[RISCV] Add the passthru operand for RVV nomask binary intrinsics.

[khchen]

Summary

The goal is support tail and mask policy in RVV builtins.
We focus on IR part first.
If the passthru operand is undef, we use tail agnostic, otherwise
use tail undisturbed.

I will update tests in upstream patch to avoid too much files
changed in phabricator.

[khchen]

Add passthru operand for VSLIDE1UP_VL and VSLIDE1DOWN_VL to handle
i64 scalar in rv32.

[rogfer01]

Looks reasonable to me. Thans @khchen!

[khchen]

Thanks @rogfer01!

So sorry, I fixed two bugs after testing more.
include VPatTiedBinaryNoMaskTU and missed tu pattern
for int_riscv_vrsub.

@rogfer01 could you please help to review the letest patch again, thanks!

[craig.topper]

llvm/lib/Target/RISCV/RISCVISelLowering.cpp

4684

Are we dropping the policy operand for masked intrinsics here?

4720

Capitalize variable name

[khchen]

address Craig's comments.

1. Add masked vslide test to test policy operand.
2. The masked VLSIDE1 would only emit mask undisturbed policy regardless of giving

mask agnostic policy until InsertVSETVLI supports mask agnostic.

[khchen]

Gentle ping.

[craig.topper]

clang/include/clang/Basic/riscv_vector.td

181

Should this be HasNoMaskPassThru rather than Policy?

llvm/lib/Target/RISCV/RISCVISelLowering.cpp

4733

VSLDIE1 -> VSLIDE1

Fix the clang-format warning

4760

I don't think you can assert the policy here. The mask can be folded to undef without the user knowing.

llvm/test/CodeGen/RISCV/rvv/masked-vslide1down-rv32.ll

56

untill -> until

78

untill -> until

[khchen]

address Craig's comments. Thanks!!

[craig.topper]

LGTM other than what I think is an unnecessary include.

llvm/lib/Target/RISCV/RISCVISelLowering.cpp

15

This file is included in RISCV.h can we remove this change?

[eopXD]

clang/include/clang/Basic/riscv_vector.td

180

Nit: V-spec prefers to call instructions without masking operands "unmasked instructions"? Maybe a RFC patch later?

180

I mean an NFC patch.

clang/include/clang/Basic/riscv_vector.td

class RVVBuiltin<string suffix, string prototype, string type_range,
 
   // If HasMask, this flag states that this builtin has a maskedoff operand. It
   // is always the first operand in builtin and IR intrinsic.
   bit HasMaskedOffOperand = true;
 
   // This builtin has a granted vector length parameter.
   bit HasVL = true;
 
-  // There are several cases for specifying tail policy.
-  // 1. Add tail policy argument to masked intrinsics. It may have the maskedoff
-  //    argument or not.
-  //  * Have the maskedoff argument: (HasPolicy = true, HasMaskedOffOperand = true)
-  //    Ex: vadd_vv_i8m1_mt(mask, maskedoff, op1, op2, vl, ta);
-  //  * Do not have the maskedoff argument: (HasPolicy = true, HasMaskedOffOperand = false)
-  //    Ex: vmacc_vv_i8m1_mt(mask, vd, vs1, vs2, vl, ta);
-  // 2. Add dest argument for no mask intrinsics. (TODO)
-  //    Ex: vmv_v_x_i8m1_t(dest, src, vl);
-  // 3. Always tail agnostic. (HasPolicy = false)
-  //    Ex: vmseq_vv_i8m1_b8_m(mask, maskedoff, op1, op2, vl);
-  // The tail policy argument is located at the last position.
+  // The masked intrinsic IR have the policy operand.
+  // The policy argument is located at the last position.
   bit HasPolicy = true;
 
+  // The nomask intrinsic IR have the passthru operand.
+  bit HasNoMaskPolicy = false;
+
   // This builtin supports non-masked function overloading api.
   // All masked operations support overloading api.
   bit HasNoMaskedOverloaded = true;
 
   // This builtin is valid for the given Log2LMULs.
   list<int> Log2LMUL = [0, 1, 2, 3, -1, -2, -3];
 
   // Manual code in clang codegen riscv_vector_builtin_cg.inc
let Name = NAME,
       IRNameMask = IR # "_mask",
       ManualCodegen = [{
       {
         // op1, vl
         IntrinsicTypes = {ResultType,
                           cast<llvm::VectorType>(ResultType)->getElementType(),
                           Ops[1]->getType()};
         Ops.insert(Ops.begin() + 1, llvm::Constant::getNullValue(IntrinsicTypes[1]));
+        // insert undef passthru
+        Ops.insert(Ops.begin(), llvm::UndefValue::get(ResultType));
         break;
       }
       }],
       ManualCodegenMask = [{
       {
         std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         Ops.push_back(ConstantInt::get(Ops.back()->getType(), TAIL_UNDISTURBED));
         // maskedoff, op1, mask, vl
let Name = NAME,
       ManualCodegen = [{
       {
         // op1, vl
         IntrinsicTypes = {ResultType,
                           cast<llvm::VectorType>(ResultType)->getElementType(),
                           Ops[1]->getType()};
         Ops.insert(Ops.begin() + 1,
                    llvm::Constant::getAllOnesValue(IntrinsicTypes[1]));
+        // insert undef passthru
+        Ops.insert(Ops.begin(), llvm::UndefValue::get(ResultType));
         break;
       }
       }],
       ManualCodegenMask = [{
       {
         std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         Ops.push_back(ConstantInt::get(Ops.back()->getType(), TAIL_UNDISTURBED));
         // maskedoff, op1, mask, vl
let Name = NAME,
       IRName = IR,
       IRNameMask = IR # "_mask",
       ManualCodegen = [{
       {
         // op1, vl
         IntrinsicTypes = {ResultType,
                           Ops[0]->getType(), Ops[1]->getType()};
         Ops.insert(Ops.begin() + 1, Ops[0]);
+        // insert undef passthru
+        Ops.insert(Ops.begin(), llvm::UndefValue::get(ResultType));
         break;
       }
       }],
       ManualCodegenMask = [{
       {
         std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         Ops.push_back(ConstantInt::get(Ops.back()->getType(), TAIL_UNDISTURBED));
         // maskedoff, op1, mask, vl
let Name = NAME,
       ManualCodegen = [{
       {
         // op1, vl
         IntrinsicTypes = {ResultType,
                           Ops[0]->getType(),
                           cast<llvm::VectorType>(Ops[0]->getType())->getElementType(),
                           Ops[1]->getType()};
         Ops.insert(Ops.begin() + 1, llvm::Constant::getNullValue(IntrinsicTypes[2]));
+        // insert undef passthru
+        Ops.insert(Ops.begin(), llvm::UndefValue::get(ResultType));
         break;
       }
       }],
       ManualCodegenMask = [{
       {
         std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         Ops.push_back(ConstantInt::get(Ops.back()->getType(), TAIL_UNDISTURBED));
         // maskedoff, op1, mask, vl
let Name = NAME,
       ManualCodegen = [{
       {
         // op1, vl
         IntrinsicTypes = {ResultType,
                           Ops[0]->getType(),
                           Ops[1]->getType(),
                           Ops[1]->getType()};
         Ops.insert(Ops.begin() + 1, llvm::Constant::getNullValue(IntrinsicTypes[2]));
+        // insert undef passthru
+        Ops.insert(Ops.begin(), llvm::UndefValue::get(ResultType));
         break;
       }
       }],
       ManualCodegenMask = [{
       {
         std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);
         Ops.push_back(ConstantInt::get(Ops.back()->getType(), TAIL_UNDISTURBED));
         // maskedoff, op1, mask, vl

 defm : RVVIndexedSegLoad<"vloxseg">;
 defm : RVVUnitStridedSegStore<"vsseg">;
 defm : RVVStridedSegStore<"vssseg">;
 defm : RVVIndexedSegStore<"vsuxseg">;
 defm : RVVIndexedSegStore<"vsoxseg">;
 
 // 12. Vector Integer Arithmetic Instructions
 // 12.1. Vector Single-Width Integer Add and Subtract
+let HasNoMaskPolicy = true in {
 defm vadd : RVVIntBinBuiltinSet;
 defm vsub : RVVIntBinBuiltinSet;
 defm vrsub : RVVOutOp1BuiltinSet<"vrsub", "csil",
                                  [["vx", "v", "vve"],
                                   ["vx", "Uv", "UvUvUe"]]>;
+}
 defm vneg_v : RVVPseudoUnaryBuiltin<"vrsub", "csil">;
 
 // 12.2. Vector Widening Integer Add/Subtract
 // Widening unsigned integer add/subtract, 2*SEW = SEW +/- SEW
+let HasNoMaskPolicy = true in {
 defm vwaddu : RVVUnsignedWidenBinBuiltinSet;
 defm vwsubu : RVVUnsignedWidenBinBuiltinSet;
 // Widening signed integer add/subtract, 2*SEW = SEW +/- SEW
 defm vwadd : RVVSignedWidenBinBuiltinSet;
 defm vwsub : RVVSignedWidenBinBuiltinSet;
 // Widening unsigned integer add/subtract, 2*SEW = 2*SEW +/- SEW
 defm vwaddu : RVVUnsignedWidenOp0BinBuiltinSet;
 defm vwsubu : RVVUnsignedWidenOp0BinBuiltinSet;
 // Widening signed integer add/subtract, 2*SEW = 2*SEW +/- SEW
 defm vwadd : RVVSignedWidenOp0BinBuiltinSet;
 defm vwsub : RVVSignedWidenOp0BinBuiltinSet;
+}
 defm vwcvtu_x_x_v : RVVPseudoVWCVTBuiltin<"vwaddu", "vwcvtu_x", "csi",
                                           [["Uw", "UwUv"]]>;
 defm vwcvt_x_x_v : RVVPseudoVWCVTBuiltin<"vwadd", "vwcvt_x", "csi",
                                          [["w", "wv"]]>;
 
 // 12.3. Vector Integer Extension
 let Log2LMUL = [-3, -2, -1, 0, 1, 2] in {
   def vsext_vf2 : RVVIntExt<"vsext", "w", "wv", "csi">;
let HasMask = false, HasPolicy = false in {
   defm vmadc : RVVCarryOutInBuiltinSet<"vmadc_carry_in">;
   defm vmadc : RVVIntMaskOutBuiltinSet;
   defm vsbc : RVVCarryinBuiltinSet;
   defm vmsbc : RVVCarryOutInBuiltinSet<"vmsbc_borrow_in">;
   defm vmsbc : RVVIntMaskOutBuiltinSet;
 }
 
