[X86] Add versions of the avx512 gather intrinsics that take the mask as a vXi1 vector instead of a scalar

[craig.topper]

Summary

In keeping with our general direction of having the vXi1 type present in IR, this patch converts the mask argument for avx512 gather to vXi1. This can avoid k-register to GPR to k-register transitions late in codegen.

I left the existing intrinsics behind because they have many out of tree users such as ISPC. They generate their own code and don't go through the autoupgrade path which only works for bitcode and ll parsing. Ideally we will get them to migrate to target independent intrinsics, but it might be easier for them to migrate to these new intrinsics.

I'll work on scatter and gatherpf/scatterpf after next. Hopefully we can get this all done for 8.0 so we can maybe migrate everyone before 9.0.

[RKSimon]

I'm not sure its a good idea to completely lose the CHECKs from the intrinsic tests

[craig.topper]

Oops that was a mistake. I meant to regenerate after using the autoupgrade code to get the new IR out of opt.

[craig.topper]

Add back the check lines. Make a copy of the test for the old intrinsics.

[craig.topper]

Fix accidental renaming of the tests that prevented the script from working

[RKSimon]

I'm not totally clear on what relationship these have to the generic gather intrinsics: llvm.masked.gather.* - shouldn't we be trying to use those? The vector of pointers vs base pointer + vector of indices I guess is the issue?

[craig.topper]

Yeah that’s the issue. The software interface for the intrinsics passes base, index, and scale separately. The generic intrinsics expect a GEP they can extract the info from. I suppose with the right casting we could form a GEP, but that seems ugly.

I do want to push the out of tree users to use the generic versions where possible since they’re doing something similar to the loop vectorizer rather than the software intrinsics. So they should be better able to form a vector GEP from the scalar GEP they probably already have.

[RKSimon]

LGTM - ideally we'd have a way for people to call autoupgrade directly for cases like this and just remove them, but looks ok as long as these aren't going to linger over too many versions (ironically like autoupgrade......).

llvm/trunk/include/llvm/IR/IntrinsicsX86.td

let TargetPrefix = "x86" in {
   def int_x86_avx512_dbpsadbw_512 :
          GCCBuiltin<"__builtin_ia32_dbpsadbw512">,
           Intrinsic<[llvm_v32i16_ty],
                     [llvm_v64i8_ty, llvm_v64i8_ty, llvm_i32_ty], [IntrNoMem]>;
 }
 
 // Gather and Scatter ops
 let TargetPrefix = "x86" in {
-  def int_x86_avx512_gather_dpd_512  : GCCBuiltin<"__builtin_ia32_gathersiv8df">,
+  // These are gather intrinsics that use a scalar integer for the mask. They
+  // have been superceded by new versions that use a vXi1 mask. Leaving these
+  // for now as they have multiple out of tree users that need to migrate.
+  // TODO: Remove when we can confirm out of tree migration.
+  def int_x86_avx512_gather_dpd_512  :
           Intrinsic<[llvm_v8f64_ty], [llvm_v8f64_ty, llvm_ptr_ty,
                      llvm_v8i32_ty, llvm_i8_ty, llvm_i32_ty],
                     [IntrReadMem, IntrArgMemOnly]>;
-  def int_x86_avx512_gather_dps_512  : GCCBuiltin<"__builtin_ia32_gathersiv16sf">,
+  def int_x86_avx512_gather_dps_512  :
           Intrinsic<[llvm_v16f32_ty], [llvm_v16f32_ty, llvm_ptr_ty,
                      llvm_v16i32_ty, llvm_i16_ty, llvm_i32_ty],
                     [IntrReadMem, IntrArgMemOnly]>;
-  def int_x86_avx512_gather_qpd_512  : GCCBuiltin<"__builtin_ia32_gatherdiv8df">,
+  def int_x86_avx512_gather_qpd_512  :
           Intrinsic<[llvm_v8f64_ty], [llvm_v8f64_ty, llvm_ptr_ty,
                      llvm_v8i64_ty, llvm_i8_ty, llvm_i32_ty],
                     [IntrReadMem, IntrArgMemOnly]>;
-  def int_x86_avx512_gather_qps_512  : GCCBuiltin<"__builtin_ia32_gatherdiv16sf">,
+  def int_x86_avx512_gather_qps_512  :
           Intrinsic<[llvm_v8f32_ty], [llvm_v8f32_ty, llvm_ptr_ty,
                      llvm_v8i64_ty, llvm_i8_ty, llvm_i32_ty],
                     [IntrReadMem, IntrArgMemOnly]>;
 
 
-  def int_x86_avx512_gather_dpq_512  : GCCBuiltin<"__builtin_ia32_gathersiv8di">,
+  def int_x86_avx512_gather_dpq_512  :
           Intrinsic<[llvm_v8i64_ty], [llvm_v8i64_ty, llvm_ptr_ty,
                      llvm_v8i32_ty, llvm_i8_ty, llvm_i32_ty],
                     [IntrReadMem, IntrArgMemOnly]>;
-  def int_x86_avx512_gather_dpi_512  : GCCBuiltin<"__builtin_ia32_gathersiv16si">,
+  def int_x86_avx512_gather_dpi_512  :
           Intrinsic<[llvm_v16i32_ty], [llvm_v16i32_ty, llvm_ptr_ty,
                      llvm_v16i32_ty, llvm_i16_ty, llvm_i32_ty],
                     [IntrReadMem, IntrArgMemOnly]>;
-  def int_x86_avx512_gather_qpq_512  : GCCBuiltin<"__builtin_ia32_gatherdiv8di">,
+  def int_x86_avx512_gather_qpq_512  :
           Intrinsic<[llvm_v8i64_ty], [llvm_v8i64_ty, llvm_ptr_ty,
                      llvm_v8i64_ty, llvm_i8_ty, llvm_i32_ty],
                     [IntrReadMem, IntrArgMemOnly]>;
-  def int_x86_avx512_gather_qpi_512  : GCCBuiltin<"__builtin_ia32_gatherdiv16si">,
+  def int_x86_avx512_gather_qpi_512  :
           Intrinsic<[llvm_v8i32_ty], [llvm_v8i32_ty, llvm_ptr_ty,
                      llvm_v8i64_ty, llvm_i8_ty, llvm_i32_ty],
                     [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3div2_df :
-        GCCBuiltin<"__builtin_ia32_gather3div2df">,
           Intrinsic<[llvm_v2f64_ty],
           [llvm_v2f64_ty, llvm_ptr_ty, llvm_v2i64_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3div2_di :
-        GCCBuiltin<"__builtin_ia32_gather3div2di">,
           Intrinsic<[llvm_v2i64_ty],
           [llvm_v2i64_ty, llvm_ptr_ty, llvm_v2i64_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3div4_df :
-        GCCBuiltin<"__builtin_ia32_gather3div4df">,
           Intrinsic<[llvm_v4f64_ty],
           [llvm_v4f64_ty, llvm_ptr_ty, llvm_v4i64_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3div4_di :
-        GCCBuiltin<"__builtin_ia32_gather3div4di">,
           Intrinsic<[llvm_v4i64_ty],
           [llvm_v4i64_ty, llvm_ptr_ty, llvm_v4i64_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3div4_sf :
-        GCCBuiltin<"__builtin_ia32_gather3div4sf">,
           Intrinsic<[llvm_v4f32_ty],
           [llvm_v4f32_ty, llvm_ptr_ty, llvm_v2i64_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3div4_si :
-        GCCBuiltin<"__builtin_ia32_gather3div4si">,
           Intrinsic<[llvm_v4i32_ty],
           [llvm_v4i32_ty, llvm_ptr_ty, llvm_v2i64_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3div8_sf :
-        GCCBuiltin<"__builtin_ia32_gather3div8sf">,
           Intrinsic<[llvm_v4f32_ty],
           [llvm_v4f32_ty, llvm_ptr_ty, llvm_v4i64_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3div8_si :
-        GCCBuiltin<"__builtin_ia32_gather3div8si">,
           Intrinsic<[llvm_v4i32_ty],
           [llvm_v4i32_ty, llvm_ptr_ty, llvm_v4i64_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3siv2_df :
-        GCCBuiltin<"__builtin_ia32_gather3siv2df">,
           Intrinsic<[llvm_v2f64_ty],
           [llvm_v2f64_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3siv2_di :
-        GCCBuiltin<"__builtin_ia32_gather3siv2di">,
           Intrinsic<[llvm_v2i64_ty],
           [llvm_v2i64_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3siv4_df :
-        GCCBuiltin<"__builtin_ia32_gather3siv4df">,
           Intrinsic<[llvm_v4f64_ty],
           [llvm_v4f64_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3siv4_di :
-        GCCBuiltin<"__builtin_ia32_gather3siv4di">,
           Intrinsic<[llvm_v4i64_ty],
           [llvm_v4i64_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3siv4_sf :
-        GCCBuiltin<"__builtin_ia32_gather3siv4sf">,
           Intrinsic<[llvm_v4f32_ty],
           [llvm_v4f32_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3siv4_si :
-        GCCBuiltin<"__builtin_ia32_gather3siv4si">,
           Intrinsic<[llvm_v4i32_ty],
           [llvm_v4i32_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3siv8_sf :
-        GCCBuiltin<"__builtin_ia32_gather3siv8sf">,
           Intrinsic<[llvm_v8f32_ty],
           [llvm_v8f32_ty, llvm_ptr_ty, llvm_v8i32_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
   def int_x86_avx512_gather3siv8_si :
-        GCCBuiltin<"__builtin_ia32_gather3siv8si">,
           Intrinsic<[llvm_v8i32_ty],
           [llvm_v8i32_ty, llvm_ptr_ty, llvm_v8i32_ty, llvm_i8_ty, llvm_i32_ty],
           [IntrReadMem, IntrArgMemOnly]>;
 
 // scatter
   def int_x86_avx512_scatter_dpd_512  : GCCBuiltin<"__builtin_ia32_scattersiv8df">,
           Intrinsic<[], [llvm_ptr_ty, llvm_i8_ty,
                         llvm_v8i32_ty, llvm_v8f64_ty, llvm_i32_ty],
// scatter
   def int_x86_avx512_scatterpf_qpd_512  : GCCBuiltin<"__builtin_ia32_scatterpfqpd">,
           Intrinsic<[], [llvm_i8_ty, llvm_v8i64_ty, llvm_ptr_ty,
                      llvm_i32_ty, llvm_i32_ty], [IntrArgMemOnly]>;
   def int_x86_avx512_scatterpf_qps_512  : GCCBuiltin<"__builtin_ia32_scatterpfqps">,
           Intrinsic<[], [llvm_i8_ty, llvm_v8i64_ty, llvm_ptr_ty,
                      llvm_i32_ty, llvm_i32_ty], [IntrArgMemOnly]>;
 }
 
+// AVX512 gather intrinsics that use vXi1 masks.
+let TargetPrefix = "x86" in {
+  def int_x86_avx512_mask_gather_dpd_512  :
+          Intrinsic<[llvm_v8f64_ty], [llvm_v8f64_ty, llvm_ptr_ty,
+                     llvm_v8i32_ty, llvm_v8i1_ty, llvm_i32_ty],
+                    [IntrReadMem, IntrArgMemOnly]>;
+  def int_x86_avx512_mask_gather_dps_512  :
+          Intrinsic<[llvm_v16f32_ty], [llvm_v16f32_ty, llvm_ptr_ty,
+                     llvm_v16i32_ty, llvm_v16i1_ty, llvm_i32_ty],
+                    [IntrReadMem, IntrArgMemOnly]>;
+  def int_x86_avx512_mask_gather_qpd_512  :
+          Intrinsic<[llvm_v8f64_ty], [llvm_v8f64_ty, llvm_ptr_ty,
+                     llvm_v8i64_ty, llvm_v8i1_ty, llvm_i32_ty],
+                    [IntrReadMem, IntrArgMemOnly]>;
+  def int_x86_avx512_mask_gather_qps_512  :
+          Intrinsic<[llvm_v8f32_ty], [llvm_v8f32_ty, llvm_ptr_ty,
+                     llvm_v8i64_ty, llvm_v8i1_ty, llvm_i32_ty],
+                    [IntrReadMem, IntrArgMemOnly]>;
+
+
+  def int_x86_avx512_mask_gather_dpq_512  :
+          Intrinsic<[llvm_v8i64_ty], [llvm_v8i64_ty, llvm_ptr_ty,
+                     llvm_v8i32_ty, llvm_v8i1_ty, llvm_i32_ty],
+                    [IntrReadMem, IntrArgMemOnly]>;
+  def int_x86_avx512_mask_gather_dpi_512  :
+          Intrinsic<[llvm_v16i32_ty], [llvm_v16i32_ty, llvm_ptr_ty,
+                     llvm_v16i32_ty, llvm_v16i1_ty, llvm_i32_ty],
+                    [IntrReadMem, IntrArgMemOnly]>;
+  def int_x86_avx512_mask_gather_qpq_512  :
+          Intrinsic<[llvm_v8i64_ty], [llvm_v8i64_ty, llvm_ptr_ty,
+                     llvm_v8i64_ty, llvm_v8i1_ty, llvm_i32_ty],
+                    [IntrReadMem, IntrArgMemOnly]>;
+  def int_x86_avx512_mask_gather_qpi_512  :
+          Intrinsic<[llvm_v8i32_ty], [llvm_v8i32_ty, llvm_ptr_ty,
+                     llvm_v8i64_ty, llvm_v8i1_ty, llvm_i32_ty],
+                    [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3div2_df :
+          Intrinsic<[llvm_v2f64_ty],
+          [llvm_v2f64_ty, llvm_ptr_ty, llvm_v2i64_ty, llvm_v2i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3div2_di :
+          Intrinsic<[llvm_v2i64_ty],
+          [llvm_v2i64_ty, llvm_ptr_ty, llvm_v2i64_ty, llvm_v2i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3div4_df :
+          Intrinsic<[llvm_v4f64_ty],
+          [llvm_v4f64_ty, llvm_ptr_ty, llvm_v4i64_ty, llvm_v4i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3div4_di :
+          Intrinsic<[llvm_v4i64_ty],
+          [llvm_v4i64_ty, llvm_ptr_ty, llvm_v4i64_ty, llvm_v4i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3div4_sf :
+          Intrinsic<[llvm_v4f32_ty],
+          [llvm_v4f32_ty, llvm_ptr_ty, llvm_v2i64_ty, llvm_v2i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3div4_si :
+          Intrinsic<[llvm_v4i32_ty],
+          [llvm_v4i32_ty, llvm_ptr_ty, llvm_v2i64_ty, llvm_v2i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3div8_sf :
+          Intrinsic<[llvm_v4f32_ty],
+          [llvm_v4f32_ty, llvm_ptr_ty, llvm_v4i64_ty, llvm_v4i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3div8_si :
+          Intrinsic<[llvm_v4i32_ty],
+          [llvm_v4i32_ty, llvm_ptr_ty, llvm_v4i64_ty, llvm_v4i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3siv2_df :
+          Intrinsic<[llvm_v2f64_ty],
+          [llvm_v2f64_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_v2i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3siv2_di :
+          Intrinsic<[llvm_v2i64_ty],
+          [llvm_v2i64_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_v2i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3siv4_df :
+          Intrinsic<[llvm_v4f64_ty],
+          [llvm_v4f64_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_v4i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3siv4_di :
+          Intrinsic<[llvm_v4i64_ty],
+          [llvm_v4i64_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_v4i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3siv4_sf :
+          Intrinsic<[llvm_v4f32_ty],
+          [llvm_v4f32_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_v4i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3siv4_si :
+          Intrinsic<[llvm_v4i32_ty],
+          [llvm_v4i32_ty, llvm_ptr_ty, llvm_v4i32_ty, llvm_v4i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3siv8_sf :
+          Intrinsic<[llvm_v8f32_ty],
+          [llvm_v8f32_ty, llvm_ptr_ty, llvm_v8i32_ty, llvm_v8i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+
+  def int_x86_avx512_mask_gather3siv8_si :
+          Intrinsic<[llvm_v8i32_ty],
+          [llvm_v8i32_ty, llvm_ptr_ty, llvm_v8i32_ty, llvm_v8i1_ty, llvm_i32_ty],
+          [IntrReadMem, IntrArgMemOnly]>;
+}
+
 // AVX-512 conflict detection instruction
 // Instructions that count the number of leading zero bits
 let TargetPrefix = "x86" in {
   def int_x86_avx512_mask_conflict_d_128 :
           GCCBuiltin<"__builtin_ia32_vpconflictsi_128_mask">,
           Intrinsic<[llvm_v4i32_ty],
                     [llvm_v4i32_ty, llvm_v4i32_ty, llvm_i8_ty],
                     [IntrNoMem]>;

