHECK: %{{.*}} = shufflevector <4 x i32> %{{.*}}, <4 x i32> %{{.*}}, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+  // CHECK: %{{.*}} = icmp slt <8 x i32> %{{.*}}, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = select <8 x i1> %{{.*}}, <8 x i32> %{{.*}}, <8 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = icmp sgt <8 x i32> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <8 x i1> %{{.*}}, <8 x i32> %{{.*}}, <8 x i32> zeroinitializer
+  // CHECK: %{{.*}} = trunc <8 x i32> %{{.*}} to <8 x i16>
   // CHECK: select <8 x i1> %{{.*}}, <8 x i16> %{{.*}}, <8 x i16> %{{.*}}
   return _mm_mask_packus_epi32(__W,__M,__A,__B); 
 }
 
 __m128i test_mm_maskz_packus_epi32(__mmask8 __M, __m128i __A, __m128i __B) {
   // CHECK-LABEL: @test_mm_maskz_packus_epi32
-  // CHECK: @llvm.x86.sse41.packusdw
+  // CHECK: %{{.*}} = shufflevector <4 x i32> %{{.*}}, <4 x i32> %{{.*}}, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+  // CHECK: %{{.*}} = icmp slt <8 x i32> %{{.*}}, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = select <8 x i1> %{{.*}}, <8 x i32> %{{.*}}, <8 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = icmp sgt <8 x i32> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <8 x i1> %{{.*}}, <8 x i32> %{{.*}}, <8 x i32> zeroinitializer
+  // CHECK: %{{.*}} = trunc <8 x i32> %{{.*}} to <8 x i16>
   // CHECK: select <8 x i1> %{{.*}}, <8 x i16> %{{.*}}, <8 x i16> %{{.*}}
   return _mm_maskz_packus_epi32(__M,__A,__B); 
 }
 
 __m256i test_mm256_maskz_packus_epi32(__mmask16 __M, __m256i __A, __m256i __B) {
   // CHECK-LABEL: @test_mm256_maskz_packus_epi32
-  // CHECK: @llvm.x86.avx2.packusdw
+  // CHECK: %{{.*}} = shufflevector <8 x i32> %{{.*}}, <8 x i32> %{{.*}}, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11, i32 4, i32 5, i32 6, i32 7, i32 12, i32 13, i32 14, i32 15>
+  // CHECK: %{{.*}} = icmp slt <16 x i32> %{{.*}}, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = icmp sgt <16 x i32> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> zeroinitializer
+  // CHECK: %{{.*}} = trunc <16 x i32> %{{.*}} to <16 x i16>
   // CHECK: select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> %{{.*}}
   return _mm256_maskz_packus_epi32(__M,__A,__B); 
 }
 
 __m256i test_mm256_mask_packus_epi32(__m256i __W, __mmask16 __M, __m256i __A,        __m256i __B) {
   // CHECK-LABEL: @test_mm256_mask_packus_epi32
-  // CHECK: @llvm.x86.avx2.packusdw
+  // CHECK: %{{.*}} = shufflevector <8 x i32> %{{.*}}, <8 x i32> %{{.*}}, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 8, i32 9, i32 10, i32 11, i32 4, i32 5, i32 6, i32 7, i32 12, i32 13, i32 14, i32 15>
+  // CHECK: %{{.*}} = icmp slt <16 x i32> %{{.*}}, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = icmp sgt <16 x i32> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i32> %{{.*}}, <16 x i32> zeroinitializer
+  // CHECK: %{{.*}} = trunc <16 x i32> %{{.*}} to <16 x i16>
   // CHECK: select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> %{{.*}}
   return _mm256_mask_packus_epi32(__W,__M,__A,__B); 
 }
 
 __m128i test_mm_maskz_packus_epi16(__mmask16 __M, __m128i __A, __m128i __B) {
   // CHECK-LABEL: @test_mm_maskz_packus_epi16
-  // CHECK: @llvm.x86.sse2.packuswb
+  // CHECK: %{{.*}} = shufflevector <8 x i16> %{{.*}}, <8 x i16> %{{.*}}, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
+  // CHECK: %{{.*}} = icmp slt <16 x i16> %{{.*}}, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = icmp sgt <16 x i16> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> zeroinitializer
+  // CHECK: %{{.*}} = trunc <16 x i16> %{{.*}} to <16 x i8>
   // CHECK: select <16 x i1> %{{.*}}, <16 x i8> %{{.*}}, <16 x i8> %{{.*}}
   return _mm_maskz_packus_epi16(__M,__A,__B); 
 }
 