 // 12.5. Vector Bitwise Logical Instructions
+let HasNoMaskPolicy = true in {
 defm vand : RVVIntBinBuiltinSet;
 defm vxor : RVVIntBinBuiltinSet;
 defm vor : RVVIntBinBuiltinSet;
+}
 defm vnot_v : RVVPseudoVNotBuiltin<"vxor", "csil">;
 
 // 12.6. Vector Single-Width Bit Shift Instructions
+let HasNoMaskPolicy = true in {
 defm vsll : RVVShiftBuiltinSet;
 defm vsrl : RVVUnsignedShiftBuiltinSet;
 defm vsra : RVVSignedShiftBuiltinSet;
 
 // 12.7. Vector Narrowing Integer Right Shift Instructions
 defm vnsrl : RVVUnsignedNShiftBuiltinSet;
 defm vnsra : RVVSignedNShiftBuiltinSet;
+}
 defm vncvt_x_x_w : RVVPseudoVNCVTBuiltin<"vnsrl", "vncvt_x", "csi",
                                          [["v", "vw"],
                                           ["Uv", "UvUw"]]>;
 
 // 12.8. Vector Integer Comparison Instructions
 let HasPolicy = false in {
 defm vmseq : RVVIntMaskOutBuiltinSet;
 defm vmsne : RVVIntMaskOutBuiltinSet;
 defm vmsltu : RVVUnsignedMaskOutBuiltinSet;
 defm vmslt : RVVSignedMaskOutBuiltinSet;
 defm vmsleu : RVVUnsignedMaskOutBuiltinSet;
 defm vmsle : RVVSignedMaskOutBuiltinSet;
 defm vmsgtu : RVVUnsignedMaskOutBuiltinSet;
 defm vmsgt : RVVSignedMaskOutBuiltinSet;
 defm vmsgeu : RVVUnsignedMaskOutBuiltinSet;
 defm vmsge : RVVSignedMaskOutBuiltinSet;
 }
 
 // 12.9. Vector Integer Min/Max Instructions
+let HasNoMaskPolicy = true in {
 defm vminu : RVVUnsignedBinBuiltinSet;
 defm vmin : RVVSignedBinBuiltinSet;
 defm vmaxu : RVVUnsignedBinBuiltinSet;
 defm vmax : RVVSignedBinBuiltinSet;
 
 // 12.10. Vector Single-Width Integer Multiply Instructions
 defm vmul : RVVIntBinBuiltinSet;
 let RequiredFeatures = ["FullMultiply"] in {
 defm vmulh : RVVSignedBinBuiltinSet;
 defm vmulhu : RVVUnsignedBinBuiltinSet;
 defm vmulhsu : RVVOutOp1BuiltinSet<"vmulhsu", "csil",
                                    [["vv", "v", "vvUv"],
                                     ["vx", "v", "vvUe"]]>;
 }
 
 // 12.11. Vector Integer Divide Instructions
 defm vdivu : RVVUnsignedBinBuiltinSet;
 defm vdiv : RVVSignedBinBuiltinSet;
 defm vremu : RVVUnsignedBinBuiltinSet;
 defm vrem : RVVSignedBinBuiltinSet;
+}
 
 // 12.12. Vector Widening Integer Multiply Instructions
-let Log2LMUL = [-3, -2, -1, 0, 1, 2] in {
+let Log2LMUL = [-3, -2, -1, 0, 1, 2], HasNoMaskPolicy = true in {
 defm vwmul : RVVOutOp0Op1BuiltinSet<"vwmul", "csi",
                                     [["vv", "w", "wvv"],
                                      ["vx", "w", "wve"]]>;
 defm vwmulu : RVVOutOp0Op1BuiltinSet<"vwmulu", "csi",
                                      [["vv", "Uw", "UwUvUv"],
                                       ["vx", "Uw", "UwUvUe"]]>;
 defm vwmulsu : RVVOutOp0Op1BuiltinSet<"vwmulsu", "csi",
                                       [["vv", "w", "wvUv"],
let HasMask = false, HasPolicy = false in {
   let HasNoMaskedOverloaded = false in
     defm vmv_v : RVVOutBuiltinSet<"vmv_v_x", "csil",
                                    [["x", "v", "ve"],
                                     ["x", "Uv", "UvUe"]]>;
 }
 
 // 13. Vector Fixed-Point Arithmetic Instructions
 // 13.1. Vector Single-Width Saturating Add and Subtract
+let HasNoMaskPolicy = true in {
 defm vsaddu : RVVUnsignedBinBuiltinSet;
 defm vsadd : RVVSignedBinBuiltinSet;
 defm vssubu : RVVUnsignedBinBuiltinSet;
 defm vssub : RVVSignedBinBuiltinSet;
 
 // 13.2. Vector Single-Width Averaging Add and Subtract
 defm vaaddu : RVVUnsignedBinBuiltinSet;
 defm vaadd : RVVSignedBinBuiltinSet;

 defm vfrdiv : RVVFloatingBinVFBuiltinSet;
 
 // 14.5. Vector Widening Floating-Point Multiply
 let Log2LMUL = [-2, -1, 0, 1, 2] in {
   defm vfwmul : RVVOutOp0Op1BuiltinSet<"vfwmul", "xf",
                                        [["vv", "w", "wvv"],
                                         ["vf", "w", "wve"]]>;
 }
+}
 
 // 14.6. Vector Single-Width Floating-Point Fused Multiply-Add Instructions
 let HasPolicy = false in {
 defm vfmacc  : RVVFloatingTerBuiltinSet;
 defm vfnmacc : RVVFloatingTerBuiltinSet;
 defm vfmsac  : RVVFloatingTerBuiltinSet;
 defm vfnmsac : RVVFloatingTerBuiltinSet;
 defm vfmadd  : RVVFloatingTerBuiltinSet;

 
 // 14.9. Vector Floating-Point Reciprocal Square-Root Estimate Instruction
 def vfrsqrt7 : RVVFloatingUnaryVVBuiltin;
 
 // 14.10. Vector Floating-Point Reciprocal Estimate Instruction
 def vfrec7 : RVVFloatingUnaryVVBuiltin;
 
 // 14.11. Vector Floating-Point MIN/MAX Instructions
+let HasNoMaskPolicy = true in {
 defm vfmin : RVVFloatingBinBuiltinSet;
 defm vfmax : RVVFloatingBinBuiltinSet;
 
 // 14.12. Vector Floating-Point Sign-Injection Instructions
 defm vfsgnj  : RVVFloatingBinBuiltinSet;
 defm vfsgnjn : RVVFloatingBinBuiltinSet;
 defm vfsgnjx : RVVFloatingBinBuiltinSet;
+}
 defm vfneg_v : RVVPseudoVFUnaryBuiltin<"vfsgnjn", "xfd">;
 defm vfabs_v : RVVPseudoVFUnaryBuiltin<"vfsgnjx", "xfd">;
 
 // 14.13. Vector Floating-Point Compare Instructions
 let HasPolicy = false in {
 defm vmfeq : RVVFloatingMaskOutBuiltinSet;
 defm vmfne : RVVFloatingMaskOutBuiltinSet;
 defm vmflt : RVVFloatingMaskOutBuiltinSet;

 let HasPolicy = false in {
 // 17.3.1. Vector Slideup Instructions
 defm vslideup   : RVVSlideBuiltinSet;
 // 17.3.2. Vector Slidedown Instructions
 defm vslidedown : RVVSlideBuiltinSet;
 }
 
 // 17.3.3. Vector Slide1up Instructions
+let HasNoMaskPolicy = true in {
 defm vslide1up : RVVSlideOneBuiltinSet;
 defm vfslide1up : RVVFloatingBinVFBuiltinSet;
 
 // 17.3.4. Vector Slide1down Instruction
 defm vslide1down : RVVSlideOneBuiltinSet;
 defm vfslide1down : RVVFloatingBinVFBuiltinSet;
 
 // 17.4. Vector Register Gather Instructions
 // signed and floating type
 defm vrgather : RVVOutBuiltinSet<"vrgather_vv", "csilxfd",
                                  [["vv", "v", "vvUv"]]>;
 defm vrgather : RVVOutBuiltinSet<"vrgather_vx", "csilxfd",
                                  [["vx", "v", "vvz"]]>;
 defm vrgatherei16 : RVVOutBuiltinSet<"vrgatherei16_vv", "csilxfd",
                                      [["vv", "v", "vv(Log2EEW:4)Uv"]]>;
 // unsigned type
 defm vrgather : RVVOutBuiltinSet<"vrgather_vv", "csil",
                                  [["vv", "Uv", "UvUvUv"]]>;
 defm vrgather : RVVOutBuiltinSet<"vrgather_vx", "csil",
                                  [["vx", "Uv", "UvUvz"]]>;
 defm vrgatherei16 : RVVOutBuiltinSet<"vrgatherei16_vv", "csil",
                                      [["vv", "Uv", "UvUv(Log2EEW:4)Uv"]]>;
+}
 
 // 17.5. Vector Compress Instruction
 let HasMask = false, HasPolicy = false,
     ManualCodegen = [{
       std::rotate(Ops.begin(), Ops.begin() + 1, Ops.begin() + 3);
       IntrinsicTypes = {ResultType, Ops[3]->getType()};
     }] in {
   // signed and floating type

clang/utils/TableGen/RISCVVEmitter.cpp

 private:
   std::string BuiltinName; // Builtin name
   std::string Name;        // C intrinsic name.
   std::string MangledName;
   std::string IRName;
   bool IsMask;
   bool HasVL;
   bool HasPolicy;
+  bool HasNoMaskPolicy;
   bool HasNoMaskedOverloaded;
   bool HasAutoDef; // There is automiatic definition in header
   std::string ManualCodegen;
   RVVTypePtr OutputType; // Builtin output type
   RVVTypes InputTypes;   // Builtin input types
   // The types we use to obtain the specific LLVM intrinsic. They are index of
   // InputTypes. -1 means the return type.
   std::vector<int64_t> IntrinsicTypes;
   RISCVPredefinedMacroT RISCVPredefinedMacros = 0;
   unsigned NF = 1;
 
 public:
   RVVIntrinsic(StringRef Name, StringRef Suffix, StringRef MangledName,
                StringRef MangledSuffix, StringRef IRName, bool IsMask,
                bool HasMaskedOffOperand, bool HasVL, bool HasPolicy,
-               bool HasNoMaskedOverloaded, bool HasAutoDef,
-               StringRef ManualCodegen, const RVVTypes &Types,
+               bool HasNoMaskPolicy, bool HasNoMaskedOverloaded,
+               bool HasAutoDef, StringRef ManualCodegen, const RVVTypes &Types,
                const std::vector<int64_t> &IntrinsicTypes,
                const std::vector<StringRef> &RequiredFeatures, unsigned NF);
   ~RVVIntrinsic() = default;
 