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

static SDValue getGatherNode(unsigned Opc, SDValue Op, SelectionDAG &DAG,
   // Scale must be constant.
   if (!C)
     return SDValue();
   SDValue Scale = DAG.getTargetConstant(C->getZExtValue(), dl, MVT::i8);
   unsigned MinElts = std::min(Index.getSimpleValueType().getVectorNumElements(),
                               VT.getVectorNumElements());
   MVT MaskVT = MVT::getVectorVT(MVT::i1, MinElts);
 
-  SDValue VMask = getMaskNode(Mask, MaskVT, Subtarget, DAG, dl);
+  // We support two versions of the gather intrinsics. One with scalar mask and
+  // one with vXi1 mask. Convert scalar to vXi1 if necessary.
+  if (Mask.getValueType() != MaskVT)
+    Mask = getMaskNode(Mask, MaskVT, Subtarget, DAG, dl);
+
   SDVTList VTs = DAG.getVTList(Op.getValueType(), MaskVT, MVT::Other);
   SDValue Disp = DAG.getTargetConstant(0, dl, MVT::i32);
   SDValue Segment = DAG.getRegister(0, MVT::i32);
   // If source is undef or we know it won't be used, use a zero vector
   // to break register dependency.
   // TODO: use undef instead and let BreakFalseDeps deal with it?
-  if (Src.isUndef() || ISD::isBuildVectorAllOnes(VMask.getNode()))
+  if (Src.isUndef() || ISD::isBuildVectorAllOnes(Mask.getNode()))
     Src = getZeroVector(Op.getSimpleValueType(), Subtarget, DAG, dl);
-  SDValue Ops[] = {Src, VMask, Base, Scale, Index, Disp, Segment, Chain};
+  SDValue Ops[] = {Src, Mask, Base, Scale, Index, Disp, Segment, Chain};
   SDNode *Res = DAG.getMachineNode(Opc, dl, VTs, Ops);
   SDValue RetOps[] = { SDValue(Res, 0), SDValue(Res, 2) };
   return DAG.getMergeValues(RetOps, dl);
 }
 
 static SDValue getScatterNode(unsigned Opc, SDValue Op, SelectionDAG &DAG,
                                SDValue Src, SDValue Mask, SDValue Base,
                                SDValue Index, SDValue ScaleOp, SDValue Chain,

llvm/trunk/lib/Target/X86/X86IntrinsicsInfo.h

X86_INTRINSIC_DATA(avx512_gatherpf_dpd_512, PREFETCH,
                      X86::VGATHERPF0DPDm, X86::VGATHERPF1DPDm),
   X86_INTRINSIC_DATA(avx512_gatherpf_dps_512, PREFETCH,
                      X86::VGATHERPF0DPSm, X86::VGATHERPF1DPSm),
   X86_INTRINSIC_DATA(avx512_gatherpf_qpd_512, PREFETCH,
                      X86::VGATHERPF0QPDm, X86::VGATHERPF1QPDm),
   X86_INTRINSIC_DATA(avx512_gatherpf_qps_512, PREFETCH,
                      X86::VGATHERPF0QPSm, X86::VGATHERPF1QPSm),
 
+  X86_INTRINSIC_DATA(avx512_mask_gather_dpd_512, GATHER, X86::VGATHERDPDZrm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather_dpi_512, GATHER, X86::VPGATHERDDZrm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather_dpq_512, GATHER, X86::VPGATHERDQZrm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather_dps_512, GATHER, X86::VGATHERDPSZrm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather_qpd_512, GATHER, X86::VGATHERQPDZrm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather_qpi_512, GATHER, X86::VPGATHERQDZrm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather_qpq_512, GATHER, X86::VPGATHERQQZrm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather_qps_512, GATHER, X86::VGATHERQPSZrm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3div2_df, GATHER, X86::VGATHERQPDZ128rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3div2_di, GATHER, X86::VPGATHERQQZ128rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3div4_df, GATHER, X86::VGATHERQPDZ256rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3div4_di, GATHER, X86::VPGATHERQQZ256rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3div4_sf, GATHER, X86::VGATHERQPSZ128rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3div4_si, GATHER, X86::VPGATHERQDZ128rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3div8_sf, GATHER, X86::VGATHERQPSZ256rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3div8_si, GATHER, X86::VPGATHERQDZ256rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3siv2_df, GATHER, X86::VGATHERDPDZ128rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3siv2_di, GATHER, X86::VPGATHERDQZ128rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3siv4_df, GATHER, X86::VGATHERDPDZ256rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3siv4_di, GATHER, X86::VPGATHERDQZ256rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3siv4_sf, GATHER, X86::VGATHERDPSZ128rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3siv4_si, GATHER, X86::VPGATHERDDZ128rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3siv8_sf, GATHER, X86::VGATHERDPSZ256rm, 0),
+  X86_INTRINSIC_DATA(avx512_mask_gather3siv8_si, GATHER, X86::VPGATHERDDZ256rm, 0),
+
   X86_INTRINSIC_DATA(avx512_mask_pmov_db_mem_128, TRUNCATE_TO_MEM_VI8,
                      X86ISD::VTRUNC, 0),
   X86_INTRINSIC_DATA(avx512_mask_pmov_db_mem_256, TRUNCATE_TO_MEM_VI8,
                      X86ISD::VTRUNC, 0),
   X86_INTRINSIC_DATA(avx512_mask_pmov_db_mem_512, TRUNCATE_TO_MEM_VI8,
                      X86ISD::VTRUNC, 0),
   X86_INTRINSIC_DATA(avx512_mask_pmov_dw_mem_128, TRUNCATE_TO_MEM_VI16,
                      X86ISD::VTRUNC, 0),