 __m128i test_mm_mask_packus_epi16(__m128i __W, __mmask16 __M, __m128i __A,           __m128i __B) {
   // CHECK-LABEL: @test_mm_mask_packus_epi16
-  // CHECK: @llvm.x86.sse2.packuswb
+  // CHECK: %{{.*}} = shufflevector <8 x i16> %{{.*}}, <8 x i16> %{{.*}}, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
+  // CHECK: %{{.*}} = icmp slt <16 x i16> %{{.*}}, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = icmp sgt <16 x i16> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> zeroinitializer
+  // CHECK: %{{.*}} = trunc <16 x i16> %{{.*}} to <16 x i8>
   // CHECK: select <16 x i1> %{{.*}}, <16 x i8> %{{.*}}, <16 x i8> %{{.*}}
   return _mm_mask_packus_epi16(__W,__M,__A,__B); 
 }
 
 __m256i test_mm256_maskz_packus_epi16(__mmask32 __M, __m256i __A, __m256i __B) {
   // CHECK-LABEL: @test_mm256_maskz_packus_epi16
-  // CHECK: @llvm.x86.avx2.packuswb
+  // CHECK: %{{.*}} = shufflevector <16 x i16> %{{.*}}, <16 x i16> %{{.*}}, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31>
+  // CHECK: %{{.*}} = icmp slt <32 x i16> %{{.*}}, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = icmp sgt <32 x i16> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> zeroinitializer
+  // CHECK: %{{.*}} = trunc <32 x i16> %{{.*}} to <32 x i8>
   // CHECK: select <32 x i1> %{{.*}}, <32 x i8> %{{.*}}, <32 x i8> %{{.*}}
   return _mm256_maskz_packus_epi16(__M,__A,__B); 
 }
 
 __m256i test_mm256_mask_packus_epi16(__m256i __W, __mmask32 __M, __m256i __A,        __m256i __B) {
   // CHECK-LABEL: @test_mm256_mask_packus_epi16
-  // CHECK: @llvm.x86.avx2.packuswb
+  // CHECK: %{{.*}} = shufflevector <16 x i16> %{{.*}}, <16 x i16> %{{.*}}, <32 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31>
+  // CHECK: %{{.*}} = icmp slt <32 x i16> %{{.*}}, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = icmp sgt <32 x i16> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <32 x i1> %{{.*}}, <32 x i16> %{{.*}}, <32 x i16> zeroinitializer
+  // CHECK: %{{.*}} = trunc <32 x i16> %{{.*}} to <32 x i8>
   // CHECK: select <32 x i1> %{{.*}}, <32 x i8> %{{.*}}, <32 x i8> %{{.*}}
   return _mm256_mask_packus_epi16(__W,__M,__A,__B); 
 }
 
 __m128i test_mm_mask_adds_epi8(__m128i __W, __mmask16 __U, __m128i __A,        __m128i __B) {
   // CHECK-LABEL: @test_mm_mask_adds_epi8
   // CHECK: @llvm.x86.sse2.padds.b
   // CHECK: select <16 x i1> %{{.*}}, <16 x i8> %{{.*}}, <16 x i8> %{{.*}}

clang/test/CodeGen/sse2-builtins.c

 __m128i test_mm_or_si128(__m128i A, __m128i B) {
   // CHECK-LABEL: test_mm_or_si128
   // CHECK: or <2 x i64> %{{.*}}, %{{.*}}
   return _mm_or_si128(A, B);
 }
 
 __m128i test_mm_packs_epi16(__m128i A, __m128i B) {
   // CHECK-LABEL: test_mm_packs_epi16
-  // CHECK: call <16 x i8> @llvm.x86.sse2.packsswb.128(<8 x i16> %{{.*}}, <8 x i16> %{{.*}})
+  // CHECK: %{{.*}} = shufflevector <8 x i16> %{{.*}}, <8 x i16> %{{.*}}, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
+  // CHECK: %{{.*}} = icmp slt <16 x i16> %{{.*}}, <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> <i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127, i16 127>
+  // CHECK: %{{.*}} = icmp sgt <16 x i16> %{{.*}}, <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> <i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128, i16 -128>
+  // CHECK: %{{.*}} = trunc <16 x i16> %{{.*}} to <16 x i8>
   return _mm_packs_epi16(A, B);
 }
 