   StringRef getBuiltinName() const { return BuiltinName; }
   StringRef getName() const { return Name; }
   StringRef getMangledName() const { return MangledName; }
   bool hasVL() const { return HasVL; }
   bool hasPolicy() const { return HasPolicy; }
+  bool hasNoMaskPolicy() const { return HasNoMaskPolicy; }
   bool hasNoMaskedOverloaded() const { return HasNoMaskedOverloaded; }
   bool hasManualCodegen() const { return !ManualCodegen.empty(); }
   bool hasAutoDef() const { return HasAutoDef; }
   bool isMask() const { return IsMask; }
   StringRef getIRName() const { return IRName; }
   StringRef getManualCodegen() const { return ManualCodegen; }
   RISCVPredefinedMacroT getRISCVPredefinedMacros() const {
     return RISCVPredefinedMacros;
default:
       PrintFatalError("Illegal non-primitive type transformer!");
     }
   }
 }
 
 //===----------------------------------------------------------------------===//
 // RVVIntrinsic implementation
 //===----------------------------------------------------------------------===//
-RVVIntrinsic::RVVIntrinsic(StringRef NewName, StringRef Suffix,
-                           StringRef NewMangledName, StringRef MangledSuffix,
-                           StringRef IRName, bool IsMask,
-                           bool HasMaskedOffOperand, bool HasVL, bool HasPolicy,
-                           bool HasNoMaskedOverloaded, bool HasAutoDef,
-                           StringRef ManualCodegen, const RVVTypes &OutInTypes,
-                           const std::vector<int64_t> &NewIntrinsicTypes,
-                           const std::vector<StringRef> &RequiredFeatures,
-                           unsigned NF)
+RVVIntrinsic::RVVIntrinsic(
+    StringRef NewName, StringRef Suffix, StringRef NewMangledName,
+    StringRef MangledSuffix, StringRef IRName, bool IsMask,
+    bool HasMaskedOffOperand, bool HasVL, bool HasPolicy, bool HasNoMaskPolicy,
+    bool HasNoMaskedOverloaded, bool HasAutoDef, StringRef ManualCodegen,
+    const RVVTypes &OutInTypes, const std::vector<int64_t> &NewIntrinsicTypes,
+    const std::vector<StringRef> &RequiredFeatures, unsigned NF)
     : IRName(IRName), IsMask(IsMask), HasVL(HasVL), HasPolicy(HasPolicy),
+      HasNoMaskPolicy(HasNoMaskPolicy),
       HasNoMaskedOverloaded(HasNoMaskedOverloaded), HasAutoDef(HasAutoDef),
       ManualCodegen(ManualCodegen.str()), NF(NF) {
 
   // Init BuiltinName, Name and MangledName
   BuiltinName = NewName.str();
   Name = BuiltinName;
   if (NewMangledName.empty())
     MangledName = NewName.split("_").first.str();
RVVIntrinsic::RVVIntrinsic(
 
   // Init OutputType and InputTypes
   OutputType = OutInTypes[0];
   InputTypes.assign(OutInTypes.begin() + 1, OutInTypes.end());
 
   // IntrinsicTypes is nonmasked version index. Need to update it
   // if there is maskedoff operand (It is always in first operand).
   IntrinsicTypes = NewIntrinsicTypes;
-  if (IsMask && HasMaskedOffOperand) {
+  if ((IsMask && HasMaskedOffOperand) || (!IsMask && HasNoMaskPolicy)) {
     for (auto &I : IntrinsicTypes) {
       if (I >= 0)
         I += NF;
     }
   }
 }
 
 std::string RVVIntrinsic::getBuiltinTypeStr() const {
if (isMask()) {
     if (hasVL()) {
       OS << "  std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end() - 1);\n";
       if (hasPolicy())
         OS << "  Ops.push_back(ConstantInt::get(Ops.back()->getType(),"
                                " TAIL_UNDISTURBED));\n";
     } else {
       OS << "  std::rotate(Ops.begin(), Ops.begin() + 1, Ops.end());\n";
     }
+  } else if (hasNoMaskPolicy()) {
+    OS << "  Ops.push_back(llvm::UndefValue::get(ResultType));\n";
+    OS << "  std::rotate(Ops.rbegin(), Ops.rbegin() + 1,  Ops.rend());\n";
   }
 
   OS << "  IntrinsicTypes = {";
   ListSeparator LS;
   for (const auto &Idx : IntrinsicTypes) {
     if (Idx == -1)
       OS << LS << "ResultType";
     else
for (auto &Def : Defs) {
     else if (P.first->second->getNF() != Def->getNF())
       PrintFatalError("Builtin with same name has different NF");
     else if (P.first->second->isMask() != Def->isMask())
       PrintFatalError("Builtin with same name has different isMask");
     else if (P.first->second->hasVL() != Def->hasVL())
       PrintFatalError("Builtin with same name has different HasPolicy");
     else if (P.first->second->hasPolicy() != Def->hasPolicy())
       PrintFatalError("Builtin with same name has different HasPolicy");
+    else if (P.first->second->hasNoMaskPolicy() != Def->hasNoMaskPolicy())
+      PrintFatalError("Builtin with same name has different HasNoMaskPolicy");
     else if (P.first->second->getIntrinsicTypes() != Def->getIntrinsicTypes())
       PrintFatalError("Builtin with same name has different IntrinsicTypes");
   }
   Defs.back()->emitCodeGenSwitchBody(OS);
   OS << "\n";
 }
 
 void RVVEmitter::parsePrototypes(StringRef Prototypes,
for (auto *R : RV) {
     StringRef MangledName = R->getValueAsString("MangledName");
     StringRef MangledSuffixProto = R->getValueAsString("MangledSuffix");
     StringRef Prototypes = R->getValueAsString("Prototype");
     StringRef TypeRange = R->getValueAsString("TypeRange");
     bool HasMask = R->getValueAsBit("HasMask");
     bool HasMaskedOffOperand = R->getValueAsBit("HasMaskedOffOperand");
     bool HasVL = R->getValueAsBit("HasVL");
     bool HasPolicy = R->getValueAsBit("HasPolicy");
+    bool HasNoMaskPolicy = R->getValueAsBit("HasNoMaskPolicy");
     bool HasNoMaskedOverloaded = R->getValueAsBit("HasNoMaskedOverloaded");
     std::vector<int64_t> Log2LMULList = R->getValueAsListOfInts("Log2LMUL");
     StringRef ManualCodegen = R->getValueAsString("ManualCodegen");
     StringRef ManualCodegenMask = R->getValueAsString("ManualCodegenMask");
     std::vector<int64_t> IntrinsicTypes =
         R->getValueAsListOfInts("IntrinsicTypes");
     std::vector<StringRef> RequiredFeatures =
         R->getValueAsListOfStrings("RequiredFeatures");
for (char I : TypeRange) {
           continue;
 
         auto SuffixStr = getSuffixStr(I, Log2LMUL, SuffixProto);
         auto MangledSuffixStr = getSuffixStr(I, Log2LMUL, MangledSuffixProto);
         // Create a non-mask intrinsic
         Out.push_back(std::make_unique<RVVIntrinsic>(
             Name, SuffixStr, MangledName, MangledSuffixStr, IRName,
             /*IsMask=*/false, /*HasMaskedOffOperand=*/false, HasVL, HasPolicy,
-            HasNoMaskedOverloaded, HasAutoDef, ManualCodegen, Types.getValue(),
-            IntrinsicTypes, RequiredFeatures, NF));
+            HasNoMaskPolicy, HasNoMaskedOverloaded, HasAutoDef, ManualCodegen,
+            Types.getValue(), IntrinsicTypes, RequiredFeatures, NF));
         if (HasMask) {
           // Create a mask intrinsic
           Optional<RVVTypes> MaskTypes =
               computeTypes(I, Log2LMUL, NF, ProtoMaskSeq);
           Out.push_back(std::make_unique<RVVIntrinsic>(
               Name, SuffixStr, MangledName, MangledSuffixStr, IRNameMask,
               /*IsMask=*/true, HasMaskedOffOperand, HasVL, HasPolicy,
-              HasNoMaskedOverloaded, HasAutoDef, ManualCodegenMask,
-              MaskTypes.getValue(), IntrinsicTypes, RequiredFeatures, NF));
+              HasNoMaskPolicy, HasNoMaskedOverloaded, HasAutoDef,
+              ManualCodegenMask, MaskTypes.getValue(), IntrinsicTypes,
+              RequiredFeatures, NF));
         }
       } // end for Log2LMULList
     }   // end for TypeRange
   }
 }
 
 void RVVEmitter::createRVVHeaders(raw_ostream &OS) {
   std::vector<Record *> RVVHeaders =