llvm/trunk/test/CodeGen/X86/avx512-gather-scatter-intrin-deprecated.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc < %s -mtriple=x86_64-apple-darwin -mattr=+avx512f,+avx512bw,+avx512vl,+avx512dq | FileCheck %s
+
+declare <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float>, i8*, <16 x i32>, i16, i32)
+declare void @llvm.x86.avx512.scatter.dps.512 (i8*, i16, <16 x i32>, <16 x float>, i32)
+declare <8 x double> @llvm.x86.avx512.gather.dpd.512 (<8 x double>, i8*, <8 x i32>, i8, i32)
+declare void @llvm.x86.avx512.scatter.dpd.512 (i8*, i8, <8 x i32>, <8 x double>, i32)
+
+declare <8 x float> @llvm.x86.avx512.gather.qps.512 (<8 x float>, i8*, <8 x i64>, i8, i32)
+declare void @llvm.x86.avx512.scatter.qps.512 (i8*, i8, <8 x i64>, <8 x float>, i32)
+declare <8 x double> @llvm.x86.avx512.gather.qpd.512 (<8 x double>, i8*, <8 x i64>, i8, i32)
+declare void @llvm.x86.avx512.scatter.qpd.512 (i8*, i8, <8 x i64>, <8 x double>, i32)
+
+define void @gather_mask_dps(<16 x i32> %ind, <16 x float> %src, i16 %mask, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: gather_mask_dps:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    kmovq %k1, %k2
+; CHECK-NEXT:    vgatherdps (%rsi,%zmm0,4), %zmm1 {%k2}
+; CHECK-NEXT:    vpaddd {{.*}}(%rip), %zmm0, %zmm0
+; CHECK-NEXT:    vscatterdps %zmm1, (%rdx,%zmm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = call <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float> %src, i8* %base, <16 x i32>%ind, i16 %mask, i32 4)
+  %ind2 = add <16 x i32> %ind, <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
+  call void @llvm.x86.avx512.scatter.dps.512 (i8* %stbuf, i16 %mask, <16 x i32>%ind2, <16 x float> %x, i32 4)
+  ret void
+}
+
+define void @gather_mask_dpd(<8 x i32> %ind, <8 x double> %src, i8 %mask, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: gather_mask_dpd:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    kmovq %k1, %k2
+; CHECK-NEXT:    vgatherdpd (%rsi,%ymm0,4), %zmm1 {%k2}
+; CHECK-NEXT:    vpaddd {{.*}}(%rip), %ymm0, %ymm0
+; CHECK-NEXT:    vscatterdpd %zmm1, (%rdx,%ymm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = call <8 x double> @llvm.x86.avx512.gather.dpd.512 (<8 x double> %src, i8* %base, <8 x i32>%ind, i8 %mask, i32 4)
+  %ind2 = add <8 x i32> %ind, <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
+  call void @llvm.x86.avx512.scatter.dpd.512 (i8* %stbuf, i8 %mask, <8 x i32>%ind2, <8 x double> %x, i32 4)
+  ret void
+}
+
+define void @gather_mask_qps(<8 x i64> %ind, <8 x float> %src, i8 %mask, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: gather_mask_qps:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    kmovq %k1, %k2
+; CHECK-NEXT:    vgatherqps (%rsi,%zmm0,4), %ymm1 {%k2}
+; CHECK-NEXT:    vpaddq {{.*}}(%rip), %zmm0, %zmm0
+; CHECK-NEXT:    vscatterqps %ymm1, (%rdx,%zmm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = call <8 x float> @llvm.x86.avx512.gather.qps.512 (<8 x float> %src, i8* %base, <8 x i64>%ind, i8 %mask, i32 4)
+  %ind2 = add <8 x i64> %ind, <i64 0, i64 1, i64 2, i64 3, i64 0, i64 1, i64 2, i64 3>
+  call void @llvm.x86.avx512.scatter.qps.512 (i8* %stbuf, i8 %mask, <8 x i64>%ind2, <8 x float> %x, i32 4)
+  ret void
+}
+
+define void @gather_mask_qpd(<8 x i64> %ind, <8 x double> %src, i8 %mask, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: gather_mask_qpd:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    kmovq %k1, %k2
+; CHECK-NEXT:    vgatherqpd (%rsi,%zmm0,4), %zmm1 {%k2}
+; CHECK-NEXT:    vpaddq {{.*}}(%rip), %zmm0, %zmm0
+; CHECK-NEXT:    vscatterqpd %zmm1, (%rdx,%zmm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = call <8 x double> @llvm.x86.avx512.gather.qpd.512 (<8 x double> %src, i8* %base, <8 x i64>%ind, i8 %mask, i32 4)
+  %ind2 = add <8 x i64> %ind, <i64 0, i64 1, i64 2, i64 3, i64 0, i64 1, i64 2, i64 3>
+  call void @llvm.x86.avx512.scatter.qpd.512 (i8* %stbuf, i8 %mask, <8 x i64>%ind2, <8 x double> %x, i32 4)
+  ret void
+}
+;;
+;; Integer Gather/Scatter
+;;
+declare <16 x i32> @llvm.x86.avx512.gather.dpi.512 (<16 x i32>, i8*, <16 x i32>, i16, i32)
+declare void @llvm.x86.avx512.scatter.dpi.512 (i8*, i16, <16 x i32>, <16 x i32>, i32)
+declare <8 x i64> @llvm.x86.avx512.gather.dpq.512 (<8 x i64>, i8*, <8 x i32>, i8, i32)
+declare void @llvm.x86.avx512.scatter.dpq.512 (i8*, i8, <8 x i32>, <8 x i64>, i32)
+
+declare <8 x i32> @llvm.x86.avx512.gather.qpi.512 (<8 x i32>, i8*, <8 x i64>, i8, i32)
+declare void @llvm.x86.avx512.scatter.qpi.512 (i8*, i8, <8 x i64>, <8 x i32>, i32)
+declare <8 x i64> @llvm.x86.avx512.gather.qpq.512 (<8 x i64>, i8*, <8 x i64>, i8, i32)
+declare void @llvm.x86.avx512.scatter.qpq.512 (i8*, i8, <8 x i64>, <8 x i64>, i32)
+
+define void @gather_mask_dd(<16 x i32> %ind, <16 x i32> %src, i16 %mask, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: gather_mask_dd:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    kmovq %k1, %k2
+; CHECK-NEXT:    vpgatherdd (%rsi,%zmm0,4), %zmm1 {%k2}
+; CHECK-NEXT:    vpaddd {{.*}}(%rip), %zmm0, %zmm0
+; CHECK-NEXT:    vpscatterdd %zmm1, (%rdx,%zmm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = call <16 x i32> @llvm.x86.avx512.gather.dpi.512 (<16 x i32> %src, i8* %base, <16 x i32>%ind, i16 %mask, i32 4)
+  %ind2 = add <16 x i32> %ind, <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
+  call void @llvm.x86.avx512.scatter.dpi.512 (i8* %stbuf, i16 %mask, <16 x i32>%ind2, <16 x i32> %x, i32 4)
+  ret void
+}
+
+define void @gather_mask_qd(<8 x i64> %ind, <8 x i32> %src, i8 %mask, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: gather_mask_qd:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    kmovq %k1, %k2
+; CHECK-NEXT:    vpgatherqd (%rsi,%zmm0,4), %ymm1 {%k2}
+; CHECK-NEXT:    vpaddq {{.*}}(%rip), %zmm0, %zmm0
+; CHECK-NEXT:    vpscatterqd %ymm1, (%rdx,%zmm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = call <8 x i32> @llvm.x86.avx512.gather.qpi.512 (<8 x i32> %src, i8* %base, <8 x i64>%ind, i8 %mask, i32 4)
+  %ind2 = add <8 x i64> %ind, <i64 0, i64 1, i64 2, i64 3, i64 0, i64 1, i64 2, i64 3>
+  call void @llvm.x86.avx512.scatter.qpi.512 (i8* %stbuf, i8 %mask, <8 x i64>%ind2, <8 x i32> %x, i32 4)
+  ret void
+}
+
+define void @gather_mask_qq(<8 x i64> %ind, <8 x i64> %src, i8 %mask, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: gather_mask_qq:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    kmovq %k1, %k2
+; CHECK-NEXT:    vpgatherqq (%rsi,%zmm0,4), %zmm1 {%k2}
+; CHECK-NEXT:    vpaddq {{.*}}(%rip), %zmm0, %zmm0
+; CHECK-NEXT:    vpscatterqq %zmm1, (%rdx,%zmm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = call <8 x i64> @llvm.x86.avx512.gather.qpq.512 (<8 x i64> %src, i8* %base, <8 x i64>%ind, i8 %mask, i32 4)
+  %ind2 = add <8 x i64> %ind, <i64 0, i64 1, i64 2, i64 3, i64 0, i64 1, i64 2, i64 3>
+  call void @llvm.x86.avx512.scatter.qpq.512 (i8* %stbuf, i8 %mask, <8 x i64>%ind2, <8 x i64> %x, i32 4)
+  ret void
+}
+
+define void @gather_mask_dq(<8 x i32> %ind, <8 x i64> %src, i8 %mask, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: gather_mask_dq:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    kmovq %k1, %k2
+; CHECK-NEXT:    vpgatherdq (%rsi,%ymm0,4), %zmm1 {%k2}
+; CHECK-NEXT:    vpaddd {{.*}}(%rip), %ymm0, %ymm0
+; CHECK-NEXT:    vpscatterdq %zmm1, (%rdx,%ymm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = call <8 x i64> @llvm.x86.avx512.gather.dpq.512 (<8 x i64> %src, i8* %base, <8 x i32>%ind, i8 %mask, i32 4)
+  %ind2 = add <8 x i32> %ind, <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
+  call void @llvm.x86.avx512.scatter.dpq.512 (i8* %stbuf, i8 %mask, <8 x i32>%ind2, <8 x i64> %x, i32 4)
+  ret void
+}
+
+define void @gather_mask_dpd_execdomain(<8 x i32> %ind, <8 x double> %src, i8 %mask, i8* %base, <8 x double>* %stbuf)  {
+; CHECK-LABEL: gather_mask_dpd_execdomain:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    vgatherdpd (%rsi,%ymm0,4), %zmm1 {%k1}
+; CHECK-NEXT:    vmovapd %zmm1, (%rdx)
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = call <8 x double> @llvm.x86.avx512.gather.dpd.512 (<8 x double> %src, i8* %base, <8 x i32>%ind, i8 %mask, i32 4)
+  store <8 x double> %x, <8 x double>* %stbuf
+  ret void
+}
+
+define void @gather_mask_qpd_execdomain(<8 x i64> %ind, <8 x double> %src, i8 %mask, i8* %base, <8 x double>* %stbuf)  {
+; CHECK-LABEL: gather_mask_qpd_execdomain:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    vgatherqpd (%rsi,%zmm0,4), %zmm1 {%k1}
+; CHECK-NEXT:    vmovapd %zmm1, (%rdx)
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = call <8 x double> @llvm.x86.avx512.gather.qpd.512 (<8 x double> %src, i8* %base, <8 x i64>%ind, i8 %mask, i32 4)
+  store <8 x double> %x, <8 x double>* %stbuf
+  ret void
+}
+
+define <16 x float> @gather_mask_dps_execdomain(<16 x i32> %ind, <16 x float> %src, i16 %mask, i8* %base)  {
+; CHECK-LABEL: gather_mask_dps_execdomain:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    vgatherdps (%rsi,%zmm0,4), %zmm1 {%k1}
+; CHECK-NEXT:    vmovaps %zmm1, %zmm0
+; CHECK-NEXT:    retq
+  %res = call <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float> %src, i8* %base, <16 x i32>%ind, i16 %mask, i32 4)
+  ret <16 x float> %res;
+}
+
+define <8 x float> @gather_mask_qps_execdomain(<8 x i64> %ind, <8 x float> %src, i8 %mask, i8* %base)  {
+; CHECK-LABEL: gather_mask_qps_execdomain:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %edi, %k1
+; CHECK-NEXT:    vgatherqps (%rsi,%zmm0,4), %ymm1 {%k1}
+; CHECK-NEXT:    vmovaps %ymm1, %ymm0
+; CHECK-NEXT:    retq
+  %res = call <8 x float> @llvm.x86.avx512.gather.qps.512 (<8 x float> %src, i8* %base, <8 x i64>%ind, i8 %mask, i32 4)
+  ret <8 x float> %res;
+}
+
+define void @scatter_mask_dpd_execdomain(<8 x i32> %ind, <8 x double>* %src, i8 %mask, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: scatter_mask_dpd_execdomain:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vmovapd (%rdi), %zmm1
+; CHECK-NEXT:    vscatterdpd %zmm1, (%rcx,%ymm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = load <8 x double>, <8 x double>* %src, align 64
+  call void @llvm.x86.avx512.scatter.dpd.512 (i8* %stbuf, i8 %mask, <8 x i32>%ind, <8 x double> %x, i32 4)
+  ret void
+}
+
+define void @scatter_mask_qpd_execdomain(<8 x i64> %ind, <8 x double>* %src, i8 %mask, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: scatter_mask_qpd_execdomain:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vmovapd (%rdi), %zmm1
+; CHECK-NEXT:    vscatterqpd %zmm1, (%rcx,%zmm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = load <8 x double>, <8 x double>* %src, align 64
+  call void @llvm.x86.avx512.scatter.qpd.512 (i8* %stbuf, i8 %mask, <8 x i64>%ind, <8 x double> %x, i32 4)
+  ret void
+}
+
+define void @scatter_mask_dps_execdomain(<16 x i32> %ind, <16 x float>* %src, i16 %mask, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: scatter_mask_dps_execdomain:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vmovaps (%rdi), %zmm1
+; CHECK-NEXT:    vscatterdps %zmm1, (%rcx,%zmm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = load <16 x float>, <16 x float>* %src, align 64
+  call void @llvm.x86.avx512.scatter.dps.512 (i8* %stbuf, i16 %mask, <16 x i32>%ind, <16 x float> %x, i32 4)
+  ret void
+}
+
+define void @scatter_mask_qps_execdomain(<8 x i64> %ind, <8 x float>* %src, i8 %mask, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: scatter_mask_qps_execdomain:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vmovaps (%rdi), %ymm1
+; CHECK-NEXT:    vscatterqps %ymm1, (%rcx,%zmm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = load <8 x float>, <8 x float>* %src, align 32
+  call void @llvm.x86.avx512.scatter.qps.512 (i8* %stbuf, i8 %mask, <8 x i64>%ind, <8 x float> %x, i32 4)
+  ret void
+}
+
+define void @gather_qps(<8 x i64> %ind, <8 x float> %src, i8* %base, i8* %stbuf)  {
+; CHECK-LABEL: gather_qps:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    kxnorw %k0, %k0, %k2
+; CHECK-NEXT:    vgatherqps (%rdi,%zmm0,4), %ymm1 {%k2}
+; CHECK-NEXT:    vpaddq {{.*}}(%rip), %zmm0, %zmm0
+; CHECK-NEXT:    vscatterqps %ymm1, (%rsi,%zmm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %x = call <8 x float> @llvm.x86.avx512.gather.qps.512 (<8 x float> %src, i8* %base, <8 x i64>%ind, i8 -1, i32 4)
+  %ind2 = add <8 x i64> %ind, <i64 0, i64 1, i64 2, i64 3, i64 0, i64 1, i64 2, i64 3>
+  call void @llvm.x86.avx512.scatter.qps.512 (i8* %stbuf, i8 -1, <8 x i64>%ind2, <8 x float> %x, i32 4)
+  ret void
+}
+
+declare  void @llvm.x86.avx512.gatherpf.qps.512(i8, <8 x i64>, i8* , i32, i32);
+declare  void @llvm.x86.avx512.scatterpf.qps.512(i8, <8 x i64>, i8* , i32, i32);
+define void @prefetch(<8 x i64> %ind, i8* %base) {
+; CHECK-LABEL: prefetch:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vgatherpf0qps (%rdi,%zmm0,4) {%k1}
+; CHECK-NEXT:    kxorw %k0, %k0, %k1
+; CHECK-NEXT:    vgatherpf1qps (%rdi,%zmm0,4) {%k1}
+; CHECK-NEXT:    movb $1, %al
+; CHECK-NEXT:    kmovd %eax, %k1
+; CHECK-NEXT:    vscatterpf0qps (%rdi,%zmm0,2) {%k1}
+; CHECK-NEXT:    movb $120, %al
+; CHECK-NEXT:    kmovd %eax, %k1
+; CHECK-NEXT:    vscatterpf1qps (%rdi,%zmm0,2) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.gatherpf.qps.512(i8 -1, <8 x i64> %ind, i8* %base, i32 4, i32 3)
+  call void @llvm.x86.avx512.gatherpf.qps.512(i8 0, <8 x i64> %ind, i8* %base, i32 4, i32 2)
+  call void @llvm.x86.avx512.scatterpf.qps.512(i8 1, <8 x i64> %ind, i8* %base, i32 2, i32 3)
+  call void @llvm.x86.avx512.scatterpf.qps.512(i8 120, <8 x i64> %ind, i8* %base, i32 2, i32 2)
+  ret void
+}
+
+declare <2 x double> @llvm.x86.avx512.gather3div2.df(<2 x double>, i8*, <2 x i64>, i8, i32)
+
+define <2 x double>@test_int_x86_avx512_gather3div2_df(<2 x double> %x0, i8* %x1, <2 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3div2_df:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vgatherqpd (%rdi,%xmm1,4), %xmm0 {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vxorpd %xmm2, %xmm2, %xmm2
+; CHECK-NEXT:    vgatherqpd (%rdi,%xmm1,2), %xmm2 {%k1}
+; CHECK-NEXT:    vaddpd %xmm2, %xmm0, %xmm0
+; CHECK-NEXT:    retq
+  %res = call <2 x double> @llvm.x86.avx512.gather3div2.df(<2 x double> %x0, i8* %x1, <2 x i64> %x2, i8 %x3, i32 4)
+  %res1 = call <2 x double> @llvm.x86.avx512.gather3div2.df(<2 x double> %x0, i8* %x1, <2 x i64> %x2, i8 -1, i32 2)
+  %res2 = fadd <2 x double> %res, %res1
+  ret <2 x double> %res2
+}
+
+declare <2 x i64> @llvm.x86.avx512.gather3div2.di(<2 x i64>, i8*, <2 x i64>, i8, i32)
+
+define <2 x i64>@test_int_x86_avx512_gather3div2_di(<2 x i64> %x0, i8* %x1, <2 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3div2_di:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpgatherqq (%rdi,%xmm1,8), %xmm0 {%k1}
+; CHECK-NEXT:    vpaddq %xmm0, %xmm0, %xmm0
+; CHECK-NEXT:    retq
+  %res = call <2 x i64> @llvm.x86.avx512.gather3div2.di(<2 x i64> %x0, i8* %x1, <2 x i64> %x2, i8 %x3, i32 8)
+  %res1 = call <2 x i64> @llvm.x86.avx512.gather3div2.di(<2 x i64> %x0, i8* %x1, <2 x i64> %x2, i8 %x3, i32 8)
+  %res2 = add <2 x i64> %res, %res1
+  ret <2 x i64> %res2
+}
+
+declare <4 x double> @llvm.x86.avx512.gather3div4.df(<4 x double>, i8*, <4 x i64>, i8, i32)
+
+define <4 x double>@test_int_x86_avx512_gather3div4_df(<4 x double> %x0, i8* %x1, <4 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3div4_df:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vgatherqpd (%rdi,%ymm1,4), %ymm0 {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vxorpd %xmm2, %xmm2, %xmm2
+; CHECK-NEXT:    vgatherqpd (%rdi,%ymm1,2), %ymm2 {%k1}
+; CHECK-NEXT:    vaddpd %ymm2, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+  %res = call <4 x double> @llvm.x86.avx512.gather3div4.df(<4 x double> %x0, i8* %x1, <4 x i64> %x2, i8 %x3, i32 4)
+  %res1 = call <4 x double> @llvm.x86.avx512.gather3div4.df(<4 x double> %x0, i8* %x1, <4 x i64> %x2, i8 -1, i32 2)
+  %res2 = fadd <4 x double> %res, %res1
+  ret <4 x double> %res2
+}
+
+declare <4 x i64> @llvm.x86.avx512.gather3div4.di(<4 x i64>, i8*, <4 x i64>, i8, i32)
+
+define <4 x i64>@test_int_x86_avx512_gather3div4_di(<4 x i64> %x0, i8* %x1, <4 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3div4_di:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpgatherqq (%rdi,%ymm1,8), %ymm0 {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vpxor %xmm2, %xmm2, %xmm2
+; CHECK-NEXT:    vpgatherqq (%rdi,%ymm1,8), %ymm2 {%k1}
+; CHECK-NEXT:    vpaddq %ymm2, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+  %res = call <4 x i64> @llvm.x86.avx512.gather3div4.di(<4 x i64> %x0, i8* %x1, <4 x i64> %x2, i8 %x3, i32 8)
+  %res1 = call <4 x i64> @llvm.x86.avx512.gather3div4.di(<4 x i64> %x0, i8* %x1, <4 x i64> %x2, i8 -1, i32 8)
+  %res2 = add <4 x i64> %res, %res1
+  ret <4 x i64> %res2
+}
+
+declare <4 x float> @llvm.x86.avx512.gather3div4.sf(<4 x float>, i8*, <2 x i64>, i8, i32)
+
+define <4 x float>@test_int_x86_avx512_gather3div4_sf(<4 x float> %x0, i8* %x1, <2 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3div4_sf:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vgatherqps (%rdi,%xmm1,4), %xmm0 {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vxorps %xmm2, %xmm2, %xmm2
+; CHECK-NEXT:    vgatherqps (%rdi,%xmm1,2), %xmm2 {%k1}
+; CHECK-NEXT:    vaddps %xmm2, %xmm0, %xmm0
+; CHECK-NEXT:    retq
+  %res = call <4 x float> @llvm.x86.avx512.gather3div4.sf(<4 x float> %x0, i8* %x1, <2 x i64> %x2, i8 %x3, i32 4)
+  %res1 = call <4 x float> @llvm.x86.avx512.gather3div4.sf(<4 x float> %x0, i8* %x1, <2 x i64> %x2, i8 -1, i32 2)
+  %res2 = fadd <4 x float> %res, %res1
+  ret <4 x float> %res2
+}
+
+declare <4 x i32> @llvm.x86.avx512.gather3div4.si(<4 x i32>, i8*, <2 x i64>, i8, i32)
+
+define <4 x i32>@test_int_x86_avx512_gather3div4_si(<4 x i32> %x0, i8* %x1, <2 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3div4_si:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    kxnorw %k0, %k0, %k2
+; CHECK-NEXT:    vpxor %xmm2, %xmm2, %xmm2
+; CHECK-NEXT:    vpgatherqd (%rdi,%xmm1,4), %xmm2 {%k2}
+; CHECK-NEXT:    vpgatherqd (%rdi,%xmm1,4), %xmm0 {%k1}
+; CHECK-NEXT:    vpaddd %xmm0, %xmm2, %xmm0
+; CHECK-NEXT:    retq
+  %res = call <4 x i32> @llvm.x86.avx512.gather3div4.si(<4 x i32> %x0, i8* %x1, <2 x i64> %x2, i8 -1, i32 4)
+  %res1 = call <4 x i32> @llvm.x86.avx512.gather3div4.si(<4 x i32> %x0, i8* %x1, <2 x i64> %x2, i8 %x3, i32 4)
+  %res2 = add <4 x i32> %res, %res1
+  ret <4 x i32> %res2
+}
+
+declare <4 x float> @llvm.x86.avx512.gather3div8.sf(<4 x float>, i8*, <4 x i64>, i8, i32)
+
+define <4 x float>@test_int_x86_avx512_gather3div8_sf(<4 x float> %x0, i8* %x1, <4 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3div8_sf:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vgatherqps (%rdi,%ymm1,4), %xmm0 {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vxorps %xmm2, %xmm2, %xmm2
+; CHECK-NEXT:    vgatherqps (%rdi,%ymm1,2), %xmm2 {%k1}
+; CHECK-NEXT:    vaddps %xmm2, %xmm0, %xmm0
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %res = call <4 x float> @llvm.x86.avx512.gather3div8.sf(<4 x float> %x0, i8* %x1, <4 x i64> %x2, i8 %x3, i32 4)
+  %res1 = call <4 x float> @llvm.x86.avx512.gather3div8.sf(<4 x float> %x0, i8* %x1, <4 x i64> %x2, i8 -1, i32 2)
+  %res2 = fadd <4 x float> %res, %res1
+  ret <4 x float> %res2
+}
+
+declare <4 x i32> @llvm.x86.avx512.gather3div8.si(<4 x i32>, i8*, <4 x i64>, i8, i32)
+
+define <4 x i32>@test_int_x86_avx512_gather3div8_si(<4 x i32> %x0, i8* %x1, <4 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3div8_si:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vmovdqa %xmm0, %xmm2
+; CHECK-NEXT:    kmovq %k1, %k2
+; CHECK-NEXT:    vpgatherqd (%rdi,%ymm1,4), %xmm2 {%k2}
+; CHECK-NEXT:    vpgatherqd (%rdi,%ymm1,2), %xmm0 {%k1}