 __m128i test_mm_packs_epi32(__m128i A, __m128i B) {
   // CHECK-LABEL: test_mm_packs_epi32
-  // CHECK: call <8 x i16> @llvm.x86.sse2.packssdw.128(<4 x i32> %{{.*}}, <4 x i32> %{{.*}})
+  // CHECK: %{{.*}} = shufflevector <4 x i32> %{{.*}}, <4 x i32> %{{.*}}, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+  // CHECK: %{{.*}} = icmp slt <8 x i32> %{{.*}}, <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = select <8 x i1> %{{.*}}, <8 x i32> %{{.*}}, <8 x i32> <i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767, i32 32767>
+  // CHECK: %{{.*}} = icmp sgt <8 x i32> %{{.*}}, <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = select <8 x i1> %{{.*}}, <8 x i32> %{{.*}}, <8 x i32> <i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768, i32 -32768>
+  // CHECK: %{{.*}} = trunc <8 x i32> %{{.*}} to <8 x i16>
   return _mm_packs_epi32(A, B);
 }
 
 __m128i test_mm_packus_epi16(__m128i A, __m128i B) {
   // CHECK-LABEL: test_mm_packus_epi16
-  // CHECK: call <16 x i8> @llvm.x86.sse2.packuswb.128(<8 x i16> %{{.*}}, <8 x i16> %{{.*}})
+  // CHECK: %{{.*}} = shufflevector <8 x i16> %{{.*}}, <8 x i16> %{{.*}}, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
+  // CHECK: %{{.*}} = icmp slt <16 x i16> %{{.*}}, <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> <i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255, i16 255>
+  // CHECK: %{{.*}} = icmp sgt <16 x i16> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <16 x i1> %{{.*}}, <16 x i16> %{{.*}}, <16 x i16> zeroinitializer
+  // CHECK: %{{.*}} = trunc <16 x i16> %{{.*}} to <16 x i8>
   return _mm_packus_epi16(A, B);
 }
 
 void test_mm_pause() {
   // CHECK-LABEL: test_mm_pause
   // CHECK: call void @llvm.x86.sse2.pause()
   return _mm_pause();
 }

clang/test/CodeGen/sse41-builtins.c

 __m128i test_mm_mullo_epi32(__m128i x, __m128i y) {
   // CHECK-LABEL: test_mm_mullo_epi32
   // CHECK: mul <4 x i32>
   return _mm_mullo_epi32(x, y);
 }
 
 __m128i test_mm_packus_epi32(__m128i x, __m128i y) {
   // CHECK-LABEL: test_mm_packus_epi32
-  // CHECK: call <8 x i16> @llvm.x86.sse41.packusdw(<4 x i32> %{{.*}}, <4 x i32> %{{.*}})
+  // CHECK: %{{.*}} = shufflevector <4 x i32> %{{.*}}, <4 x i32> %{{.*}}, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
+  // CHECK: %{{.*}} = icmp slt <8 x i32> %{{.*}}, <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = select <8 x i1> %{{.*}}, <8 x i32> %{{.*}}, <8 x i32> <i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535, i32 65535>
+  // CHECK: %{{.*}} = icmp sgt <8 x i32> %{{.*}}, zeroinitializer
+  // CHECK: %{{.*}} = select <8 x i1> %{{.*}}, <8 x i32> %{{.*}}, <8 x i32> zeroinitializer
+  // CHECK: %{{.*}} = trunc <8 x i32> %{{.*}} to <8 x i16>
   return _mm_packus_epi32(x, y);
 }
 
 __m128d test_mm_round_pd(__m128d x) {
   // CHECK-LABEL: test_mm_round_pd
   // CHECK: call <2 x double> @llvm.x86.sse41.round.pd(<2 x double> %{{.*}}, i32 4)
   return _mm_round_pd(x, 4);
 }