llvm/include/llvm/IR/IntrinsicsRISCV.td

let TargetPrefix = "riscv" in {
   // Input: (vector_in, vector_in, vl)
   class RISCVBinaryAAANoMask
         : Intrinsic<[llvm_anyvector_ty],
                     [LLVMMatchType<0>, LLVMMatchType<0>, llvm_anyint_ty],
                     [IntrNoMem]>, RISCVVIntrinsic {
     let VLOperand = 2;
   }
   // For destination vector type is the same as first and second source vector.
-  // Input: (vector_in, int_vector_in, vl)
+  // Input: (passthru, vector_in, int_vector_in, vl)
   class RISCVRGatherVVNoMask
         : Intrinsic<[llvm_anyvector_ty],
-                    [LLVMMatchType<0>, LLVMVectorOfBitcastsToInt<0>, llvm_anyint_ty],
+                    [LLVMMatchType<0>, LLVMMatchType<0>,
+                     LLVMVectorOfBitcastsToInt<0>, llvm_anyint_ty],
                     [IntrNoMem]>, RISCVVIntrinsic {
-    let VLOperand = 2;
+    let VLOperand = 3;
   }
   // For destination vector type is the same as first and second source vector.
   // Input: (vector_in, vector_in, int_vector_in, vl, ta)
   class RISCVRGatherVVMask
         : Intrinsic<[llvm_anyvector_ty],
                     [LLVMMatchType<0>, LLVMMatchType<0>, LLVMVectorOfBitcastsToInt<0>,
                      LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>, llvm_anyint_ty,
                      LLVMMatchType<1>],
                     [ImmArg<ArgIndex<5>>, IntrNoMem]>, RISCVVIntrinsic {
     let VLOperand = 4;
   }
-  // Input: (vector_in, int16_vector_in, vl)
+  // Input: (passthru, vector_in, int16_vector_in, vl)
   class RISCVRGatherEI16VVNoMask
         : Intrinsic<[llvm_anyvector_ty],
-                    [LLVMMatchType<0>, LLVMScalarOrSameVectorWidth<0, llvm_i16_ty>,
+                    [LLVMMatchType<0>, LLVMMatchType<0>,
+                     LLVMScalarOrSameVectorWidth<0, llvm_i16_ty>,
                      llvm_anyint_ty],
                     [IntrNoMem]>, RISCVVIntrinsic {
-    let VLOperand = 2;
+    let VLOperand = 3;
   }
   // For destination vector type is the same as first and second source vector.
   // Input: (vector_in, vector_in, int16_vector_in, vl, ta)
   class RISCVRGatherEI16VVMask
         : Intrinsic<[llvm_anyvector_ty],
                     [LLVMMatchType<0>, LLVMMatchType<0>,
                      LLVMScalarOrSameVectorWidth<0, llvm_i16_ty>,
                      LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>, llvm_anyint_ty,
                      LLVMMatchType<1>],
                     [ImmArg<ArgIndex<5>>, IntrNoMem]>, RISCVVIntrinsic {
     let VLOperand = 4;
   }
   // For destination vector type is the same as first source vector, and the
   // second operand is XLen.
-  // Input: (vector_in, xlen_in, vl)
+  // Input: (passthru, vector_in, xlen_in, vl)
   class RISCVGatherVXNoMask
         : Intrinsic<[llvm_anyvector_ty],
-                    [LLVMMatchType<0>, llvm_anyint_ty, LLVMMatchType<1>],
+                    [LLVMMatchType<0>, LLVMMatchType<0>, llvm_anyint_ty, 
+                     LLVMMatchType<1>],
                     [IntrNoMem]>, RISCVVIntrinsic {
-    let VLOperand = 2;
+    let VLOperand = 3;
   }
   // For destination vector type is the same as first source vector (with mask).
   // Second operand is XLen.
   // Input: (maskedoff, vector_in, xlen_in, mask, vl, ta)
   class RISCVGatherVXMask
        : Intrinsic<[llvm_anyvector_ty],
                    [LLVMMatchType<0>, LLVMMatchType<0>, llvm_anyint_ty,
                     LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>, LLVMMatchType<1>,
                     LLVMMatchType<1>],
                    [ImmArg<ArgIndex<5>>, IntrNoMem]>, RISCVVIntrinsic {
     let VLOperand = 4;
   }
   // For destination vector type is the same as first source vector.
-  // Input: (vector_in, vector_in/scalar_in, vl)
+  // Input: (passthru, vector_in, vector_in/scalar_in, vl)
   class RISCVBinaryAAXNoMask
         : Intrinsic<[llvm_anyvector_ty],
-                    [LLVMMatchType<0>, llvm_any_ty, llvm_anyint_ty],
+                    [LLVMMatchType<0>, LLVMMatchType<0>, llvm_any_ty,
+                     llvm_anyint_ty],
                     [IntrNoMem]>, RISCVVIntrinsic {
-    let SplatOperand = 1;
-    let VLOperand = 2;
+    let SplatOperand = 2;
+    let VLOperand = 3;
   }
   // For destination vector type is the same as first source vector (with mask).
   // Input: (maskedoff, vector_in, vector_in/scalar_in, mask, vl, ta)
   class RISCVBinaryAAXMask
        : Intrinsic<[llvm_anyvector_ty],
                    [LLVMMatchType<0>, LLVMMatchType<0>, llvm_any_ty,
                     LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>, llvm_anyint_ty,
                     LLVMMatchType<2>],
                    [ImmArg<ArgIndex<5>>, IntrNoMem]>, RISCVVIntrinsic {
     let SplatOperand = 2;
     let VLOperand = 4;
   }
   // For destination vector type is the same as first source vector. The
   // second source operand must match the destination type or be an XLen scalar.
-  // Input: (vector_in, vector_in/scalar_in, vl)
+  // Input: (passthru, vector_in, vector_in/scalar_in, vl)
   class RISCVBinaryAAShiftNoMask
         : Intrinsic<[llvm_anyvector_ty],
-                    [LLVMMatchType<0>, llvm_any_ty, llvm_anyint_ty],
+                    [LLVMMatchType<0>, LLVMMatchType<0>, llvm_any_ty,
+                     llvm_anyint_ty],
                     [IntrNoMem]>, RISCVVIntrinsic {
-    let VLOperand = 2;
+    let VLOperand = 3;
   }
   // For destination vector type is the same as first source vector (with mask).
   // The second source operand must match the destination type or be an XLen scalar.
   // Input: (maskedoff, vector_in, vector_in/scalar_in, mask, vl, ta)
   class RISCVBinaryAAShiftMask
        : Intrinsic<[llvm_anyvector_ty],
                    [LLVMMatchType<0>, LLVMMatchType<0>, llvm_any_ty,
                     LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>, llvm_anyint_ty,
                     LLVMMatchType<2>],
                    [ImmArg<ArgIndex<5>>, IntrNoMem]>, RISCVVIntrinsic {
     let VLOperand = 4;
   }
   // For destination vector type is NOT the same as first source vector.
-  // Input: (vector_in, vector_in/scalar_in, vl)
+  // Input: (passthru, vector_in, vector_in/scalar_in, vl)
   class RISCVBinaryABXNoMask
         : Intrinsic<[llvm_anyvector_ty],
-                    [llvm_anyvector_ty, llvm_any_ty, llvm_anyint_ty],
+                    [LLVMMatchType<0>, llvm_anyvector_ty, llvm_any_ty,
+                     llvm_anyint_ty],
                     [IntrNoMem]>, RISCVVIntrinsic {
-    let SplatOperand = 1;
-    let VLOperand = 2;
+    let SplatOperand = 2;
+    let VLOperand = 3;
   }
   // For destination vector type is NOT the same as first source vector (with mask).
   // Input: (maskedoff, vector_in, vector_in/scalar_in, mask, vl, ta)
   class RISCVBinaryABXMask
         : Intrinsic<[llvm_anyvector_ty],
                     [LLVMMatchType<0>, llvm_anyvector_ty, llvm_any_ty,
                      LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>, llvm_anyint_ty,
                      LLVMMatchType<3>],
                     [ImmArg<ArgIndex<5>>, IntrNoMem]>, RISCVVIntrinsic {
     let SplatOperand = 2;
     let VLOperand = 4;
   }
   // For destination vector type is NOT the same as first source vector. The
   // second source operand must match the destination type or be an XLen scalar.
-  // Input: (vector_in, vector_in/scalar_in, vl)
+  // Input: (passthru, vector_in, vector_in/scalar_in, vl)
   class RISCVBinaryABShiftNoMask
         : Intrinsic<[llvm_anyvector_ty],
-                    [llvm_anyvector_ty, llvm_any_ty, llvm_anyint_ty],
+                    [LLVMMatchType<0>, llvm_anyvector_ty, llvm_any_ty,
+                     llvm_anyint_ty],
                     [IntrNoMem]>, RISCVVIntrinsic {
-    let VLOperand = 2;
+    let VLOperand = 3;
   }
   // For destination vector type is NOT the same as first source vector (with mask).
   // The second source operand must match the destination type or be an XLen scalar.
   // Input: (maskedoff, vector_in, vector_in/scalar_in, mask, vl, ta)
   class RISCVBinaryABShiftMask
         : Intrinsic<[llvm_anyvector_ty],
                     [LLVMMatchType<0>, llvm_anyvector_ty, llvm_any_ty,
                      LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>, llvm_anyint_ty,
class RISCVClassifyMask
         : Intrinsic<[LLVMVectorOfBitcastsToInt<0>],
                     [LLVMVectorOfBitcastsToInt<0>, llvm_anyvector_ty,
                      LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>, llvm_anyint_ty],
                     [IntrNoMem]>, RISCVVIntrinsic {
     let VLOperand = 3;
   }
   // For Saturating binary operations.
   // The destination vector type is the same as first source vector.
-  // Input: (vector_in, vector_in/scalar_in, vl)
+  // Input: (passthru, vector_in, vector_in/scalar_in, vl)
   class RISCVSaturatingBinaryAAXNoMask
         : Intrinsic<[llvm_anyvector_ty],
-                    [LLVMMatchType<0>, llvm_any_ty, llvm_anyint_ty],
+                    [LLVMMatchType<0>, LLVMMatchType<0>, llvm_any_ty,
+                     llvm_anyint_ty],
                     [IntrNoMem, IntrHasSideEffects]>, RISCVVIntrinsic {
-    let SplatOperand = 1;
-    let VLOperand = 2;
+    let SplatOperand = 2;
+    let VLOperand = 3;
   }
   // For Saturating binary operations with mask.
   // The destination vector type is the same as first source vector.
   // Input: (maskedoff, vector_in, vector_in/scalar_in, mask, vl, ta)
   class RISCVSaturatingBinaryAAXMask
         : Intrinsic<[llvm_anyvector_ty],
                     [LLVMMatchType<0>, LLVMMatchType<0>, llvm_any_ty,
                      LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>, llvm_anyint_ty,
                      LLVMMatchType<2>],
                     [ImmArg<ArgIndex<5>>, IntrNoMem, IntrHasSideEffects]>, RISCVVIntrinsic {
     let SplatOperand = 2;
     let VLOperand = 4;
   }
   // For Saturating binary operations.
   // The destination vector type is the same as first source vector.
   // The second source operand matches the destination type or is an XLen scalar.
-  // Input: (vector_in, vector_in/scalar_in, vl)
+  // Input: (passthru, vector_in, vector_in/scalar_in, vl)
   class RISCVSaturatingBinaryAAShiftNoMask
         : Intrinsic<[llvm_anyvector_ty],
-                    [LLVMMatchType<0>, llvm_any_ty, llvm_anyint_ty],
+                    [LLVMMatchType<0>, LLVMMatchType<0>, llvm_any_ty,
+                     llvm_anyint_ty],
                     [IntrNoMem, IntrHasSideEffects]>, RISCVVIntrinsic {
-    let VLOperand = 2;
+    let VLOperand = 3;
   }
   // For Saturating binary operations with mask.
   // The destination vector type is the same as first source vector.
   // The second source operand matches the destination type or is an XLen scalar.
   // Input: (maskedoff, vector_in, vector_in/scalar_in, mask, vl, ta)
   class RISCVSaturatingBinaryAAShiftMask
         : Intrinsic<[llvm_anyvector_ty],
                     [LLVMMatchType<0>, LLVMMatchType<0>, llvm_any_ty,
                      LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>, llvm_anyint_ty,
                      LLVMMatchType<2>],
                     [ImmArg<ArgIndex<5>>, IntrNoMem, IntrHasSideEffects]>, RISCVVIntrinsic {
     let VLOperand = 4;
   }
   // For Saturating binary operations.
   // The destination vector type is NOT the same as first source vector.
   // The second source operand matches the destination type or is an XLen scalar.
-  // Input: (vector_in, vector_in/scalar_in, vl)
+  // Input: (passthru, vector_in, vector_in/scalar_in, vl)
   class RISCVSaturatingBinaryABShiftNoMask
         : Intrinsic<[llvm_anyvector_ty],
-                    [llvm_anyvector_ty, llvm_any_ty, llvm_anyint_ty],
+                    [LLVMMatchType<0>, llvm_anyvector_ty, llvm_any_ty,
+                     llvm_anyint_ty],
                     [IntrNoMem, IntrHasSideEffects]>, RISCVVIntrinsic {
-    let VLOperand = 2;
+    let VLOperand = 3;
   }
   // For Saturating binary operations with mask.
   // The destination vector type is NOT the same as first source vector (with mask).
   // The second source operand matches the destination type or is an XLen scalar.
   // Input: (maskedoff, vector_in, vector_in/scalar_in, mask, vl, ta)
   class RISCVSaturatingBinaryABShiftMask
         : Intrinsic<[llvm_anyvector_ty],
                     [LLVMMatchType<0>, llvm_anyvector_ty, llvm_any_ty,