+; CHECK-NEXT:    vpaddd %xmm0, %xmm2, %xmm0
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  %res = call <4 x i32> @llvm.x86.avx512.gather3div8.si(<4 x i32> %x0, i8* %x1, <4 x i64> %x2, i8 %x3, i32 4)
+  %res1 = call <4 x i32> @llvm.x86.avx512.gather3div8.si(<4 x i32> %x0, i8* %x1, <4 x i64> %x2, i8 %x3, i32 2)
+  %res2 = add <4 x i32> %res, %res1
+  ret <4 x i32> %res2
+}
+
+declare <2 x double> @llvm.x86.avx512.gather3siv2.df(<2 x double>, i8*, <4 x i32>, i8, i32)
+
+define <2 x double>@test_int_x86_avx512_gather3siv2_df(<2 x double> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3siv2_df:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vgatherdpd (%rdi,%xmm1,4), %xmm0 {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vxorpd %xmm2, %xmm2, %xmm2
+; CHECK-NEXT:    vgatherdpd (%rdi,%xmm1,2), %xmm2 {%k1}
+; CHECK-NEXT:    vaddpd %xmm2, %xmm0, %xmm0
+; CHECK-NEXT:    retq
+  %res = call <2 x double> @llvm.x86.avx512.gather3siv2.df(<2 x double> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 4)
+  %res1 = call <2 x double> @llvm.x86.avx512.gather3siv2.df(<2 x double> %x0, i8* %x1, <4 x i32> %x2, i8 -1, i32 2)
+  %res2 = fadd <2 x double> %res, %res1
+  ret <2 x double> %res2
+}
+
+declare <2 x i64> @llvm.x86.avx512.gather3siv2.di(<2 x i64>, i8*, <4 x i32>, i8, i32)
+
+define <2 x i64>@test_int_x86_avx512_gather3siv2_di(<2 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3siv2_di:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpgatherdq (%rdi,%xmm1,8), %xmm0 {%k1}
+; CHECK-NEXT:    vpaddq %xmm0, %xmm0, %xmm0
+; CHECK-NEXT:    retq
+  %res = call <2 x i64> @llvm.x86.avx512.gather3siv2.di(<2 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 8)
+  %res1 = call <2 x i64> @llvm.x86.avx512.gather3siv2.di(<2 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 8)
+  %res2 = add <2 x i64> %res, %res1
+  ret <2 x i64> %res2
+}
+
+declare <4 x double> @llvm.x86.avx512.gather3siv4.df(<4 x double>, i8*, <4 x i32>, i8, i32)
+
+define <4 x double>@test_int_x86_avx512_gather3siv4_df(<4 x double> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3siv4_df:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vgatherdpd (%rdi,%xmm1,4), %ymm0 {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vxorpd %xmm2, %xmm2, %xmm2
+; CHECK-NEXT:    vgatherdpd (%rdi,%xmm1,2), %ymm2 {%k1}
+; CHECK-NEXT:    vaddpd %ymm2, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+  %res = call <4 x double> @llvm.x86.avx512.gather3siv4.df(<4 x double> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 4)
+  %res1 = call <4 x double> @llvm.x86.avx512.gather3siv4.df(<4 x double> %x0, i8* %x1, <4 x i32> %x2, i8 -1, i32 2)
+  %res2 = fadd <4 x double> %res, %res1
+  ret <4 x double> %res2
+}
+
+declare <4 x i64> @llvm.x86.avx512.gather3siv4.di(<4 x i64>, i8*, <4 x i32>, i8, i32)
+
+define <4 x i64>@test_int_x86_avx512_gather3siv4_di(<4 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3siv4_di:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpgatherdq (%rdi,%xmm1,8), %ymm0 {%k1}
+; CHECK-NEXT:    vpaddq %ymm0, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+  %res = call <4 x i64> @llvm.x86.avx512.gather3siv4.di(<4 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 8)
+  %res1 = call <4 x i64> @llvm.x86.avx512.gather3siv4.di(<4 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 8)
+  %res2 = add <4 x i64> %res, %res1
+  ret <4 x i64> %res2
+}
+
+declare <4 x float> @llvm.x86.avx512.gather3siv4.sf(<4 x float>, i8*, <4 x i32>, i8, i32)
+
+define <4 x float>@test_int_x86_avx512_gather3siv4_sf(<4 x float> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3siv4_sf:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vgatherdps (%rdi,%xmm1,4), %xmm0 {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vxorps %xmm2, %xmm2, %xmm2
+; CHECK-NEXT:    vgatherdps (%rdi,%xmm1,2), %xmm2 {%k1}
+; CHECK-NEXT:    vaddps %xmm2, %xmm0, %xmm0
+; CHECK-NEXT:    retq
+  %res = call <4 x float> @llvm.x86.avx512.gather3siv4.sf(<4 x float> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 4)
+  %res1 = call <4 x float> @llvm.x86.avx512.gather3siv4.sf(<4 x float> %x0, i8* %x1, <4 x i32> %x2, i8 -1, i32 2)
+  %res2 = fadd <4 x float> %res, %res1
+  ret <4 x float> %res2
+}
+
+declare <4 x i32> @llvm.x86.avx512.gather3siv4.si(<4 x i32>, i8*, <4 x i32>, i8, i32)
+
+define <4 x i32>@test_int_x86_avx512_gather3siv4_si(<4 x i32> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3siv4_si:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    kxnorw %k0, %k0, %k2
+; CHECK-NEXT:    vpxor %xmm2, %xmm2, %xmm2
+; CHECK-NEXT:    vpgatherdd (%rdi,%xmm1,4), %xmm2 {%k2}
+; CHECK-NEXT:    vpgatherdd (%rdi,%xmm1,2), %xmm0 {%k1}
+; CHECK-NEXT:    vpaddd %xmm0, %xmm2, %xmm0
+; CHECK-NEXT:    retq
+  %res = call <4 x i32> @llvm.x86.avx512.gather3siv4.si(<4 x i32> %x0, i8* %x1, <4 x i32> %x2, i8 -1, i32 4)
+  %res1 = call <4 x i32> @llvm.x86.avx512.gather3siv4.si(<4 x i32> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 2)
+  %res2 = add <4 x i32> %res, %res1
+  ret <4 x i32> %res2
+}
+
+declare <8 x float> @llvm.x86.avx512.gather3siv8.sf(<8 x float>, i8*, <8 x i32>, i8, i32)
+
+define <8 x float>@test_int_x86_avx512_gather3siv8_sf(<8 x float> %x0, i8* %x1, <8 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3siv8_sf:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vgatherdps (%rdi,%ymm1,4), %ymm0 {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vxorps %xmm2, %xmm2, %xmm2
+; CHECK-NEXT:    vgatherdps (%rdi,%ymm1,2), %ymm2 {%k1}
+; CHECK-NEXT:    vaddps %ymm2, %ymm0, %ymm0
+; CHECK-NEXT:    retq
+  %res = call <8 x float> @llvm.x86.avx512.gather3siv8.sf(<8 x float> %x0, i8* %x1, <8 x i32> %x2, i8 %x3, i32 4)
+  %res1 = call <8 x float> @llvm.x86.avx512.gather3siv8.sf(<8 x float> %x0, i8* %x1, <8 x i32> %x2, i8 -1, i32 2)
+  %res2 = fadd <8 x float> %res, %res1
+  ret <8 x float> %res2
+}
+
+declare <8 x i32> @llvm.x86.avx512.gather3siv8.si(<8 x i32>, i8*, <8 x i32>, i8, i32)
+
+define <8 x i32>@test_int_x86_avx512_gather3siv8_si(<8 x i32> %x0, i8* %x1, <8 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_gather3siv8_si:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vmovdqa %ymm0, %ymm2
+; CHECK-NEXT:    kmovq %k1, %k2
+; CHECK-NEXT:    vpgatherdd (%rdi,%ymm1,4), %ymm2 {%k2}
+; CHECK-NEXT:    vpgatherdd (%rdi,%ymm1,2), %ymm0 {%k1}
+; CHECK-NEXT:    vpaddd %ymm0, %ymm2, %ymm0
+; CHECK-NEXT:    retq
+  %res = call <8 x i32> @llvm.x86.avx512.gather3siv8.si(<8 x i32> %x0, i8* %x1, <8 x i32> %x2, i8 %x3, i32 4)
+  %res1 = call <8 x i32> @llvm.x86.avx512.gather3siv8.si(<8 x i32> %x0, i8* %x1, <8 x i32> %x2, i8 %x3, i32 2)
+  %res2 = add <8 x i32> %res, %res1
+  ret <8 x i32> %res2
+}
+
+declare void @llvm.x86.avx512.scatterdiv2.df(i8*, i8, <2 x i64>, <2 x double>, i32)
+
+define void@test_int_x86_avx512_scatterdiv2_df(i8* %x0, i8 %x1, <2 x i64> %x2, <2 x double> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scatterdiv2_df:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    kxnorw %k0, %k0, %k2
+; CHECK-NEXT:    vscatterqpd %xmm1, (%rdi,%xmm0,2) {%k2}
+; CHECK-NEXT:    vscatterqpd %xmm1, (%rdi,%xmm0,4) {%k1}
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scatterdiv2.df(i8* %x0, i8 -1, <2 x i64> %x2, <2 x double> %x3, i32 2)
+  call void @llvm.x86.avx512.scatterdiv2.df(i8* %x0, i8 %x1, <2 x i64> %x2, <2 x double> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scatterdiv2.di(i8*, i8, <2 x i64>, <2 x i64>, i32)
+
+define void@test_int_x86_avx512_scatterdiv2_di(i8* %x0, i8 %x1, <2 x i64> %x2, <2 x i64> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scatterdiv2_di:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpscatterqq %xmm1, (%rdi,%xmm0,2) {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vpscatterqq %xmm1, (%rdi,%xmm0,4) {%k1}
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scatterdiv2.di(i8* %x0, i8 %x1, <2 x i64> %x2, <2 x i64> %x3, i32 2)
+  call void @llvm.x86.avx512.scatterdiv2.di(i8* %x0, i8 -1, <2 x i64> %x2, <2 x i64> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scatterdiv4.df(i8*, i8, <4 x i64>, <4 x double>, i32)
+
+define void@test_int_x86_avx512_scatterdiv4_df(i8* %x0, i8 %x1, <4 x i64> %x2, <4 x double> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scatterdiv4_df:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vscatterqpd %ymm1, (%rdi,%ymm0,2) {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vscatterqpd %ymm1, (%rdi,%ymm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scatterdiv4.df(i8* %x0, i8 %x1, <4 x i64> %x2, <4 x double> %x3, i32 2)
+  call void @llvm.x86.avx512.scatterdiv4.df(i8* %x0, i8 -1, <4 x i64> %x2, <4 x double> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scatterdiv4.di(i8*, i8, <4 x i64>, <4 x i64>, i32)
+
+define void@test_int_x86_avx512_scatterdiv4_di(i8* %x0, i8 %x1, <4 x i64> %x2, <4 x i64> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scatterdiv4_di:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpscatterqq %ymm1, (%rdi,%ymm0,2) {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vpscatterqq %ymm1, (%rdi,%ymm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scatterdiv4.di(i8* %x0, i8 %x1, <4 x i64> %x2, <4 x i64> %x3, i32 2)
+  call void @llvm.x86.avx512.scatterdiv4.di(i8* %x0, i8 -1, <4 x i64> %x2, <4 x i64> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scatterdiv4.sf(i8*, i8, <2 x i64>, <4 x float>, i32)
+
+define void@test_int_x86_avx512_scatterdiv4_sf(i8* %x0, i8 %x1, <2 x i64> %x2, <4 x float> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scatterdiv4_sf:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vscatterqps %xmm1, (%rdi,%xmm0,2) {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vscatterqps %xmm1, (%rdi,%xmm0,4) {%k1}
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scatterdiv4.sf(i8* %x0, i8 %x1, <2 x i64> %x2, <4 x float> %x3, i32 2)
+  call void @llvm.x86.avx512.scatterdiv4.sf(i8* %x0, i8 -1, <2 x i64> %x2, <4 x float> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scatterdiv4.si(i8*, i8, <2 x i64>, <4 x i32>, i32)
+
+define void@test_int_x86_avx512_scatterdiv4_si(i8* %x0, i8 %x1, <2 x i64> %x2, <4 x i32> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scatterdiv4_si:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    kxnorw %k0, %k0, %k2
+; CHECK-NEXT:    vpscatterqd %xmm1, (%rdi,%xmm0,2) {%k2}
+; CHECK-NEXT:    vpscatterqd %xmm1, (%rdi,%xmm0,4) {%k1}
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scatterdiv4.si(i8* %x0, i8 -1, <2 x i64> %x2, <4 x i32> %x3, i32 2)
+  call void @llvm.x86.avx512.scatterdiv4.si(i8* %x0, i8 %x1, <2 x i64> %x2, <4 x i32> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scatterdiv8.sf(i8*, i8, <4 x i64>, <4 x float>, i32)
+
+define void@test_int_x86_avx512_scatterdiv8_sf(i8* %x0, i8 %x1, <4 x i64> %x2, <4 x float> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scatterdiv8_sf:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vscatterqps %xmm1, (%rdi,%ymm0,2) {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vscatterqps %xmm1, (%rdi,%ymm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scatterdiv8.sf(i8* %x0, i8 %x1, <4 x i64> %x2, <4 x float> %x3, i32 2)
+  call void @llvm.x86.avx512.scatterdiv8.sf(i8* %x0, i8 -1, <4 x i64> %x2, <4 x float> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scatterdiv8.si(i8*, i8, <4 x i64>, <4 x i32>, i32)
+
+define void@test_int_x86_avx512_scatterdiv8_si(i8* %x0, i8 %x1, <4 x i64> %x2, <4 x i32> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scatterdiv8_si:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpscatterqd %xmm1, (%rdi,%ymm0,2) {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vpscatterqd %xmm1, (%rdi,%ymm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scatterdiv8.si(i8* %x0, i8 %x1, <4 x i64> %x2, <4 x i32> %x3, i32 2)
+  call void @llvm.x86.avx512.scatterdiv8.si(i8* %x0, i8 -1, <4 x i64> %x2, <4 x i32> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scattersiv2.df(i8*, i8, <4 x i32>, <2 x double>, i32)
+
+define void@test_int_x86_avx512_scattersiv2_df(i8* %x0, i8 %x1, <4 x i32> %x2, <2 x double> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scattersiv2_df:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    kxnorw %k0, %k0, %k2
+; CHECK-NEXT:    vscatterdpd %xmm1, (%rdi,%xmm0,2) {%k2}
+; CHECK-NEXT:    vscatterdpd %xmm1, (%rdi,%xmm0,4) {%k1}
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scattersiv2.df(i8* %x0, i8 -1, <4 x i32> %x2, <2 x double> %x3, i32 2)
+  call void @llvm.x86.avx512.scattersiv2.df(i8* %x0, i8 %x1, <4 x i32> %x2, <2 x double> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scattersiv2.di(i8*, i8, <4 x i32>, <2 x i64>, i32)
+
+define void@test_int_x86_avx512_scattersiv2_di(i8* %x0, i8 %x1, <4 x i32> %x2, <2 x i64> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scattersiv2_di:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    kxnorw %k0, %k0, %k2
+; CHECK-NEXT:    vpscatterdq %xmm1, (%rdi,%xmm0,2) {%k2}
+; CHECK-NEXT:    vpscatterdq %xmm1, (%rdi,%xmm0,4) {%k1}
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scattersiv2.di(i8* %x0, i8 -1, <4 x i32> %x2, <2 x i64> %x3, i32 2)
+  call void @llvm.x86.avx512.scattersiv2.di(i8* %x0, i8 %x1, <4 x i32> %x2, <2 x i64> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scattersiv4.df(i8*, i8, <4 x i32>, <4 x double>, i32)
+
+define void@test_int_x86_avx512_scattersiv4_df(i8* %x0, i8 %x1, <4 x i32> %x2, <4 x double> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scattersiv4_df:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vscatterdpd %ymm1, (%rdi,%xmm0,2) {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vscatterdpd %ymm1, (%rdi,%xmm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scattersiv4.df(i8* %x0, i8 %x1, <4 x i32> %x2, <4 x double> %x3, i32 2)
+  call void @llvm.x86.avx512.scattersiv4.df(i8* %x0, i8 -1, <4 x i32> %x2, <4 x double> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scattersiv4.di(i8*, i8, <4 x i32>, <4 x i64>, i32)
+
+define void@test_int_x86_avx512_scattersiv4_di(i8* %x0, i8 %x1, <4 x i32> %x2, <4 x i64> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scattersiv4_di:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    kxnorw %k0, %k0, %k2
+; CHECK-NEXT:    vpscatterdq %ymm1, (%rdi,%xmm0,2) {%k2}
+; CHECK-NEXT:    vpscatterdq %ymm1, (%rdi,%xmm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scattersiv4.di(i8* %x0, i8 -1, <4 x i32> %x2, <4 x i64> %x3, i32 2)
+  call void @llvm.x86.avx512.scattersiv4.di(i8* %x0, i8 %x1, <4 x i32> %x2, <4 x i64> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scattersiv4.sf(i8*, i8, <4 x i32>, <4 x float>, i32)
+
+define void@test_int_x86_avx512_scattersiv4_sf(i8* %x0, i8 %x1, <4 x i32> %x2, <4 x float> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scattersiv4_sf:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vscatterdps %xmm1, (%rdi,%xmm0,2) {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vscatterdps %xmm1, (%rdi,%xmm0,4) {%k1}
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scattersiv4.sf(i8* %x0, i8 %x1, <4 x i32> %x2, <4 x float> %x3, i32 2)
+  call void @llvm.x86.avx512.scattersiv4.sf(i8* %x0, i8 -1, <4 x i32> %x2, <4 x float> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scattersiv4.si(i8*, i8, <4 x i32>, <4 x i32>, i32)
+
+define void@test_int_x86_avx512_scattersiv4_si(i8* %x0, i8 %x1, <4 x i32> %x2, <4 x i32> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scattersiv4_si:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpscatterdd %xmm1, (%rdi,%xmm0,2) {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vpscatterdd %xmm1, (%rdi,%xmm0,4) {%k1}
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scattersiv4.si(i8* %x0, i8 %x1, <4 x i32> %x2, <4 x i32> %x3, i32 2)
+  call void @llvm.x86.avx512.scattersiv4.si(i8* %x0, i8 -1, <4 x i32> %x2, <4 x i32> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scattersiv8.sf(i8*, i8, <8 x i32>, <8 x float>, i32)
+
+define void@test_int_x86_avx512_scattersiv8_sf(i8* %x0, i8 %x1, <8 x i32> %x2, <8 x float> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scattersiv8_sf:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vscatterdps %ymm1, (%rdi,%ymm0,2) {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vscatterdps %ymm1, (%rdi,%ymm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scattersiv8.sf(i8* %x0, i8 %x1, <8 x i32> %x2, <8 x float> %x3, i32 2)
+  call void @llvm.x86.avx512.scattersiv8.sf(i8* %x0, i8 -1, <8 x i32> %x2, <8 x float> %x3, i32 4)
+  ret void
+}
+
+declare void @llvm.x86.avx512.scattersiv8.si(i8*, i8, <8 x i32>, <8 x i32>, i32)
+
+define void@test_int_x86_avx512_scattersiv8_si(i8* %x0, i8 %x1, <8 x i32> %x2, <8 x i32> %x3) {
+; CHECK-LABEL: test_int_x86_avx512_scattersiv8_si:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kmovd %esi, %k1
+; CHECK-NEXT:    vpscatterdd %ymm1, (%rdi,%ymm0,2) {%k1}
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vpscatterdd %ymm1, (%rdi,%ymm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scattersiv8.si(i8* %x0, i8 %x1, <8 x i32> %x2, <8 x i32> %x3, i32 2)
+  call void @llvm.x86.avx512.scattersiv8.si(i8* %x0, i8 -1, <8 x i32> %x2, <8 x i32> %x3, i32 4)
+  ret void
+}
+
+define void @scatter_mask_test(i8* %x0, <8 x i32> %x2, <8 x i32> %x3) {
+; CHECK-LABEL: scatter_mask_test:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vpscatterdd %ymm1, (%rdi,%ymm0,2) {%k1}
+; CHECK-NEXT:    kxorw %k0, %k0, %k1
+; CHECK-NEXT:    vpscatterdd %ymm1, (%rdi,%ymm0,4) {%k1}
+; CHECK-NEXT:    movb $1, %al
+; CHECK-NEXT:    kmovd %eax, %k1
+; CHECK-NEXT:    vpscatterdd %ymm1, (%rdi,%ymm0,2) {%k1}
+; CHECK-NEXT:    movb $96, %al
+; CHECK-NEXT:    kmovd %eax, %k1
+; CHECK-NEXT:    vpscatterdd %ymm1, (%rdi,%ymm0,4) {%k1}
+; CHECK-NEXT:    vzeroupper
+; CHECK-NEXT:    retq
+  call void @llvm.x86.avx512.scattersiv8.si(i8* %x0, i8 -1, <8 x i32> %x2, <8 x i32> %x3, i32 2)
+  call void @llvm.x86.avx512.scattersiv8.si(i8* %x0, i8 0, <8 x i32> %x2, <8 x i32> %x3, i32 4)
+  call void @llvm.x86.avx512.scattersiv8.si(i8* %x0, i8 1, <8 x i32> %x2, <8 x i32> %x3, i32 2)
+  call void @llvm.x86.avx512.scattersiv8.si(i8* %x0, i8 96, <8 x i32> %x2, <8 x i32> %x3, i32 4)
+  ret void
+}
+
+define <16 x float> @gather_mask_test(<16 x i32> %ind, <16 x float> %src, i8* %base)  {
+; CHECK-LABEL: gather_mask_test:
+; CHECK:       ## %bb.0:
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vxorps %xmm2, %xmm2, %xmm2
+; CHECK-NEXT:    vgatherdps (%rdi,%zmm0,4), %zmm2 {%k1}
+; CHECK-NEXT:    kxorw %k0, %k0, %k1
+; CHECK-NEXT:    vmovaps %zmm1, %zmm3
+; CHECK-NEXT:    vgatherdps (%rdi,%zmm0,4), %zmm3 {%k1}
+; CHECK-NEXT:    vaddps %zmm3, %zmm2, %zmm2
+; CHECK-NEXT:    movw $1, %ax
+; CHECK-NEXT:    kmovd %eax, %k1
+; CHECK-NEXT:    vmovaps %zmm1, %zmm3
+; CHECK-NEXT:    vgatherdps (%rdi,%zmm0,4), %zmm3 {%k1}
+; CHECK-NEXT:    movw $220, %ax
+; CHECK-NEXT:    kmovd %eax, %k1
+; CHECK-NEXT:    vgatherdps (%rdi,%zmm0,4), %zmm1 {%k1}
+; CHECK-NEXT:    vaddps %zmm3, %zmm1, %zmm0
+; CHECK-NEXT:    vaddps %zmm2, %zmm0, %zmm0
+; CHECK-NEXT:    retq
+  %res = call <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float> %src, i8* %base, <16 x i32>%ind, i16 -1, i32 4)
+  %res1 = call <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float> %src, i8* %base, <16 x i32>%ind, i16 0, i32 4)
+  %res2 = call <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float> %src, i8* %base, <16 x i32>%ind, i16 1, i32 4)
+  %res3 = call <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float> %src, i8* %base, <16 x i32>%ind, i16 220, i32 4)
+
+  %res4 = fadd <16 x float> %res, %res1
+  %res5 = fadd <16 x float> %res3, %res2
+  %res6 = fadd <16 x float> %res5, %res4
+  ret <16 x float> %res6
+}