llvm/lib/Target/RISCV/MCTargetDesc/RISCVBaseInfo.h

enum VLMUL : uint8_t {
   LMUL_F8,
   LMUL_F4,
   LMUL_F2
 };
 
 enum {
   TAIL_UNDISTURBED = 0,
   TAIL_AGNOSTIC = 1,
+  MASK_AGNOSTIC = 2,
 };
 
 // Helper functions to read TSFlags.
 /// \returns the format of the instruction.
 static inline unsigned getFormat(uint64_t TSFlags) {
   return (TSFlags & InstFormatMask) >> InstFormatShift;
 }
 /// \returns the constraint for the instruction.

llvm/lib/Target/RISCV/RISCVISelLowering.h

enum NodeType : unsigned {
   // mask and VL operand.
   TRUNCATE_VECTOR_VL,
   // Matches the semantics of vslideup/vslidedown. The first operand is the
   // pass-thru operand, the second is the source vector, the third is the
   // XLenVT index (either constant or non-constant), the fourth is the mask
   // and the fifth the VL.
   VSLIDEUP_VL,
   VSLIDEDOWN_VL,
-  // Matches the semantics of vslide1up/slide1down. The first operand is the
-  // source vector, the second is the XLenVT scalar value. The third and fourth
-  // operands are the mask and VL operands.
+  // Matches the semantics of vslide1up/slide1down. The first operand is
+  // passthru operand, the second is source vector, third is the XLenVT scalar
+  // value. The fourth and fifth operands are the mask and VL operands.
   VSLIDE1UP_VL,
   VSLIDE1DOWN_VL,
   // Matches the semantics of the vid.v instruction, with a mask and VL
   // operand.
   VID_VL,
   // Matches the semantics of the vfcnvt.rod function (Convert double-width
   // float to single-width float, rounding towards odd). Takes a double-width
   // float vector and produces a single-width float vector. Also has a mask and

llvm/lib/Target/RISCV/RISCVISelLowering.cpp

 //===-- RISCVISelLowering.cpp - RISCV DAG Lowering Implementation  --------===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 //
 // This file defines the interfaces that RISCV uses to lower LLVM code into a
 // selection DAG.
 //
 //===----------------------------------------------------------------------===//
 
 #include "RISCVISelLowering.h"
+#include "MCTargetDesc/RISCVBaseInfo.h"
 #include "MCTargetDesc/RISCVMatInt.h"
 #include "RISCV.h"
 #include "RISCVMachineFunctionInfo.h"
 #include "RISCVRegisterInfo.h"
 #include "RISCVSubtarget.h"
 #include "RISCVTargetMachine.h"
 #include "llvm/ADT/SmallSet.h"
 #include "llvm/ADT/Statistic.h"
SDValue I32Mask =
         getDefaultScalableVLOps(I32ContainerVT, DL, DAG, Subtarget).first;
     // Limit the active VL to two.
     SDValue InsertI64VL = DAG.getConstant(2, DL, XLenVT);
     // Note: We can't pass a UNDEF to the first VSLIDE1UP_VL since an untied
     // undef doesn't obey the earlyclobber constraint. Just splat a zero value.
     ValInVec = DAG.getNode(RISCVISD::VMV_V_X_VL, DL, I32ContainerVT, Zero,
                            InsertI64VL);
     // First slide in the hi value, then the lo in underneath it.
-    ValInVec = DAG.getNode(RISCVISD::VSLIDE1UP_VL, DL, I32ContainerVT, ValInVec,
-                           ValHi, I32Mask, InsertI64VL);
-    ValInVec = DAG.getNode(RISCVISD::VSLIDE1UP_VL, DL, I32ContainerVT, ValInVec,
-                           ValLo, I32Mask, InsertI64VL);
+    ValInVec = DAG.getNode(RISCVISD::VSLIDE1UP_VL, DL, I32ContainerVT,
+                           DAG.getUNDEF(I32ContainerVT), ValInVec, ValHi,
+                           I32Mask, InsertI64VL);
+    ValInVec = DAG.getNode(RISCVISD::VSLIDE1UP_VL, DL, I32ContainerVT,
+                           DAG.getUNDEF(I32ContainerVT), ValInVec, ValLo,
+                           I32Mask, InsertI64VL);
     // Bitcast back to the right container type.
     ValInVec = DAG.getBitcast(ContainerVT, ValInVec);
   }
 
   // Now that the value is in a vector, slide it into position.
   SDValue InsertVL =
       DAG.getNode(ISD::ADD, DL, XLenVT, Idx, DAG.getConstant(1, DL, XLenVT));
   SDValue Slideup = DAG.getNode(RISCVISD::VSLIDEUP_VL, DL, ContainerVT, Vec,
SDValue SelectCond =
                     DAG.getCondCode(ISD::SETEQ), Mask, VL);
     return DAG.getNode(RISCVISD::VSELECT_VL, DL, VT, SelectCond, SplattedVal,
                        Vec, VL);
   }
   case Intrinsic::riscv_vslide1up:
   case Intrinsic::riscv_vslide1down:
   case Intrinsic::riscv_vslide1up_mask:
   case Intrinsic::riscv_vslide1down_mask: {
     // We need to special case these when the scalar is larger than XLen.
     unsigned NumOps = Op.getNumOperands();
     bool IsMasked = NumOps == 7;
-    unsigned OpOffset = IsMasked ? 1 : 0;
-    SDValue Scalar = Op.getOperand(2 + OpOffset);
+    SDValue Scalar = Op.getOperand(3);
     if (Scalar.getValueType().bitsLE(XLenVT))
       break;
 
     // Splatting a sign extended constant is fine.
     if (auto *CVal = dyn_cast<ConstantSDNode>(Scalar))
       if (isInt<32>(CVal->getSExtValue()))
         break;
 
     MVT VT = Op.getSimpleValueType();
     assert(VT.getVectorElementType() == MVT::i64 &&
            Scalar.getValueType() == MVT::i64 && "Unexpected VTs");
 
     // Convert the vector source to the equivalent nxvXi32 vector.
     MVT I32VT = MVT::getVectorVT(MVT::i32, VT.getVectorElementCount() * 2);
-    SDValue Vec = DAG.getBitcast(I32VT, Op.getOperand(1 + OpOffset));
+    SDValue Vec = DAG.getBitcast(I32VT, Op.getOperand(2));
 
     SDValue ScalarLo = DAG.getNode(ISD::EXTRACT_ELEMENT, DL, MVT::i32, Scalar,
                                    DAG.getConstant(0, DL, XLenVT));
     SDValue ScalarHi = DAG.getNode(ISD::EXTRACT_ELEMENT, DL, MVT::i32, Scalar,
                                    DAG.getConstant(1, DL, XLenVT));
 
     // Double the VL since we halved SEW.
     SDValue VL = getVLOperand(Op);
     SDValue I32VL =
         DAG.getNode(ISD::SHL, DL, XLenVT, VL, DAG.getConstant(1, DL, XLenVT));
 
     MVT I32MaskVT = MVT::getVectorVT(MVT::i1, I32VT.getVectorElementCount());
     SDValue I32Mask = DAG.getNode(RISCVISD::VMSET_VL, DL, I32MaskVT, VL);
 
     // Shift the two scalar parts in using SEW=32 slide1up/slide1down
     // instructions.
-    if (IntNo == Intrinsic::riscv_vslide1up ||
-        IntNo == Intrinsic::riscv_vslide1up_mask) {
-      Vec = DAG.getNode(RISCVISD::VSLIDE1UP_VL, DL, I32VT, Vec, ScalarHi,
-                        I32Mask, I32VL);
-      Vec = DAG.getNode(RISCVISD::VSLIDE1UP_VL, DL, I32VT, Vec, ScalarLo,
-                        I32Mask, I32VL);
-    } else {
-      Vec = DAG.getNode(RISCVISD::VSLIDE1DOWN_VL, DL, I32VT, Vec, ScalarLo,
-                        I32Mask, I32VL);
-      Vec = DAG.getNode(RISCVISD::VSLIDE1DOWN_VL, DL, I32VT, Vec, ScalarHi,
-                        I32Mask, I32VL);
+    SDValue Passthru = DAG.getBitcast(I32VT, Op.getOperand(1));
+    if (!IsMasked) {
+      if (IntNo == Intrinsic::riscv_vslide1up) {
+        Vec = DAG.getNode(RISCVISD::VSLIDE1UP_VL, DL, I32VT, Passthru, Vec,
+                          ScalarHi, I32Mask, I32VL);
+        Vec = DAG.getNode(RISCVISD::VSLIDE1UP_VL, DL, I32VT, Passthru, Vec,
+                          ScalarLo, I32Mask, I32VL);
+      } else {
+        Vec = DAG.getNode(RISCVISD::VSLIDE1DOWN_VL, DL, I32VT, Passthru, Vec,
+                          ScalarLo, I32Mask, I32VL);
+        Vec = DAG.getNode(RISCVISD::VSLIDE1DOWN_VL, DL, I32VT, Passthru, Vec,
+                          ScalarHi, I32Mask, I32VL);
+      }
+    } else {
+      // TODO Those VSLDIE1 could be TAMA because we use vmerge to select maskedoff
+      SDValue Undef = DAG.getUNDEF(I32VT);
+      if (IntNo == Intrinsic::riscv_vslide1up_mask) {
+        Vec = DAG.getNode(RISCVISD::VSLIDE1UP_VL, DL, I32VT, Undef, Vec,
+                          ScalarHi, I32Mask, I32VL);
+        Vec = DAG.getNode(RISCVISD::VSLIDE1UP_VL, DL, I32VT, Undef, Vec,
+                          ScalarLo, I32Mask, I32VL);
+      } else {
+        Vec = DAG.getNode(RISCVISD::VSLIDE1DOWN_VL, DL, I32VT, Undef, Vec,
+                          ScalarLo, I32Mask, I32VL);
+        Vec = DAG.getNode(RISCVISD::VSLIDE1DOWN_VL, DL, I32VT, Undef, Vec,
+                          ScalarHi, I32Mask, I32VL);
+      }
     }
 
     // Convert back to nxvXi64.
     Vec = DAG.getBitcast(VT, Vec);
 
     if (!IsMasked)
       return Vec;
-
     // Apply mask after the operation.
     SDValue Mask = Op.getOperand(NumOps - 3);
     SDValue MaskedOff = Op.getOperand(1);
-    return DAG.getNode(RISCVISD::VSELECT_VL, DL, VT, Mask, Vec, MaskedOff, VL);
+    // Assume Policy operand is the last operand.
+    uint64_t Policy = Op.getConstantOperandVal(NumOps - 1);
+    // TAMA don't need to select value from maskedoff.
+    if (MaskedOff.isUndef()) {
+      assert(Policy == (RISCVII::TAIL_AGNOSTIC | RISCVII::MASK_AGNOSTIC) &&
+             "Invalid policy, undef maskedoff need to have tail and mask "
+             "agnostic");
+      return Vec;
+    }
+    // TAMU
+    if (Policy == RISCVII::TAIL_AGNOSTIC)
+      return DAG.getNode(RISCVISD::VSELECT_VL, DL, VT, Mask, Vec, MaskedOff,
+                         VL);
+    // TUMA or TUMU: Currently we always emit tumu policy regardless of tuma.
+    // It's fine because vmerge does not care mask policy.
+    return DAG.getNode(RISCVISD::VP_MERGE_VL, DL, VT, Mask, Vec, MaskedOff, VL);
   }
   }
 
   return lowerVectorIntrinsicSplats(Op, DAG, Subtarget);
 }
 
 SDValue RISCVTargetLowering::LowerINTRINSIC_W_CHAIN(SDValue Op,
                                                     SelectionDAG &DAG) const {

llvm/lib/Target/RISCV/RISCVInstrInfoVPseudos.td

class VPseudoBinaryNoMask<VReg RetClass,
   let hasSideEffects = 0;
   let Constraints = Constraint;
   let HasVLOp = 1;
   let HasSEWOp = 1;
   let HasDummyMask = 1;
   let BaseInstr = !cast<Instruction>(PseudoToVInst<NAME>.VInst);
 }
 
+class VPseudoBinaryNoMaskTU<VReg RetClass,
+                            VReg Op1Class,
+                            DAGOperand Op2Class,
+                            string Constraint> :
+        Pseudo<(outs RetClass:$rd),
+               (ins RetClass:$merge, Op1Class:$rs2, Op2Class:$rs1, AVL:$vl, ixlenimm:$sew), []>,
+        RISCVVPseudo {
+  let mayLoad = 0;
+  let mayStore = 0;
+  let hasSideEffects = 0;
+  let Constraints = Join<[Constraint, "$rd = $merge"], ",">.ret;
+  let HasVLOp = 1;
+  let HasSEWOp = 1;
+  let HasDummyMask = 1;
+  let HasMergeOp = 1;
+  let BaseInstr = !cast<Instruction>(PseudoToVInst<NAME>.VInst);
+}
+
 // Special version of VPseudoBinaryNoMask where we pretend the first source is
 // tied to the destination.
 // This allows maskedoff and rs2 to be the same register.
 class VPseudoTiedBinaryNoMask<VReg RetClass,
                               DAGOperand Op2Class,
                               string Constraint> :
         Pseudo<(outs RetClass:$rd),
                (ins RetClass:$rs2, Op2Class:$rs1, AVL:$vl, ixlenimm:$sew), []>,
         RISCVVPseudo {
   let mayLoad = 0;
   let mayStore = 0;
   let hasSideEffects = 0;
   let Constraints = Join<[Constraint, "$rd = $rs2"], ",">.ret;
   let HasVLOp = 1;
   let HasSEWOp = 1;
   let HasDummyMask = 1;
   let ForceTailAgnostic = 1;
   let isConvertibleToThreeAddress = 1;
   let BaseInstr = !cast<Instruction>(PseudoToVInst<NAME>.VInst);
 }
 
+class VPseudoTiedBinaryNoMaskTU<VReg RetClass,
+                                DAGOperand Op2Class,
+                                string Constraint> :
+        Pseudo<(outs RetClass:$rd),
+               (ins RetClass:$merge,
+                    Op2Class:$rs1,
+                    AVL:$vl, ixlenimm:$sew), []>,
+        RISCVVPseudo {
+  let mayLoad = 0;
+  let mayStore = 0;
+  let hasSideEffects = 0;
+  let Constraints = Join<[Constraint, "$rd = $merge"], ",">.ret;
+  let HasVLOp = 1;
+  let HasSEWOp = 1;
+  let HasMergeOp = 0; // Merge is also rs2.
+  let HasDummyMask = 1;
+  let BaseInstr = !cast<Instruction>(PseudoToVInst<NAME>.VInst);
+}
+
 class VPseudoIStoreNoMask<VReg StClass, VReg IdxClass, int EEW, bits<3> LMUL,
                           bit Ordered>:
       Pseudo<(outs),
               (ins StClass:$rd, GPR:$rs1, IdxClass:$rs2, AVL:$vl, ixlenimm:$sew),[]>,
       RISCVVPseudo,
       RISCVVSX</*Masked*/0, Ordered, log2<EEW>.val, VLMul, LMUL> {
   let mayLoad = 0;
   let mayStore = 1;

 multiclass VPseudoBinary<VReg RetClass,
                          VReg Op1Class,
                          DAGOperand Op2Class,
                          LMULInfo MInfo,
                          string Constraint = ""> {
   let VLMul = MInfo.value in {
     def "_" # MInfo.MX : VPseudoBinaryNoMask<RetClass, Op1Class, Op2Class,
                                              Constraint>;
+    def "_" # MInfo.MX # "_TU" : VPseudoBinaryNoMaskTU<RetClass, Op1Class, Op2Class,
+                                                       Constraint>;
     def "_" # MInfo.MX # "_MASK" : VPseudoBinaryMaskTA<RetClass, Op1Class, Op2Class,
                                                        Constraint>,
                                    RISCVMaskedPseudo</*MaskOpIdx*/ 3>;
   }
 }
 
 multiclass VPseudoBinaryM<VReg RetClass,
                           VReg Op1Class,
multiclass VPseudoBinaryEmul<VReg RetClass,
                              VReg Op1Class,
                              DAGOperand Op2Class,
                              LMULInfo lmul,
                              LMULInfo emul,
                              string Constraint = ""> {
   let VLMul = lmul.value in {
     def "_" # lmul.MX # "_" # emul.MX : VPseudoBinaryNoMask<RetClass, Op1Class, Op2Class,
                                                             Constraint>;
+    def "_" # lmul.MX # "_" # emul.MX # "_TU": VPseudoBinaryNoMaskTU<RetClass, Op1Class, Op2Class,
+                                                                     Constraint>;
     def "_" # lmul.MX # "_" # emul.MX # "_MASK" : VPseudoBinaryMaskTA<RetClass, Op1Class, Op2Class,
                                                                       Constraint>;
   }
 }
 
 multiclass VPseudoTiedBinary<VReg RetClass,
                              DAGOperand Op2Class,
                              LMULInfo MInfo,
                              string Constraint = ""> {
   let VLMul = MInfo.value in {
     def "_" # MInfo.MX # "_TIED": VPseudoTiedBinaryNoMask<RetClass, Op2Class,
                                                           Constraint>;
+    def "_" # MInfo.MX # "_TIED_TU": VPseudoTiedBinaryNoMaskTU<RetClass, Op2Class,
+                                                               Constraint>;
     def "_" # MInfo.MX # "_MASK_TIED" : VPseudoTiedBinaryMask<RetClass, Op2Class,
                                                          Constraint>;
   }
 }
 
 multiclass VPseudoBinaryV_VV<string Constraint = ""> {
   foreach m = MxList in
     defm _VV : VPseudoBinary<m.vrclass, m.vrclass, m.vrclass, m, Constraint>;
Pat<(result_type (!cast<Intrinsic>(intrinsic)
                    (mask_type VR:$rs2),
                    VLOpFrag)),
                    (!cast<Instruction>(inst#"_"#kind#"_"#vlmul.MX)
                    (result_type result_reg_class:$merge),
                    (op1_type op1_reg_class:$rs1),
                    (mask_type VR:$rs2),
                    GPR:$vl, sew)>;
 
-class VPatBinaryNoMask<string intrinsic_name,
+class VPatBinaryM<string intrinsic_name,
+                  string inst,
+                  ValueType result_type,
+                  ValueType op1_type,
+                  ValueType op2_type,
+                  int sew,
+                  VReg op1_reg_class,
+                  DAGOperand op2_kind> :
+  Pat<(result_type (!cast<Intrinsic>(intrinsic_name)
+                   (op1_type op1_reg_class:$rs1),
+                   (op2_type op2_kind:$rs2),
+                   VLOpFrag)),
+                   (!cast<Instruction>(inst)
+                   (op1_type op1_reg_class:$rs1),
+                   (op2_type op2_kind:$rs2),
+                   GPR:$vl, sew)>;
+
+class VPatBinaryNoMaskTA<string intrinsic_name,
                          string inst,
                          ValueType result_type,
                          ValueType op1_type,
                          ValueType op2_type,
                          int sew,
                          VReg op1_reg_class,
                          DAGOperand op2_kind> :
   Pat<(result_type (!cast<Intrinsic>(intrinsic_name)
+                   (result_type (undef)),
                    (op1_type op1_reg_class:$rs1),
                    (op2_type op2_kind:$rs2),
                    VLOpFrag)),
                    (!cast<Instruction>(inst)
                    (op1_type op1_reg_class:$rs1),
                    (op2_type op2_kind:$rs2),
                    GPR:$vl, sew)>;
 