llvm/trunk/test/CodeGen/X86/avx512-gather-scatter-intrin.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mattr=+avx512f,+avx512bw,+avx512vl,+avx512dq | FileCheck %s
 
-declare <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float>, i8*, <16 x i32>, i16, i32)
 declare void @llvm.x86.avx512.scatter.dps.512 (i8*, i16, <16 x i32>, <16 x float>, i32)
-declare <8 x double> @llvm.x86.avx512.gather.dpd.512 (<8 x double>, i8*, <8 x i32>, i8, i32)
 declare void @llvm.x86.avx512.scatter.dpd.512 (i8*, i8, <8 x i32>, <8 x double>, i32)
 
-declare <8 x float> @llvm.x86.avx512.gather.qps.512 (<8 x float>, i8*, <8 x i64>, i8, i32)
 declare void @llvm.x86.avx512.scatter.qps.512 (i8*, i8, <8 x i64>, <8 x float>, i32)
-declare <8 x double> @llvm.x86.avx512.gather.qpd.512 (<8 x double>, i8*, <8 x i64>, i8, i32)
 declare void @llvm.x86.avx512.scatter.qpd.512 (i8*, i8, <8 x i64>, <8 x double>, i32)
 
-define void @gather_mask_dps(<16 x i32> %ind, <16 x float> %src, i16 %mask, i8* %base, i8* %stbuf)  {
+define void @gather_mask_dps(<16 x i32> %ind, <16 x float> %src, i16 %mask, i8* %base, i8* %stbuf) {
 ; CHECK-LABEL: gather_mask_dps:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %edi, %k1
 ; CHECK-NEXT:    kmovq %k1, %k2
 ; CHECK-NEXT:    vgatherdps (%rsi,%zmm0,4), %zmm1 {%k2}
 ; CHECK-NEXT:    vpaddd {{.*}}(%rip), %zmm0, %zmm0
 ; CHECK-NEXT:    vscatterdps %zmm1, (%rdx,%zmm0,4) {%k1}
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %x = call <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float> %src, i8* %base, <16 x i32>%ind, i16 %mask, i32 4)
+  %1 = bitcast i16 %mask to <16 x i1>
+  %x = call <16 x float> @llvm.x86.avx512.mask.gather.dps.512(<16 x float> %src, i8* %base, <16 x i32> %ind, <16 x i1> %1, i32 4)
   %ind2 = add <16 x i32> %ind, <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
   call void @llvm.x86.avx512.scatter.dps.512 (i8* %stbuf, i16 %mask, <16 x i32>%ind2, <16 x float> %x, i32 4)
   ret void
 }
 
-define void @gather_mask_dpd(<8 x i32> %ind, <8 x double> %src, i8 %mask, i8* %base, i8* %stbuf)  {
+define void @gather_mask_dpd(<8 x i32> %ind, <8 x double> %src, i8 %mask, i8* %base, i8* %stbuf) {
 ; CHECK-LABEL: gather_mask_dpd:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %edi, %k1
 ; CHECK-NEXT:    kmovq %k1, %k2
 ; CHECK-NEXT:    vgatherdpd (%rsi,%ymm0,4), %zmm1 {%k2}
 ; CHECK-NEXT:    vpaddd {{.*}}(%rip), %ymm0, %ymm0
 ; CHECK-NEXT:    vscatterdpd %zmm1, (%rdx,%ymm0,4) {%k1}
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %x = call <8 x double> @llvm.x86.avx512.gather.dpd.512 (<8 x double> %src, i8* %base, <8 x i32>%ind, i8 %mask, i32 4)
+  %1 = bitcast i8 %mask to <8 x i1>
+  %x = call <8 x double> @llvm.x86.avx512.mask.gather.dpd.512(<8 x double> %src, i8* %base, <8 x i32> %ind, <8 x i1> %1, i32 4)
   %ind2 = add <8 x i32> %ind, <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
   call void @llvm.x86.avx512.scatter.dpd.512 (i8* %stbuf, i8 %mask, <8 x i32>%ind2, <8 x double> %x, i32 4)
   ret void
 }
 
-define void @gather_mask_qps(<8 x i64> %ind, <8 x float> %src, i8 %mask, i8* %base, i8* %stbuf)  {
+define void @gather_mask_qps(<8 x i64> %ind, <8 x float> %src, i8 %mask, i8* %base, i8* %stbuf) {
 ; CHECK-LABEL: gather_mask_qps:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %edi, %k1
 ; CHECK-NEXT:    kmovq %k1, %k2
 ; CHECK-NEXT:    vgatherqps (%rsi,%zmm0,4), %ymm1 {%k2}
 ; CHECK-NEXT:    vpaddq {{.*}}(%rip), %zmm0, %zmm0
 ; CHECK-NEXT:    vscatterqps %ymm1, (%rdx,%zmm0,4) {%k1}
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %x = call <8 x float> @llvm.x86.avx512.gather.qps.512 (<8 x float> %src, i8* %base, <8 x i64>%ind, i8 %mask, i32 4)
+  %1 = bitcast i8 %mask to <8 x i1>
+  %x = call <8 x float> @llvm.x86.avx512.mask.gather.qps.512(<8 x float> %src, i8* %base, <8 x i64> %ind, <8 x i1> %1, i32 4)
   %ind2 = add <8 x i64> %ind, <i64 0, i64 1, i64 2, i64 3, i64 0, i64 1, i64 2, i64 3>
   call void @llvm.x86.avx512.scatter.qps.512 (i8* %stbuf, i8 %mask, <8 x i64>%ind2, <8 x float> %x, i32 4)
   ret void
 }
 
-define void @gather_mask_qpd(<8 x i64> %ind, <8 x double> %src, i8 %mask, i8* %base, i8* %stbuf)  {
+define void @gather_mask_qpd(<8 x i64> %ind, <8 x double> %src, i8 %mask, i8* %base, i8* %stbuf) {
 ; CHECK-LABEL: gather_mask_qpd:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %edi, %k1
 ; CHECK-NEXT:    kmovq %k1, %k2
 ; CHECK-NEXT:    vgatherqpd (%rsi,%zmm0,4), %zmm1 {%k2}
 ; CHECK-NEXT:    vpaddq {{.*}}(%rip), %zmm0, %zmm0
 ; CHECK-NEXT:    vscatterqpd %zmm1, (%rdx,%zmm0,4) {%k1}
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %x = call <8 x double> @llvm.x86.avx512.gather.qpd.512 (<8 x double> %src, i8* %base, <8 x i64>%ind, i8 %mask, i32 4)
+  %1 = bitcast i8 %mask to <8 x i1>
+  %x = call <8 x double> @llvm.x86.avx512.mask.gather.qpd.512(<8 x double> %src, i8* %base, <8 x i64> %ind, <8 x i1> %1, i32 4)
   %ind2 = add <8 x i64> %ind, <i64 0, i64 1, i64 2, i64 3, i64 0, i64 1, i64 2, i64 3>
   call void @llvm.x86.avx512.scatter.qpd.512 (i8* %stbuf, i8 %mask, <8 x i64>%ind2, <8 x double> %x, i32 4)
   ret void
 }
 ;;
 ;; Integer Gather/Scatter
 ;;
-declare <16 x i32> @llvm.x86.avx512.gather.dpi.512 (<16 x i32>, i8*, <16 x i32>, i16, i32)
 declare void @llvm.x86.avx512.scatter.dpi.512 (i8*, i16, <16 x i32>, <16 x i32>, i32)
-declare <8 x i64> @llvm.x86.avx512.gather.dpq.512 (<8 x i64>, i8*, <8 x i32>, i8, i32)
 declare void @llvm.x86.avx512.scatter.dpq.512 (i8*, i8, <8 x i32>, <8 x i64>, i32)
 
-declare <8 x i32> @llvm.x86.avx512.gather.qpi.512 (<8 x i32>, i8*, <8 x i64>, i8, i32)
 declare void @llvm.x86.avx512.scatter.qpi.512 (i8*, i8, <8 x i64>, <8 x i32>, i32)
-declare <8 x i64> @llvm.x86.avx512.gather.qpq.512 (<8 x i64>, i8*, <8 x i64>, i8, i32)
 declare void @llvm.x86.avx512.scatter.qpq.512 (i8*, i8, <8 x i64>, <8 x i64>, i32)
 
-define void @gather_mask_dd(<16 x i32> %ind, <16 x i32> %src, i16 %mask, i8* %base, i8* %stbuf)  {
+define void @gather_mask_dd(<16 x i32> %ind, <16 x i32> %src, i16 %mask, i8* %base, i8* %stbuf) {
 ; CHECK-LABEL: gather_mask_dd:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %edi, %k1
 ; CHECK-NEXT:    kmovq %k1, %k2
 ; CHECK-NEXT:    vpgatherdd (%rsi,%zmm0,4), %zmm1 {%k2}
 ; CHECK-NEXT:    vpaddd {{.*}}(%rip), %zmm0, %zmm0
 ; CHECK-NEXT:    vpscatterdd %zmm1, (%rdx,%zmm0,4) {%k1}
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %x = call <16 x i32> @llvm.x86.avx512.gather.dpi.512 (<16 x i32> %src, i8* %base, <16 x i32>%ind, i16 %mask, i32 4)
+  %1 = bitcast i16 %mask to <16 x i1>
+  %x = call <16 x i32> @llvm.x86.avx512.mask.gather.dpi.512(<16 x i32> %src, i8* %base, <16 x i32> %ind, <16 x i1> %1, i32 4)
   %ind2 = add <16 x i32> %ind, <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
   call void @llvm.x86.avx512.scatter.dpi.512 (i8* %stbuf, i16 %mask, <16 x i32>%ind2, <16 x i32> %x, i32 4)
   ret void
 }
 
-define void @gather_mask_qd(<8 x i64> %ind, <8 x i32> %src, i8 %mask, i8* %base, i8* %stbuf)  {
+define void @gather_mask_qd(<8 x i64> %ind, <8 x i32> %src, i8 %mask, i8* %base, i8* %stbuf) {
 ; CHECK-LABEL: gather_mask_qd:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %edi, %k1
 ; CHECK-NEXT:    kmovq %k1, %k2
 ; CHECK-NEXT:    vpgatherqd (%rsi,%zmm0,4), %ymm1 {%k2}
 ; CHECK-NEXT:    vpaddq {{.*}}(%rip), %zmm0, %zmm0
 ; CHECK-NEXT:    vpscatterqd %ymm1, (%rdx,%zmm0,4) {%k1}
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %x = call <8 x i32> @llvm.x86.avx512.gather.qpi.512 (<8 x i32> %src, i8* %base, <8 x i64>%ind, i8 %mask, i32 4)
+  %1 = bitcast i8 %mask to <8 x i1>
+  %x = call <8 x i32> @llvm.x86.avx512.mask.gather.qpi.512(<8 x i32> %src, i8* %base, <8 x i64> %ind, <8 x i1> %1, i32 4)
   %ind2 = add <8 x i64> %ind, <i64 0, i64 1, i64 2, i64 3, i64 0, i64 1, i64 2, i64 3>
   call void @llvm.x86.avx512.scatter.qpi.512 (i8* %stbuf, i8 %mask, <8 x i64>%ind2, <8 x i32> %x, i32 4)
   ret void
 }
 
-define void @gather_mask_qq(<8 x i64> %ind, <8 x i64> %src, i8 %mask, i8* %base, i8* %stbuf)  {
+define void @gather_mask_qq(<8 x i64> %ind, <8 x i64> %src, i8 %mask, i8* %base, i8* %stbuf) {
 ; CHECK-LABEL: gather_mask_qq:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %edi, %k1
 ; CHECK-NEXT:    kmovq %k1, %k2
 ; CHECK-NEXT:    vpgatherqq (%rsi,%zmm0,4), %zmm1 {%k2}
 ; CHECK-NEXT:    vpaddq {{.*}}(%rip), %zmm0, %zmm0
 ; CHECK-NEXT:    vpscatterqq %zmm1, (%rdx,%zmm0,4) {%k1}
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %x = call <8 x i64> @llvm.x86.avx512.gather.qpq.512 (<8 x i64> %src, i8* %base, <8 x i64>%ind, i8 %mask, i32 4)
+  %1 = bitcast i8 %mask to <8 x i1>
+  %x = call <8 x i64> @llvm.x86.avx512.mask.gather.qpq.512(<8 x i64> %src, i8* %base, <8 x i64> %ind, <8 x i1> %1, i32 4)
   %ind2 = add <8 x i64> %ind, <i64 0, i64 1, i64 2, i64 3, i64 0, i64 1, i64 2, i64 3>
   call void @llvm.x86.avx512.scatter.qpq.512 (i8* %stbuf, i8 %mask, <8 x i64>%ind2, <8 x i64> %x, i32 4)
   ret void
 }
 
-define void @gather_mask_dq(<8 x i32> %ind, <8 x i64> %src, i8 %mask, i8* %base, i8* %stbuf)  {
+define void @gather_mask_dq(<8 x i32> %ind, <8 x i64> %src, i8 %mask, i8* %base, i8* %stbuf) {
 ; CHECK-LABEL: gather_mask_dq:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %edi, %k1
 ; CHECK-NEXT:    kmovq %k1, %k2
 ; CHECK-NEXT:    vpgatherdq (%rsi,%ymm0,4), %zmm1 {%k2}
 ; CHECK-NEXT:    vpaddd {{.*}}(%rip), %ymm0, %ymm0
 ; CHECK-NEXT:    vpscatterdq %zmm1, (%rdx,%ymm0,4) {%k1}
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %x = call <8 x i64> @llvm.x86.avx512.gather.dpq.512 (<8 x i64> %src, i8* %base, <8 x i32>%ind, i8 %mask, i32 4)
+  %1 = bitcast i8 %mask to <8 x i1>
+  %x = call <8 x i64> @llvm.x86.avx512.mask.gather.dpq.512(<8 x i64> %src, i8* %base, <8 x i32> %ind, <8 x i1> %1, i32 4)
   %ind2 = add <8 x i32> %ind, <i32 0, i32 1, i32 2, i32 3, i32 0, i32 1, i32 2, i32 3>
   call void @llvm.x86.avx512.scatter.dpq.512 (i8* %stbuf, i8 %mask, <8 x i32>%ind2, <8 x i64> %x, i32 4)
   ret void
 }
 
-define void @gather_mask_dpd_execdomain(<8 x i32> %ind, <8 x double> %src, i8 %mask, i8* %base, <8 x double>* %stbuf)  {
+define void @gather_mask_dpd_execdomain(<8 x i32> %ind, <8 x double> %src, i8 %mask, i8* %base, <8 x double>* %stbuf) {
 ; CHECK-LABEL: gather_mask_dpd_execdomain:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %edi, %k1
 ; CHECK-NEXT:    vgatherdpd (%rsi,%ymm0,4), %zmm1 {%k1}
 ; CHECK-NEXT:    vmovapd %zmm1, (%rdx)
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %x = call <8 x double> @llvm.x86.avx512.gather.dpd.512 (<8 x double> %src, i8* %base, <8 x i32>%ind, i8 %mask, i32 4)
+  %1 = bitcast i8 %mask to <8 x i1>
+  %x = call <8 x double> @llvm.x86.avx512.mask.gather.dpd.512(<8 x double> %src, i8* %base, <8 x i32> %ind, <8 x i1> %1, i32 4)
   store <8 x double> %x, <8 x double>* %stbuf
   ret void
 }
 
-define void @gather_mask_qpd_execdomain(<8 x i64> %ind, <8 x double> %src, i8 %mask, i8* %base, <8 x double>* %stbuf)  {
+define void @gather_mask_qpd_execdomain(<8 x i64> %ind, <8 x double> %src, i8 %mask, i8* %base, <8 x double>* %stbuf) {
 ; CHECK-LABEL: gather_mask_qpd_execdomain:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %edi, %k1
 ; CHECK-NEXT:    vgatherqpd (%rsi,%zmm0,4), %zmm1 {%k1}
 ; CHECK-NEXT:    vmovapd %zmm1, (%rdx)
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %x = call <8 x double> @llvm.x86.avx512.gather.qpd.512 (<8 x double> %src, i8* %base, <8 x i64>%ind, i8 %mask, i32 4)
+  %1 = bitcast i8 %mask to <8 x i1>
+  %x = call <8 x double> @llvm.x86.avx512.mask.gather.qpd.512(<8 x double> %src, i8* %base, <8 x i64> %ind, <8 x i1> %1, i32 4)
   store <8 x double> %x, <8 x double>* %stbuf
   ret void
 }
 
-define <16 x float> @gather_mask_dps_execdomain(<16 x i32> %ind, <16 x float> %src, i16 %mask, i8* %base)  {
+define <16 x float> @gather_mask_dps_execdomain(<16 x i32> %ind, <16 x float> %src, i16 %mask, i8* %base) {
 ; CHECK-LABEL: gather_mask_dps_execdomain:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %edi, %k1
 ; CHECK-NEXT:    vgatherdps (%rsi,%zmm0,4), %zmm1 {%k1}
 ; CHECK-NEXT:    vmovaps %zmm1, %zmm0
 ; CHECK-NEXT:    retq
-  %res = call <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float> %src, i8* %base, <16 x i32>%ind, i16 %mask, i32 4)
-  ret <16 x float> %res;
+  %1 = bitcast i16 %mask to <16 x i1>
+  %res = call <16 x float> @llvm.x86.avx512.mask.gather.dps.512(<16 x float> %src, i8* %base, <16 x i32> %ind, <16 x i1> %1, i32 4)
+  ret <16 x float> %res
 }
 
-define <8 x float> @gather_mask_qps_execdomain(<8 x i64> %ind, <8 x float> %src, i8 %mask, i8* %base)  {
+define <8 x float> @gather_mask_qps_execdomain(<8 x i64> %ind, <8 x float> %src, i8 %mask, i8* %base) {
 ; CHECK-LABEL: gather_mask_qps_execdomain:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %edi, %k1
 ; CHECK-NEXT:    vgatherqps (%rsi,%zmm0,4), %ymm1 {%k1}
 ; CHECK-NEXT:    vmovaps %ymm1, %ymm0
 ; CHECK-NEXT:    retq
-  %res = call <8 x float> @llvm.x86.avx512.gather.qps.512 (<8 x float> %src, i8* %base, <8 x i64>%ind, i8 %mask, i32 4)
-  ret <8 x float> %res;
+  %1 = bitcast i8 %mask to <8 x i1>
+  %res = call <8 x float> @llvm.x86.avx512.mask.gather.qps.512(<8 x float> %src, i8* %base, <8 x i64> %ind, <8 x i1> %1, i32 4)
+  ret <8 x float> %res
 }
 
-define void @scatter_mask_dpd_execdomain(<8 x i32> %ind, <8 x double>* %src, i8 %mask, i8* %base, i8* %stbuf)  {
+define void @scatter_mask_dpd_execdomain(<8 x i32> %ind, <8 x double>* %src, i8 %mask, i8* %base, i8* %stbuf) {
 ; CHECK-LABEL: scatter_mask_dpd_execdomain:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vmovapd (%rdi), %zmm1
 ; CHECK-NEXT:    vscatterdpd %zmm1, (%rcx,%ymm0,4) {%k1}
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
   %x = load <8 x double>, <8 x double>* %src, align 64

 ; CHECK-NEXT:    vscatterqps %ymm1, (%rcx,%zmm0,4) {%k1}
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
   %x = load <8 x float>, <8 x float>* %src, align 32
   call void @llvm.x86.avx512.scatter.qps.512 (i8* %stbuf, i8 %mask, <8 x i64>%ind, <8 x float> %x, i32 4)
   ret void
 }
 
-define void @gather_qps(<8 x i64> %ind, <8 x float> %src, i8* %base, i8* %stbuf)  {
+define void @gather_qps(<8 x i64> %ind, <8 x float> %src, i8* %base, i8* %stbuf) {
 ; CHECK-LABEL: gather_qps:
 ; CHECK:       ## %bb.0:
-; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
 ; CHECK-NEXT:    kxnorw %k0, %k0, %k1
+; CHECK-NEXT:    vxorps %xmm1, %xmm1, %xmm1
 ; CHECK-NEXT:    kxnorw %k0, %k0, %k2
 ; CHECK-NEXT:    vgatherqps (%rdi,%zmm0,4), %ymm1 {%k2}
 ; CHECK-NEXT:    vpaddq {{.*}}(%rip), %zmm0, %zmm0
 ; CHECK-NEXT:    vscatterqps %ymm1, (%rsi,%zmm0,4) {%k1}
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %x = call <8 x float> @llvm.x86.avx512.gather.qps.512 (<8 x float> %src, i8* %base, <8 x i64>%ind, i8 -1, i32 4)
+  %x = call <8 x float> @llvm.x86.avx512.mask.gather.qps.512(<8 x float> %src, i8* %base, <8 x i64> %ind, <8 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>, i32 4)
   %ind2 = add <8 x i64> %ind, <i64 0, i64 1, i64 2, i64 3, i64 0, i64 1, i64 2, i64 3>
   call void @llvm.x86.avx512.scatter.qps.512 (i8* %stbuf, i8 -1, <8 x i64>%ind2, <8 x float> %x, i32 4)
   ret void
 }
 
 declare  void @llvm.x86.avx512.gatherpf.qps.512(i8, <8 x i64>, i8* , i32, i32);
 declare  void @llvm.x86.avx512.scatterpf.qps.512(i8, <8 x i64>, i8* , i32, i32);
 define void @prefetch(<8 x i64> %ind, i8* %base) {

 ; CHECK-NEXT:    retq
   call void @llvm.x86.avx512.gatherpf.qps.512(i8 -1, <8 x i64> %ind, i8* %base, i32 4, i32 3)
   call void @llvm.x86.avx512.gatherpf.qps.512(i8 0, <8 x i64> %ind, i8* %base, i32 4, i32 2)
   call void @llvm.x86.avx512.scatterpf.qps.512(i8 1, <8 x i64> %ind, i8* %base, i32 2, i32 3)
   call void @llvm.x86.avx512.scatterpf.qps.512(i8 120, <8 x i64> %ind, i8* %base, i32 2, i32 2)
   ret void
 }
 
-declare <2 x double> @llvm.x86.avx512.gather3div2.df(<2 x double>, i8*, <2 x i64>, i8, i32)
-
-define <2 x double>@test_int_x86_avx512_gather3div2_df(<2 x double> %x0, i8* %x1, <2 x i64> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3div2_df:
+define <2 x double> @test_int_x86_avx512_mask_gather3div2_df(<2 x double> %x0, i8* %x1, <2 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3div2_df:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vgatherqpd (%rdi,%xmm1,4), %xmm0 {%k1}
 ; CHECK-NEXT:    kxnorw %k0, %k0, %k1
 ; CHECK-NEXT:    vxorpd %xmm2, %xmm2, %xmm2
 ; CHECK-NEXT:    vgatherqpd (%rdi,%xmm1,2), %xmm2 {%k1}
 ; CHECK-NEXT:    vaddpd %xmm2, %xmm0, %xmm0
 ; CHECK-NEXT:    retq
-  %res = call <2 x double> @llvm.x86.avx512.gather3div2.df(<2 x double> %x0, i8* %x1, <2 x i64> %x2, i8 %x3, i32 4)
-  %res1 = call <2 x double> @llvm.x86.avx512.gather3div2.df(<2 x double> %x0, i8* %x1, <2 x i64> %x2, i8 -1, i32 2)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %1, <8 x i1> %1, <2 x i32> <i32 0, i32 1>
+  %res = call <2 x double> @llvm.x86.avx512.mask.gather3div2.df(<2 x double> %x0, i8* %x1, <2 x i64> %x2, <2 x i1> %extract, i32 4)
+  %res1 = call <2 x double> @llvm.x86.avx512.mask.gather3div2.df(<2 x double> %x0, i8* %x1, <2 x i64> %x2, <2 x i1> <i1 true, i1 true>, i32 2)
   %res2 = fadd <2 x double> %res, %res1
   ret <2 x double> %res2
 }
 
-declare <2 x i64> @llvm.x86.avx512.gather3div2.di(<2 x i64>, i8*, <2 x i64>, i8, i32)
-
-define <2 x i64>@test_int_x86_avx512_gather3div2_di(<2 x i64> %x0, i8* %x1, <2 x i64> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3div2_di:
+define <2 x i64> @test_int_x86_avx512_mask_gather3div2_di(<2 x i64> %x0, i8* %x1, <2 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3div2_di:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vpgatherqq (%rdi,%xmm1,8), %xmm0 {%k1}
 ; CHECK-NEXT:    vpaddq %xmm0, %xmm0, %xmm0
 ; CHECK-NEXT:    retq
-  %res = call <2 x i64> @llvm.x86.avx512.gather3div2.di(<2 x i64> %x0, i8* %x1, <2 x i64> %x2, i8 %x3, i32 8)
-  %res1 = call <2 x i64> @llvm.x86.avx512.gather3div2.di(<2 x i64> %x0, i8* %x1, <2 x i64> %x2, i8 %x3, i32 8)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract1 = shufflevector <8 x i1> %1, <8 x i1> %1, <2 x i32> <i32 0, i32 1>
+  %res = call <2 x i64> @llvm.x86.avx512.mask.gather3div2.di(<2 x i64> %x0, i8* %x1, <2 x i64> %x2, <2 x i1> %extract1, i32 8)
+  %2 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %2, <8 x i1> %2, <2 x i32> <i32 0, i32 1>
+  %res1 = call <2 x i64> @llvm.x86.avx512.mask.gather3div2.di(<2 x i64> %x0, i8* %x1, <2 x i64> %x2, <2 x i1> %extract, i32 8)
   %res2 = add <2 x i64> %res, %res1
   ret <2 x i64> %res2
 }
 
-declare <4 x double> @llvm.x86.avx512.gather3div4.df(<4 x double>, i8*, <4 x i64>, i8, i32)
-
-define <4 x double>@test_int_x86_avx512_gather3div4_df(<4 x double> %x0, i8* %x1, <4 x i64> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3div4_df:
+define <4 x double> @test_int_x86_avx512_mask_gather3div4_df(<4 x double> %x0, i8* %x1, <4 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3div4_df:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vgatherqpd (%rdi,%ymm1,4), %ymm0 {%k1}
 ; CHECK-NEXT:    kxnorw %k0, %k0, %k1
 ; CHECK-NEXT:    vxorpd %xmm2, %xmm2, %xmm2
 ; CHECK-NEXT:    vgatherqpd (%rdi,%ymm1,2), %ymm2 {%k1}
 ; CHECK-NEXT:    vaddpd %ymm2, %ymm0, %ymm0
 ; CHECK-NEXT:    retq
-  %res = call <4 x double> @llvm.x86.avx512.gather3div4.df(<4 x double> %x0, i8* %x1, <4 x i64> %x2, i8 %x3, i32 4)
-  %res1 = call <4 x double> @llvm.x86.avx512.gather3div4.df(<4 x double> %x0, i8* %x1, <4 x i64> %x2, i8 -1, i32 2)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %1, <8 x i1> %1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %res = call <4 x double> @llvm.x86.avx512.mask.gather3div4.df(<4 x double> %x0, i8* %x1, <4 x i64> %x2, <4 x i1> %extract, i32 4)
+  %res1 = call <4 x double> @llvm.x86.avx512.mask.gather3div4.df(<4 x double> %x0, i8* %x1, <4 x i64> %x2, <4 x i1> <i1 true, i1 true, i1 true, i1 true>, i32 2)
   %res2 = fadd <4 x double> %res, %res1
   ret <4 x double> %res2
 }
 