+class VPatBinaryNoMaskTU<string intrinsic_name,
+                         string inst,
+                         ValueType result_type,
+                         ValueType op1_type,
+                         ValueType op2_type,
+                         int sew,
+                         VReg result_reg_class,
+                         VReg op1_reg_class,
+                         DAGOperand op2_kind> :
+  Pat<(result_type (!cast<Intrinsic>(intrinsic_name)
+                   (result_type result_reg_class:$merge),
+                   (op1_type op1_reg_class:$rs1),
+                   (op2_type op2_kind:$rs2),
+                   VLOpFrag)),
+                   (!cast<Instruction>(inst#"_TU")
+                   (result_type result_reg_class:$merge),
+                   (op1_type op1_reg_class:$rs1),
+                   (op2_type op2_kind:$rs2),
+                   GPR:$vl, sew)>;
+
 // Same as above but source operands are swapped.
 class VPatBinaryNoMaskSwapped<string intrinsic_name,
                               string inst,
                               ValueType result_type,
                               ValueType op1_type,
                               ValueType op2_type,
                               int sew,
                               VReg op1_reg_class,

 class VPatTiedBinaryNoMask<string intrinsic_name,
                            string inst,
                            ValueType result_type,
                            ValueType op2_type,
                            int sew,
                            VReg result_reg_class,
                            DAGOperand op2_kind> :
   Pat<(result_type (!cast<Intrinsic>(intrinsic_name)
+                   (result_type (undef)),
                    (result_type result_reg_class:$rs1),
                    (op2_type op2_kind:$rs2),
                    VLOpFrag)),
                    (!cast<Instruction>(inst#"_TIED")
                    (result_type result_reg_class:$rs1),
                    (op2_type op2_kind:$rs2),
                    GPR:$vl, sew)>;
 
+class VPatTiedBinaryNoMaskTU<string intrinsic_name,
+                             string inst,
+                             ValueType result_type,
+                             ValueType op2_type,
+                             int sew,
+                             VReg result_reg_class,
+                             DAGOperand op2_kind> :
+  Pat<(result_type (!cast<Intrinsic>(intrinsic_name)
+                   (result_type result_reg_class:$merge),
+                   (result_type result_reg_class:$merge),
+                   (op2_type op2_kind:$rs2),
+                   VLOpFrag)),
+                   (!cast<Instruction>(inst#"_TIED_TU")
+                   (result_type result_reg_class:$merge),
+                   (op2_type op2_kind:$rs2),
+                   GPR:$vl, sew)>;
+
 class VPatTiedBinaryMask<string intrinsic_name,
                          string inst,
                          ValueType result_type,
                          ValueType op2_type,
                          ValueType mask_type,
                          int sew,
                          VReg result_reg_class,
                          DAGOperand op2_kind> :

 multiclass VPatNullaryM<string intrinsic, string inst> {
   foreach mti = AllMasks in
     def : Pat<(mti.Mask (!cast<Intrinsic>(intrinsic)
                         (XLenVT (VLOp (XLenVT (XLenVT GPR:$vl)))))),
                         (!cast<Instruction>(inst#"_M_"#mti.BX)
                         GPR:$vl, mti.Log2SEW)>;
 }
 
-multiclass VPatBinary<string intrinsic,
+multiclass VPatBinaryM<string intrinsic,
                       string inst,
                       ValueType result_type,
                       ValueType op1_type,
                       ValueType op2_type,
                       ValueType mask_type,
                       int sew,
                       VReg result_reg_class,
                       VReg op1_reg_class,
                       DAGOperand op2_kind>
 {
-  def : VPatBinaryNoMask<intrinsic, inst, result_type, op1_type, op2_type,
-                         sew, op1_reg_class, op2_kind>;
+  def : VPatBinaryM<intrinsic, inst, result_type, op1_type, op2_type,
   def : VPatBinaryMask<intrinsic, inst, result_type, op1_type, op2_type,
                        mask_type, sew, result_reg_class, op1_reg_class,
                        op2_kind>;
 }
 
 multiclass VPatBinaryTA<string intrinsic,
                         string inst,
                         ValueType result_type,
                         ValueType op1_type,
                         ValueType op2_type,
                         ValueType mask_type,
                         int sew,
                         VReg result_reg_class,
                         VReg op1_reg_class,
                         DAGOperand op2_kind>
 {
-  def : VPatBinaryNoMask<intrinsic, inst, result_type, op1_type, op2_type,
+  def : VPatBinaryNoMaskTA<intrinsic, inst, result_type, op1_type, op2_type,
                            sew, op1_reg_class, op2_kind>;
+  def : VPatBinaryNoMaskTU<intrinsic, inst, result_type, op1_type, op2_type,
+                           sew, result_reg_class, op1_reg_class, op2_kind>;
   def : VPatBinaryMaskTA<intrinsic, inst, result_type, op1_type, op2_type,
                          mask_type, sew, result_reg_class, op1_reg_class,
                          op2_kind>;
 }
 
 multiclass VPatBinarySwapped<string intrinsic,
                       string inst,
                       ValueType result_type,
foreach vti = vtilist in
     defm : VPatBinaryTA<intrinsic, instruction # "_VI_" # vti.LMul.MX,
                         vti.Vector, vti.Vector, XLenVT, vti.Mask,
                         vti.Log2SEW, vti.RegClass,
                         vti.RegClass, imm_type>;
 }
 
 multiclass VPatBinaryM_MM<string intrinsic, string instruction> {
   foreach mti = AllMasks in
-    def : VPatBinaryNoMask<intrinsic, instruction # "_MM_" # mti.LMul.MX,
-                           mti.Mask, mti.Mask, mti.Mask,
-                           mti.Log2SEW, VR, VR>;
+    def : VPatBinaryM<intrinsic, instruction # "_MM_" # mti.LMul.MX,
 }
 
 multiclass VPatBinaryW_VV<string intrinsic, string instruction,
                           list<VTypeInfoToWide> vtilist> {
   foreach VtiToWti = vtilist in {
     defvar Vti = VtiToWti.Vti;
     defvar Wti = VtiToWti.Wti;
     defm : VPatBinaryTA<intrinsic, instruction # "_VV_" # Vti.LMul.MX,

 multiclass VPatBinaryW_WV<string intrinsic, string instruction,
                           list<VTypeInfoToWide> vtilist> {
   foreach VtiToWti = vtilist in {
     defvar Vti = VtiToWti.Vti;
     defvar Wti = VtiToWti.Wti;
     def : VPatTiedBinaryNoMask<intrinsic, instruction # "_WV_" # Vti.LMul.MX,
                                Wti.Vector, Vti.Vector,
                                Vti.Log2SEW, Wti.RegClass, Vti.RegClass>;
-    let AddedComplexity = 1 in
+    def : VPatBinaryNoMaskTU<intrinsic, instruction # "_WV_" # Vti.LMul.MX,
+                             Wti.Vector, Wti.Vector, Vti.Vector, Vti.Log2SEW,
+                             Wti.RegClass, Wti.RegClass, Vti.RegClass>;
+    let AddedComplexity = 1 in {
+    def : VPatTiedBinaryNoMaskTU<intrinsic, instruction # "_WV_" # Vti.LMul.MX,
+                                 Wti.Vector, Vti.Vector,
+                                 Vti.Log2SEW, Wti.RegClass, Vti.RegClass>;
     def : VPatTiedBinaryMask<intrinsic, instruction # "_WV_" # Vti.LMul.MX,
                              Wti.Vector, Vti.Vector, Vti.Mask,
                              Vti.Log2SEW, Wti.RegClass, Vti.RegClass>;
+    }
     def : VPatBinaryMaskTA<intrinsic, instruction # "_WV_" # Vti.LMul.MX,
                            Wti.Vector, Wti.Vector, Vti.Vector, Vti.Mask,
                            Vti.Log2SEW, Wti.RegClass,
                            Wti.RegClass, Vti.RegClass>;
   }
 }
 
 multiclass VPatBinaryW_WX<string intrinsic, string instruction,
defm : VPatBinaryMaskOut<intrinsic, instruction, "VI",
                              vti.Mask, vti.Vector, XLenVT,
                              vti.Log2SEW, vti.LMul,
                              vti.RegClass, simm5>;
 }
 
 multiclass VPatBinaryM_VV<string intrinsic, string instruction,
                           list<VTypeInfo> vtilist> {
   foreach vti = vtilist in
-    defm : VPatBinary<intrinsic, instruction # "_VV_" # vti.LMul.MX,
+    defm : VPatBinaryM<intrinsic, instruction # "_VV_" # vti.LMul.MX,
                        vti.Mask, vti.Vector, vti.Vector, vti.Mask,
                        vti.Log2SEW, VR,
                        vti.RegClass, vti.RegClass>;
 }
 
 multiclass VPatBinarySwappedM_VV<string intrinsic, string instruction,
                                  list<VTypeInfo> vtilist> {
   foreach vti = vtilist in
     defm : VPatBinarySwapped<intrinsic, instruction # "_VV_" # vti.LMul.MX,
                              vti.Mask, vti.Vector, vti.Vector, vti.Mask,
                              vti.Log2SEW, VR,
                              vti.RegClass, vti.RegClass>;
 }
 
 multiclass VPatBinaryM_VX<string intrinsic, string instruction,
                           list<VTypeInfo> vtilist> {
   foreach vti = vtilist in {
     defvar kind = "V"#vti.ScalarSuffix;
-    defm : VPatBinary<intrinsic, instruction#"_"#kind#"_"#vti.LMul.MX,
+    defm : VPatBinaryM<intrinsic, instruction#"_"#kind#"_"#vti.LMul.MX,
                        vti.Mask, vti.Vector, vti.Scalar, vti.Mask,
                        vti.Log2SEW, VR,
                        vti.RegClass, vti.ScalarRegClass>;
   }
 }
 