-declare <4 x i64> @llvm.x86.avx512.gather3div4.di(<4 x i64>, i8*, <4 x i64>, i8, i32)
-
-define <4 x i64>@test_int_x86_avx512_gather3div4_di(<4 x i64> %x0, i8* %x1, <4 x i64> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3div4_di:
+define <4 x i64> @test_int_x86_avx512_mask_gather3div4_di(<4 x i64> %x0, i8* %x1, <4 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3div4_di:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vpgatherqq (%rdi,%ymm1,8), %ymm0 {%k1}
 ; CHECK-NEXT:    kxnorw %k0, %k0, %k1
 ; CHECK-NEXT:    vpxor %xmm2, %xmm2, %xmm2
 ; CHECK-NEXT:    vpgatherqq (%rdi,%ymm1,8), %ymm2 {%k1}
 ; CHECK-NEXT:    vpaddq %ymm2, %ymm0, %ymm0
 ; CHECK-NEXT:    retq
-  %res = call <4 x i64> @llvm.x86.avx512.gather3div4.di(<4 x i64> %x0, i8* %x1, <4 x i64> %x2, i8 %x3, i32 8)
-  %res1 = call <4 x i64> @llvm.x86.avx512.gather3div4.di(<4 x i64> %x0, i8* %x1, <4 x i64> %x2, i8 -1, i32 8)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %1, <8 x i1> %1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %res = call <4 x i64> @llvm.x86.avx512.mask.gather3div4.di(<4 x i64> %x0, i8* %x1, <4 x i64> %x2, <4 x i1> %extract, i32 8)
+  %res1 = call <4 x i64> @llvm.x86.avx512.mask.gather3div4.di(<4 x i64> %x0, i8* %x1, <4 x i64> %x2, <4 x i1> <i1 true, i1 true, i1 true, i1 true>, i32 8)
   %res2 = add <4 x i64> %res, %res1
   ret <4 x i64> %res2
 }
 
-declare <4 x float> @llvm.x86.avx512.gather3div4.sf(<4 x float>, i8*, <2 x i64>, i8, i32)
-
-define <4 x float>@test_int_x86_avx512_gather3div4_sf(<4 x float> %x0, i8* %x1, <2 x i64> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3div4_sf:
+define <4 x float> @test_int_x86_avx512_mask_gather3div4_sf(<4 x float> %x0, i8* %x1, <2 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3div4_sf:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vgatherqps (%rdi,%xmm1,4), %xmm0 {%k1}
 ; CHECK-NEXT:    kxnorw %k0, %k0, %k1
 ; CHECK-NEXT:    vxorps %xmm2, %xmm2, %xmm2
 ; CHECK-NEXT:    vgatherqps (%rdi,%xmm1,2), %xmm2 {%k1}
 ; CHECK-NEXT:    vaddps %xmm2, %xmm0, %xmm0
 ; CHECK-NEXT:    retq
-  %res = call <4 x float> @llvm.x86.avx512.gather3div4.sf(<4 x float> %x0, i8* %x1, <2 x i64> %x2, i8 %x3, i32 4)
-  %res1 = call <4 x float> @llvm.x86.avx512.gather3div4.sf(<4 x float> %x0, i8* %x1, <2 x i64> %x2, i8 -1, i32 2)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %1, <8 x i1> %1, <2 x i32> <i32 0, i32 1>
+  %res = call <4 x float> @llvm.x86.avx512.mask.gather3div4.sf(<4 x float> %x0, i8* %x1, <2 x i64> %x2, <2 x i1> %extract, i32 4)
+  %res1 = call <4 x float> @llvm.x86.avx512.mask.gather3div4.sf(<4 x float> %x0, i8* %x1, <2 x i64> %x2, <2 x i1> <i1 true, i1 true>, i32 2)
   %res2 = fadd <4 x float> %res, %res1
   ret <4 x float> %res2
 }
 
-declare <4 x i32> @llvm.x86.avx512.gather3div4.si(<4 x i32>, i8*, <2 x i64>, i8, i32)
-
-define <4 x i32>@test_int_x86_avx512_gather3div4_si(<4 x i32> %x0, i8* %x1, <2 x i64> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3div4_si:
+define <4 x i32> @test_int_x86_avx512_mask_gather3div4_si(<4 x i32> %x0, i8* %x1, <2 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3div4_si:
 ; CHECK:       ## %bb.0:
-; CHECK-NEXT:    kmovd %esi, %k1
-; CHECK-NEXT:    kxnorw %k0, %k0, %k2
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
 ; CHECK-NEXT:    vpxor %xmm2, %xmm2, %xmm2
-; CHECK-NEXT:    vpgatherqd (%rdi,%xmm1,4), %xmm2 {%k2}
+; CHECK-NEXT:    vpgatherqd (%rdi,%xmm1,4), %xmm2 {%k1}
+; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vpgatherqd (%rdi,%xmm1,4), %xmm0 {%k1}
 ; CHECK-NEXT:    vpaddd %xmm0, %xmm2, %xmm0
 ; CHECK-NEXT:    retq
-  %res = call <4 x i32> @llvm.x86.avx512.gather3div4.si(<4 x i32> %x0, i8* %x1, <2 x i64> %x2, i8 -1, i32 4)
-  %res1 = call <4 x i32> @llvm.x86.avx512.gather3div4.si(<4 x i32> %x0, i8* %x1, <2 x i64> %x2, i8 %x3, i32 4)
+  %res = call <4 x i32> @llvm.x86.avx512.mask.gather3div4.si(<4 x i32> %x0, i8* %x1, <2 x i64> %x2, <2 x i1> <i1 true, i1 true>, i32 4)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %1, <8 x i1> %1, <2 x i32> <i32 0, i32 1>
+  %res1 = call <4 x i32> @llvm.x86.avx512.mask.gather3div4.si(<4 x i32> %x0, i8* %x1, <2 x i64> %x2, <2 x i1> %extract, i32 4)
   %res2 = add <4 x i32> %res, %res1
   ret <4 x i32> %res2
 }
 
-declare <4 x float> @llvm.x86.avx512.gather3div8.sf(<4 x float>, i8*, <4 x i64>, i8, i32)
-
-define <4 x float>@test_int_x86_avx512_gather3div8_sf(<4 x float> %x0, i8* %x1, <4 x i64> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3div8_sf:
+define <4 x float> @test_int_x86_avx512_mask_gather3div8_sf(<4 x float> %x0, i8* %x1, <4 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3div8_sf:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vgatherqps (%rdi,%ymm1,4), %xmm0 {%k1}
 ; CHECK-NEXT:    kxnorw %k0, %k0, %k1
 ; CHECK-NEXT:    vxorps %xmm2, %xmm2, %xmm2
 ; CHECK-NEXT:    vgatherqps (%rdi,%ymm1,2), %xmm2 {%k1}
 ; CHECK-NEXT:    vaddps %xmm2, %xmm0, %xmm0
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %res = call <4 x float> @llvm.x86.avx512.gather3div8.sf(<4 x float> %x0, i8* %x1, <4 x i64> %x2, i8 %x3, i32 4)
-  %res1 = call <4 x float> @llvm.x86.avx512.gather3div8.sf(<4 x float> %x0, i8* %x1, <4 x i64> %x2, i8 -1, i32 2)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %1, <8 x i1> %1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %res = call <4 x float> @llvm.x86.avx512.mask.gather3div8.sf(<4 x float> %x0, i8* %x1, <4 x i64> %x2, <4 x i1> %extract, i32 4)
+  %res1 = call <4 x float> @llvm.x86.avx512.mask.gather3div8.sf(<4 x float> %x0, i8* %x1, <4 x i64> %x2, <4 x i1> <i1 true, i1 true, i1 true, i1 true>, i32 2)
   %res2 = fadd <4 x float> %res, %res1
   ret <4 x float> %res2
 }
 
-declare <4 x i32> @llvm.x86.avx512.gather3div8.si(<4 x i32>, i8*, <4 x i64>, i8, i32)
-
-define <4 x i32>@test_int_x86_avx512_gather3div8_si(<4 x i32> %x0, i8* %x1, <4 x i64> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3div8_si:
+define <4 x i32> @test_int_x86_avx512_mask_gather3div8_si(<4 x i32> %x0, i8* %x1, <4 x i64> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3div8_si:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vmovdqa %xmm0, %xmm2
 ; CHECK-NEXT:    kmovq %k1, %k2
 ; CHECK-NEXT:    vpgatherqd (%rdi,%ymm1,4), %xmm2 {%k2}
 ; CHECK-NEXT:    vpgatherqd (%rdi,%ymm1,2), %xmm0 {%k1}
 ; CHECK-NEXT:    vpaddd %xmm0, %xmm2, %xmm0
 ; CHECK-NEXT:    vzeroupper
 ; CHECK-NEXT:    retq
-  %res = call <4 x i32> @llvm.x86.avx512.gather3div8.si(<4 x i32> %x0, i8* %x1, <4 x i64> %x2, i8 %x3, i32 4)
-  %res1 = call <4 x i32> @llvm.x86.avx512.gather3div8.si(<4 x i32> %x0, i8* %x1, <4 x i64> %x2, i8 %x3, i32 2)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract1 = shufflevector <8 x i1> %1, <8 x i1> %1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %res = call <4 x i32> @llvm.x86.avx512.mask.gather3div8.si(<4 x i32> %x0, i8* %x1, <4 x i64> %x2, <4 x i1> %extract1, i32 4)
+  %2 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %2, <8 x i1> %2, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %res1 = call <4 x i32> @llvm.x86.avx512.mask.gather3div8.si(<4 x i32> %x0, i8* %x1, <4 x i64> %x2, <4 x i1> %extract, i32 2)
   %res2 = add <4 x i32> %res, %res1
   ret <4 x i32> %res2
 }
 
-declare <2 x double> @llvm.x86.avx512.gather3siv2.df(<2 x double>, i8*, <4 x i32>, i8, i32)
-
-define <2 x double>@test_int_x86_avx512_gather3siv2_df(<2 x double> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3siv2_df:
+define <2 x double> @test_int_x86_avx512_mask_gather3siv2_df(<2 x double> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3siv2_df:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vgatherdpd (%rdi,%xmm1,4), %xmm0 {%k1}
 ; CHECK-NEXT:    kxnorw %k0, %k0, %k1
 ; CHECK-NEXT:    vxorpd %xmm2, %xmm2, %xmm2
 ; CHECK-NEXT:    vgatherdpd (%rdi,%xmm1,2), %xmm2 {%k1}
 ; CHECK-NEXT:    vaddpd %xmm2, %xmm0, %xmm0
 ; CHECK-NEXT:    retq
-  %res = call <2 x double> @llvm.x86.avx512.gather3siv2.df(<2 x double> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 4)
-  %res1 = call <2 x double> @llvm.x86.avx512.gather3siv2.df(<2 x double> %x0, i8* %x1, <4 x i32> %x2, i8 -1, i32 2)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %1, <8 x i1> %1, <2 x i32> <i32 0, i32 1>
+  %res = call <2 x double> @llvm.x86.avx512.mask.gather3siv2.df(<2 x double> %x0, i8* %x1, <4 x i32> %x2, <2 x i1> %extract, i32 4)
+  %res1 = call <2 x double> @llvm.x86.avx512.mask.gather3siv2.df(<2 x double> %x0, i8* %x1, <4 x i32> %x2, <2 x i1> <i1 true, i1 true>, i32 2)
   %res2 = fadd <2 x double> %res, %res1
   ret <2 x double> %res2
 }
 
-declare <2 x i64> @llvm.x86.avx512.gather3siv2.di(<2 x i64>, i8*, <4 x i32>, i8, i32)
-
-define <2 x i64>@test_int_x86_avx512_gather3siv2_di(<2 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3siv2_di:
+define <2 x i64> @test_int_x86_avx512_mask_gather3siv2_di(<2 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3siv2_di:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vpgatherdq (%rdi,%xmm1,8), %xmm0 {%k1}
 ; CHECK-NEXT:    vpaddq %xmm0, %xmm0, %xmm0
 ; CHECK-NEXT:    retq
-  %res = call <2 x i64> @llvm.x86.avx512.gather3siv2.di(<2 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 8)
-  %res1 = call <2 x i64> @llvm.x86.avx512.gather3siv2.di(<2 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 8)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract1 = shufflevector <8 x i1> %1, <8 x i1> %1, <2 x i32> <i32 0, i32 1>
+  %res = call <2 x i64> @llvm.x86.avx512.mask.gather3siv2.di(<2 x i64> %x0, i8* %x1, <4 x i32> %x2, <2 x i1> %extract1, i32 8)
+  %2 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %2, <8 x i1> %2, <2 x i32> <i32 0, i32 1>
+  %res1 = call <2 x i64> @llvm.x86.avx512.mask.gather3siv2.di(<2 x i64> %x0, i8* %x1, <4 x i32> %x2, <2 x i1> %extract, i32 8)
   %res2 = add <2 x i64> %res, %res1
   ret <2 x i64> %res2
 }
 
-declare <4 x double> @llvm.x86.avx512.gather3siv4.df(<4 x double>, i8*, <4 x i32>, i8, i32)
-
-define <4 x double>@test_int_x86_avx512_gather3siv4_df(<4 x double> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3siv4_df:
+define <4 x double> @test_int_x86_avx512_mask_gather3siv4_df(<4 x double> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3siv4_df:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vgatherdpd (%rdi,%xmm1,4), %ymm0 {%k1}
 ; CHECK-NEXT:    kxnorw %k0, %k0, %k1
 ; CHECK-NEXT:    vxorpd %xmm2, %xmm2, %xmm2
 ; CHECK-NEXT:    vgatherdpd (%rdi,%xmm1,2), %ymm2 {%k1}
 ; CHECK-NEXT:    vaddpd %ymm2, %ymm0, %ymm0
 ; CHECK-NEXT:    retq
-  %res = call <4 x double> @llvm.x86.avx512.gather3siv4.df(<4 x double> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 4)
-  %res1 = call <4 x double> @llvm.x86.avx512.gather3siv4.df(<4 x double> %x0, i8* %x1, <4 x i32> %x2, i8 -1, i32 2)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %1, <8 x i1> %1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %res = call <4 x double> @llvm.x86.avx512.mask.gather3siv4.df(<4 x double> %x0, i8* %x1, <4 x i32> %x2, <4 x i1> %extract, i32 4)
+  %res1 = call <4 x double> @llvm.x86.avx512.mask.gather3siv4.df(<4 x double> %x0, i8* %x1, <4 x i32> %x2, <4 x i1> <i1 true, i1 true, i1 true, i1 true>, i32 2)
   %res2 = fadd <4 x double> %res, %res1
   ret <4 x double> %res2
 }
 
-declare <4 x i64> @llvm.x86.avx512.gather3siv4.di(<4 x i64>, i8*, <4 x i32>, i8, i32)
-
-define <4 x i64>@test_int_x86_avx512_gather3siv4_di(<4 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3siv4_di:
+define <4 x i64> @test_int_x86_avx512_mask_gather3siv4_di(<4 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3siv4_di:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vpgatherdq (%rdi,%xmm1,8), %ymm0 {%k1}
 ; CHECK-NEXT:    vpaddq %ymm0, %ymm0, %ymm0
 ; CHECK-NEXT:    retq
-  %res = call <4 x i64> @llvm.x86.avx512.gather3siv4.di(<4 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 8)
-  %res1 = call <4 x i64> @llvm.x86.avx512.gather3siv4.di(<4 x i64> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 8)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract1 = shufflevector <8 x i1> %1, <8 x i1> %1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %res = call <4 x i64> @llvm.x86.avx512.mask.gather3siv4.di(<4 x i64> %x0, i8* %x1, <4 x i32> %x2, <4 x i1> %extract1, i32 8)
+  %2 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %2, <8 x i1> %2, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %res1 = call <4 x i64> @llvm.x86.avx512.mask.gather3siv4.di(<4 x i64> %x0, i8* %x1, <4 x i32> %x2, <4 x i1> %extract, i32 8)
   %res2 = add <4 x i64> %res, %res1
   ret <4 x i64> %res2
 }
 
-declare <4 x float> @llvm.x86.avx512.gather3siv4.sf(<4 x float>, i8*, <4 x i32>, i8, i32)
-
-define <4 x float>@test_int_x86_avx512_gather3siv4_sf(<4 x float> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3siv4_sf:
+define <4 x float> @test_int_x86_avx512_mask_gather3siv4_sf(<4 x float> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3siv4_sf:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vgatherdps (%rdi,%xmm1,4), %xmm0 {%k1}
 ; CHECK-NEXT:    kxnorw %k0, %k0, %k1
 ; CHECK-NEXT:    vxorps %xmm2, %xmm2, %xmm2
 ; CHECK-NEXT:    vgatherdps (%rdi,%xmm1,2), %xmm2 {%k1}
 ; CHECK-NEXT:    vaddps %xmm2, %xmm0, %xmm0
 ; CHECK-NEXT:    retq
-  %res = call <4 x float> @llvm.x86.avx512.gather3siv4.sf(<4 x float> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 4)
-  %res1 = call <4 x float> @llvm.x86.avx512.gather3siv4.sf(<4 x float> %x0, i8* %x1, <4 x i32> %x2, i8 -1, i32 2)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %1, <8 x i1> %1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %res = call <4 x float> @llvm.x86.avx512.mask.gather3siv4.sf(<4 x float> %x0, i8* %x1, <4 x i32> %x2, <4 x i1> %extract, i32 4)
+  %res1 = call <4 x float> @llvm.x86.avx512.mask.gather3siv4.sf(<4 x float> %x0, i8* %x1, <4 x i32> %x2, <4 x i1> <i1 true, i1 true, i1 true, i1 true>, i32 2)
   %res2 = fadd <4 x float> %res, %res1
   ret <4 x float> %res2
 }
 
-declare <4 x i32> @llvm.x86.avx512.gather3siv4.si(<4 x i32>, i8*, <4 x i32>, i8, i32)
-
-define <4 x i32>@test_int_x86_avx512_gather3siv4_si(<4 x i32> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3siv4_si:
+define <4 x i32> @test_int_x86_avx512_mask_gather3siv4_si(<4 x i32> %x0, i8* %x1, <4 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3siv4_si:
 ; CHECK:       ## %bb.0:
-; CHECK-NEXT:    kmovd %esi, %k1
-; CHECK-NEXT:    kxnorw %k0, %k0, %k2
+; CHECK-NEXT:    kxnorw %k0, %k0, %k1
 ; CHECK-NEXT:    vpxor %xmm2, %xmm2, %xmm2
-; CHECK-NEXT:    vpgatherdd (%rdi,%xmm1,4), %xmm2 {%k2}
+; CHECK-NEXT:    vpgatherdd (%rdi,%xmm1,4), %xmm2 {%k1}
+; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vpgatherdd (%rdi,%xmm1,2), %xmm0 {%k1}
 ; CHECK-NEXT:    vpaddd %xmm0, %xmm2, %xmm0
 ; CHECK-NEXT:    retq
-  %res = call <4 x i32> @llvm.x86.avx512.gather3siv4.si(<4 x i32> %x0, i8* %x1, <4 x i32> %x2, i8 -1, i32 4)
-  %res1 = call <4 x i32> @llvm.x86.avx512.gather3siv4.si(<4 x i32> %x0, i8* %x1, <4 x i32> %x2, i8 %x3, i32 2)
+  %res = call <4 x i32> @llvm.x86.avx512.mask.gather3siv4.si(<4 x i32> %x0, i8* %x1, <4 x i32> %x2, <4 x i1> <i1 true, i1 true, i1 true, i1 true>, i32 4)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %extract = shufflevector <8 x i1> %1, <8 x i1> %1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
+  %res1 = call <4 x i32> @llvm.x86.avx512.mask.gather3siv4.si(<4 x i32> %x0, i8* %x1, <4 x i32> %x2, <4 x i1> %extract, i32 2)
   %res2 = add <4 x i32> %res, %res1
   ret <4 x i32> %res2
 }
 
-declare <8 x float> @llvm.x86.avx512.gather3siv8.sf(<8 x float>, i8*, <8 x i32>, i8, i32)
-
-define <8 x float>@test_int_x86_avx512_gather3siv8_sf(<8 x float> %x0, i8* %x1, <8 x i32> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3siv8_sf:
+define <8 x float> @test_int_x86_avx512_mask_gather3siv8_sf(<8 x float> %x0, i8* %x1, <8 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3siv8_sf:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vgatherdps (%rdi,%ymm1,4), %ymm0 {%k1}
 ; CHECK-NEXT:    kxnorw %k0, %k0, %k1
 ; CHECK-NEXT:    vxorps %xmm2, %xmm2, %xmm2
 ; CHECK-NEXT:    vgatherdps (%rdi,%ymm1,2), %ymm2 {%k1}
 ; CHECK-NEXT:    vaddps %ymm2, %ymm0, %ymm0
 ; CHECK-NEXT:    retq
-  %res = call <8 x float> @llvm.x86.avx512.gather3siv8.sf(<8 x float> %x0, i8* %x1, <8 x i32> %x2, i8 %x3, i32 4)
-  %res1 = call <8 x float> @llvm.x86.avx512.gather3siv8.sf(<8 x float> %x0, i8* %x1, <8 x i32> %x2, i8 -1, i32 2)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %res = call <8 x float> @llvm.x86.avx512.mask.gather3siv8.sf(<8 x float> %x0, i8* %x1, <8 x i32> %x2, <8 x i1> %1, i32 4)
+  %res1 = call <8 x float> @llvm.x86.avx512.mask.gather3siv8.sf(<8 x float> %x0, i8* %x1, <8 x i32> %x2, <8 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>, i32 2)
   %res2 = fadd <8 x float> %res, %res1
   ret <8 x float> %res2
 }
 
-declare <8 x i32> @llvm.x86.avx512.gather3siv8.si(<8 x i32>, i8*, <8 x i32>, i8, i32)
-
-define <8 x i32>@test_int_x86_avx512_gather3siv8_si(<8 x i32> %x0, i8* %x1, <8 x i32> %x2, i8 %x3) {
-; CHECK-LABEL: test_int_x86_avx512_gather3siv8_si:
+define <8 x i32> @test_int_x86_avx512_mask_gather3siv8_si(<8 x i32> %x0, i8* %x1, <8 x i32> %x2, i8 %x3) {
+; CHECK-LABEL: test_int_x86_avx512_mask_gather3siv8_si:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kmovd %esi, %k1
 ; CHECK-NEXT:    vmovdqa %ymm0, %ymm2
 ; CHECK-NEXT:    kmovq %k1, %k2
 ; CHECK-NEXT:    vpgatherdd (%rdi,%ymm1,4), %ymm2 {%k2}
 ; CHECK-NEXT:    vpgatherdd (%rdi,%ymm1,2), %ymm0 {%k1}
 ; CHECK-NEXT:    vpaddd %ymm0, %ymm2, %ymm0
 ; CHECK-NEXT:    retq
-  %res = call <8 x i32> @llvm.x86.avx512.gather3siv8.si(<8 x i32> %x0, i8* %x1, <8 x i32> %x2, i8 %x3, i32 4)
-  %res1 = call <8 x i32> @llvm.x86.avx512.gather3siv8.si(<8 x i32> %x0, i8* %x1, <8 x i32> %x2, i8 %x3, i32 2)
+  %1 = bitcast i8 %x3 to <8 x i1>
+  %res = call <8 x i32> @llvm.x86.avx512.mask.gather3siv8.si(<8 x i32> %x0, i8* %x1, <8 x i32> %x2, <8 x i1> %1, i32 4)
+  %2 = bitcast i8 %x3 to <8 x i1>
+  %res1 = call <8 x i32> @llvm.x86.avx512.mask.gather3siv8.si(<8 x i32> %x0, i8* %x1, <8 x i32> %x2, <8 x i1> %2, i32 2)
   %res2 = add <8 x i32> %res, %res1
   ret <8 x i32> %res2
 }
 
 declare void @llvm.x86.avx512.scatterdiv2.df(i8*, i8, <2 x i64>, <2 x double>, i32)
 
 define void@test_int_x86_avx512_scatterdiv2_df(i8* %x0, i8 %x1, <2 x i64> %x2, <2 x double> %x3) {
 ; CHECK-LABEL: test_int_x86_avx512_scatterdiv2_df:

 ; CHECK-NEXT:    retq
   call void @llvm.x86.avx512.scattersiv8.si(i8* %x0, i8 -1, <8 x i32> %x2, <8 x i32> %x3, i32 2)
   call void @llvm.x86.avx512.scattersiv8.si(i8* %x0, i8 0, <8 x i32> %x2, <8 x i32> %x3, i32 4)
   call void @llvm.x86.avx512.scattersiv8.si(i8* %x0, i8 1, <8 x i32> %x2, <8 x i32> %x3, i32 2)
   call void @llvm.x86.avx512.scattersiv8.si(i8* %x0, i8 96, <8 x i32> %x2, <8 x i32> %x3, i32 4)
   ret void
 }
 
-define <16 x float> @gather_mask_test(<16 x i32> %ind, <16 x float> %src, i8* %base)  {
+define <16 x float> @gather_mask_test(<16 x i32> %ind, <16 x float> %src, i8* %base) {
 ; CHECK-LABEL: gather_mask_test:
 ; CHECK:       ## %bb.0:
 ; CHECK-NEXT:    kxnorw %k0, %k0, %k1
 ; CHECK-NEXT:    vxorps %xmm2, %xmm2, %xmm2
 ; CHECK-NEXT:    vgatherdps (%rdi,%zmm0,4), %zmm2 {%k1}
 ; CHECK-NEXT:    kxorw %k0, %k0, %k1
 ; CHECK-NEXT:    vmovaps %zmm1, %zmm3
 ; CHECK-NEXT:    vgatherdps (%rdi,%zmm0,4), %zmm3 {%k1}
 ; CHECK-NEXT:    vaddps %zmm3, %zmm2, %zmm2
 ; CHECK-NEXT:    movw $1, %ax
 ; CHECK-NEXT:    kmovd %eax, %k1
 ; CHECK-NEXT:    vmovaps %zmm1, %zmm3
 ; CHECK-NEXT:    vgatherdps (%rdi,%zmm0,4), %zmm3 {%k1}
 ; CHECK-NEXT:    movw $220, %ax
 ; CHECK-NEXT:    kmovd %eax, %k1
 ; CHECK-NEXT:    vgatherdps (%rdi,%zmm0,4), %zmm1 {%k1}
 ; CHECK-NEXT:    vaddps %zmm3, %zmm1, %zmm0
 ; CHECK-NEXT:    vaddps %zmm2, %zmm0, %zmm0
 ; CHECK-NEXT:    retq
-  %res = call <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float> %src, i8* %base, <16 x i32>%ind, i16 -1, i32 4)
-  %res1 = call <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float> %src, i8* %base, <16 x i32>%ind, i16 0, i32 4)
-  %res2 = call <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float> %src, i8* %base, <16 x i32>%ind, i16 1, i32 4)
-  %res3 = call <16 x float> @llvm.x86.avx512.gather.dps.512 (<16 x float> %src, i8* %base, <16 x i32>%ind, i16 220, i32 4)
-
+  %res = call <16 x float> @llvm.x86.avx512.mask.gather.dps.512(<16 x float> %src, i8* %base, <16 x i32> %ind, <16 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>, i32 4)
+  %res1 = call <16 x float> @llvm.x86.avx512.mask.gather.dps.512(<16 x float> %src, i8* %base, <16 x i32> %ind, <16 x i1> zeroinitializer, i32 4)
+  %res2 = call <16 x float> @llvm.x86.avx512.mask.gather.dps.512(<16 x float> %src, i8* %base, <16 x i32> %ind, <16 x i1> bitcast (<1 x i16> <i16 1> to <16 x i1>), i32 4)
+  %res3 = call <16 x float> @llvm.x86.avx512.mask.gather.dps.512(<16 x float> %src, i8* %base, <16 x i32> %ind, <16 x i1> bitcast (<1 x i16> <i16 220> to <16 x i1>), i32 4)
   %res4 = fadd <16 x float> %res, %res1
   %res5 = fadd <16 x float> %res3, %res2
   %res6 = fadd <16 x float> %res5, %res4
   ret <16 x float> %res6
 }
+
+declare <16 x float> @llvm.x86.avx512.mask.gather.dps.512(<16 x float>, i8*, <16 x i32>, <16 x i1>, i32)
+declare <8 x double> @llvm.x86.avx512.mask.gather.dpd.512(<8 x double>, i8*, <8 x i32>, <8 x i1>, i32)
+declare <8 x float> @llvm.x86.avx512.mask.gather.qps.512(<8 x float>, i8*, <8 x i64>, <8 x i1>, i32)
+declare <8 x double> @llvm.x86.avx512.mask.gather.qpd.512(<8 x double>, i8*, <8 x i64>, <8 x i1>, i32)
+declare <16 x i32> @llvm.x86.avx512.mask.gather.dpi.512(<16 x i32>, i8*, <16 x i32>, <16 x i1>, i32)
+declare <8 x i64> @llvm.x86.avx512.mask.gather.dpq.512(<8 x i64>, i8*, <8 x i32>, <8 x i1>, i32)
+declare <8 x i32> @llvm.x86.avx512.mask.gather.qpi.512(<8 x i32>, i8*, <8 x i64>, <8 x i1>, i32)
+declare <8 x i64> @llvm.x86.avx512.mask.gather.qpq.512(<8 x i64>, i8*, <8 x i64>, <8 x i1>, i32)
+declare <2 x double> @llvm.x86.avx512.mask.gather3div2.df(<2 x double>, i8*, <2 x i64>, <2 x i1>, i32)
+declare <2 x i64> @llvm.x86.avx512.mask.gather3div2.di(<2 x i64>, i8*, <2 x i64>, <2 x i1>, i32)
+declare <4 x double> @llvm.x86.avx512.mask.gather3div4.df(<4 x double>, i8*, <4 x i64>, <4 x i1>, i32)
+declare <4 x i64> @llvm.x86.avx512.mask.gather3div4.di(<4 x i64>, i8*, <4 x i64>, <4 x i1>, i32)
+declare <4 x float> @llvm.x86.avx512.mask.gather3div4.sf(<4 x float>, i8*, <2 x i64>, <2 x i1>, i32)
+declare <4 x i32> @llvm.x86.avx512.mask.gather3div4.si(<4 x i32>, i8*, <2 x i64>, <2 x i1>, i32)
+declare <4 x float> @llvm.x86.avx512.mask.gather3div8.sf(<4 x float>, i8*, <4 x i64>, <4 x i1>, i32)
+declare <4 x i32> @llvm.x86.avx512.mask.gather3div8.si(<4 x i32>, i8*, <4 x i64>, <4 x i1>, i32)
+declare <2 x double> @llvm.x86.avx512.mask.gather3siv2.df(<2 x double>, i8*, <4 x i32>, <2 x i1>, i32)
+declare <2 x i64> @llvm.x86.avx512.mask.gather3siv2.di(<2 x i64>, i8*, <4 x i32>, <2 x i1>, i32)
+declare <4 x double> @llvm.x86.avx512.mask.gather3siv4.df(<4 x double>, i8*, <4 x i32>, <4 x i1>, i32)
+declare <4 x i64> @llvm.x86.avx512.mask.gather3siv4.di(<4 x i64>, i8*, <4 x i32>, <4 x i1>, i32)
+declare <4 x float> @llvm.x86.avx512.mask.gather3siv4.sf(<4 x float>, i8*, <4 x i32>, <4 x i1>, i32)
+declare <4 x i32> @llvm.x86.avx512.mask.gather3siv4.si(<4 x i32>, i8*, <4 x i32>, <4 x i1>, i32)
+declare <8 x float> @llvm.x86.avx512.mask.gather3siv8.sf(<8 x float>, i8*, <8 x i32>, <8 x i1>, i32)
+declare <8 x i32> @llvm.x86.avx512.mask.gather3siv8.si(<8 x i32>, i8*, <8 x i32>, <8 x i1>, i32)