 multiclass VPatBinaryM_VI<string intrinsic, string instruction,
                           list<VTypeInfo> vtilist> {
   foreach vti = vtilist in
-    defm : VPatBinary<intrinsic, instruction # "_VI_" # vti.LMul.MX,
+    defm : VPatBinaryM<intrinsic, instruction # "_VI_" # vti.LMul.MX,
                        vti.Mask, vti.Vector, XLenVT, vti.Mask,
                        vti.Log2SEW, VR,
                        vti.RegClass, simm5>;
 }
 
 multiclass VPatBinaryV_VV_VX_VI<string intrinsic, string instruction,
                                 list<VTypeInfo> vtilist, Operand ImmType = simm5>
     : VPatBinaryV_VV<intrinsic, instruction, vtilist>,
       VPatBinaryV_VX<intrinsic, instruction, vtilist>,
       VPatBinaryV_VI<intrinsic, instruction, vtilist, ImmType>;
 

 defm PseudoVSUB   : VPseudoVALU_VV_VX;
 defm PseudoVRSUB  : VPseudoVALU_VX_VI;
 
 foreach vti = AllIntegerVectors in {
   // Match vrsub with 2 vector operands to vsub.vv by swapping operands. This
   // Occurs when legalizing vrsub.vx intrinsics for i64 on RV32 since we need
   // to use a more complex splat sequence. Add the pattern for all VTs for
   // consistency.
-  def : Pat<(vti.Vector (int_riscv_vrsub (vti.Vector vti.RegClass:$rs2),
+  def : Pat<(vti.Vector (int_riscv_vrsub (vti.Vector (undef)),
+                                         (vti.Vector vti.RegClass:$rs2),
                                          (vti.Vector vti.RegClass:$rs1),
                                          VLOpFrag)),
             (!cast<Instruction>("PseudoVSUB_VV_"#vti.LMul.MX) vti.RegClass:$rs1,
                                                               vti.RegClass:$rs2,
                                                               GPR:$vl,
                                                               vti.Log2SEW)>;
+  def : Pat<(vti.Vector (int_riscv_vrsub (vti.Vector vti.RegClass:$merge),
+                                         (vti.Vector vti.RegClass:$rs2),
+                                         (vti.Vector vti.RegClass:$rs1),
+                                         VLOpFrag)),
+            (!cast<Instruction>("PseudoVSUB_VV_"#vti.LMul.MX#"_TU")
+                                                      vti.RegClass:$merge,
+                                                      vti.RegClass:$rs1,
+                                                      vti.RegClass:$rs2,
+                                                      GPR:$vl,
+                                                      vti.Log2SEW)>;
   def : Pat<(vti.Vector (int_riscv_vrsub_mask (vti.Vector vti.RegClass:$merge),
                                               (vti.Vector vti.RegClass:$rs2),
                                               (vti.Vector vti.RegClass:$rs1),
                                               (vti.Mask V0),
                                               VLOpFrag,
                                               (XLenVT timm:$policy))),
             (!cast<Instruction>("PseudoVSUB_VV_"#vti.LMul.MX#"_MASK")
                                                       vti.RegClass:$merge,
                                                       vti.RegClass:$rs1,
                                                       vti.RegClass:$rs2,
                                                       (vti.Mask V0),
                                                       GPR:$vl,
                                                       vti.Log2SEW,
                                                       (XLenVT timm:$policy))>;
 
   // Match VSUB with a small immediate to vadd.vi by negating the immediate.
-  def : Pat<(vti.Vector (int_riscv_vsub (vti.Vector vti.RegClass:$rs1),
+  def : Pat<(vti.Vector (int_riscv_vsub (vti.Vector (undef)),
+                                        (vti.Vector vti.RegClass:$rs1),
                                         (vti.Scalar simm5_plus1:$rs2),
                                         VLOpFrag)),
             (!cast<Instruction>("PseudoVADD_VI_"#vti.LMul.MX) vti.RegClass:$rs1,
                                                               (NegImm simm5_plus1:$rs2),
                                                               GPR:$vl,
                                                               vti.Log2SEW)>;
   def : Pat<(vti.Vector (int_riscv_vsub_mask (vti.Vector vti.RegClass:$merge),
                                              (vti.Vector vti.RegClass:$rs1),
defm : VPatBinaryV_VV_VX_VI<"int_riscv_vsll", "PseudoVSLL", AllIntegerVectors,
                             uimm5>;
 defm : VPatBinaryV_VV_VX_VI<"int_riscv_vsrl", "PseudoVSRL", AllIntegerVectors,
                             uimm5>;
 defm : VPatBinaryV_VV_VX_VI<"int_riscv_vsra", "PseudoVSRA", AllIntegerVectors,
                             uimm5>;
 
 foreach vti = AllIntegerVectors in {
   // Emit shift by 1 as an add since it might be faster.
-  def : Pat<(vti.Vector (int_riscv_vsll (vti.Vector vti.RegClass:$rs1),
+  def : Pat<(vti.Vector (int_riscv_vsll (vti.Vector undef),
+                                        (vti.Vector vti.RegClass:$rs1),
                                         (XLenVT 1), VLOpFrag)),
             (!cast<Instruction>("PseudoVADD_VV_"#vti.LMul.MX) vti.RegClass:$rs1,
                                                               vti.RegClass:$rs1,
                                                               GPR:$vl,
                                                               vti.Log2SEW)>;
   def : Pat<(vti.Vector (int_riscv_vsll_mask (vti.Vector vti.RegClass:$merge),
                                              (vti.Vector vti.RegClass:$rs1),
                                              (XLenVT 1),

llvm/lib/Target/RISCV/RISCVInstrInfoVVLPatterns.td

 def riscv_vid_vl : SDNode<"RISCVISD::VID_VL", SDTypeProfile<1, 2,
                           [SDTCisVec<0>, SDTCVecEltisVT<1, i1>,
                            SDTCisSameNumEltsAs<0, 1>, SDTCisVT<2, XLenVT>]>, []>;
 
 def SDTRVVSlide : SDTypeProfile<1, 5, [
   SDTCisVec<0>, SDTCisSameAs<1, 0>, SDTCisSameAs<2, 0>, SDTCisVT<3, XLenVT>,
   SDTCVecEltisVT<4, i1>, SDTCisSameNumEltsAs<0, 4>, SDTCisVT<5, XLenVT>
 ]>;
-def SDTRVVSlide1 : SDTypeProfile<1, 4, [
-  SDTCisVec<0>, SDTCisSameAs<1, 0>, SDTCisInt<0>, SDTCisVT<2, XLenVT>,
-  SDTCVecEltisVT<3, i1>, SDTCisSameNumEltsAs<0, 3>, SDTCisVT<4, XLenVT>
+def SDTRVVSlide1 : SDTypeProfile<1, 5, [
+  SDTCisVec<0>, SDTCisSameAs<1, 0>, SDTCisSameAs<2, 0>, SDTCisInt<0>,
+  SDTCisVT<3, XLenVT>, SDTCVecEltisVT<4, i1>, SDTCisSameNumEltsAs<0, 4>,
+  SDTCisVT<5, XLenVT>
 ]>;
 
 def riscv_slideup_vl   : SDNode<"RISCVISD::VSLIDEUP_VL", SDTRVVSlide, []>;
 def riscv_slide1up_vl  : SDNode<"RISCVISD::VSLIDE1UP_VL", SDTRVVSlide1, []>;
 def riscv_slidedown_vl : SDNode<"RISCVISD::VSLIDEDOWN_VL", SDTRVVSlide, []>;
 def riscv_slide1down_vl  : SDNode<"RISCVISD::VSLIDE1DOWN_VL", SDTRVVSlide1, []>;
 
 let Predicates = [HasVInstructions] in {
 
 foreach vti = AllIntegerVectors in {
   def : Pat<(vti.Vector (riscv_vid_vl (vti.Mask true_mask),
                                       VLOpFrag)),
             (!cast<Instruction>("PseudoVID_V_"#vti.LMul.MX) GPR:$vl, vti.Log2SEW)>;
 
-  def : Pat<(vti.Vector (riscv_slide1up_vl (vti.Vector vti.RegClass:$rs1),
+  def : Pat<(vti.Vector (riscv_slide1up_vl (vti.Vector undef),
+                                           (vti.Vector vti.RegClass:$rs1),
                                            GPR:$rs2, (vti.Mask true_mask),
                                            VLOpFrag)),
             (!cast<Instruction>("PseudoVSLIDE1UP_VX_"#vti.LMul.MX)
                 vti.RegClass:$rs1, GPR:$rs2, GPR:$vl, vti.Log2SEW)>;
-  def : Pat<(vti.Vector (riscv_slide1down_vl (vti.Vector vti.RegClass:$rs1),
+  def : Pat<(vti.Vector (riscv_slide1up_vl (vti.Vector vti.RegClass:$rd),
+                                           (vti.Vector vti.RegClass:$rs1),
+                                           GPR:$rs2, (vti.Mask true_mask),
+                                           VLOpFrag)),
+            (!cast<Instruction>("PseudoVSLIDE1UP_VX_"#vti.LMul.MX#"_TU")
+                vti.RegClass:$rd, vti.RegClass:$rs1, GPR:$rs2, GPR:$vl, vti.Log2SEW)>;
+  def : Pat<(vti.Vector (riscv_slide1down_vl (vti.Vector undef),
+                                             (vti.Vector vti.RegClass:$rs1),
                                              GPR:$rs2, (vti.Mask true_mask),
                                              VLOpFrag)),
             (!cast<Instruction>("PseudoVSLIDE1DOWN_VX_"#vti.LMul.MX)
                 vti.RegClass:$rs1, GPR:$rs2, GPR:$vl, vti.Log2SEW)>;
+  def : Pat<(vti.Vector (riscv_slide1down_vl (vti.Vector vti.RegClass:$rd),
+                                             (vti.Vector vti.RegClass:$rs1),
+                                             GPR:$rs2, (vti.Mask true_mask),
+                                             VLOpFrag)),
+            (!cast<Instruction>("PseudoVSLIDE1DOWN_VX_"#vti.LMul.MX#"_TU")
+                vti.RegClass:$rd, vti.RegClass:$rs1, GPR:$rs2, GPR:$vl, vti.Log2SEW)>;
 }
 
 foreach vti = !listconcat(AllIntegerVectors, AllFloatVectors) in {
   def : Pat<(vti.Vector (riscv_slideup_vl (vti.Vector vti.RegClass:$rs3),
                                           (vti.Vector vti.RegClass:$rs1),
                                           uimm5:$rs2, (vti.Mask true_mask),
                                           VLOpFrag)),
             (!cast<Instruction>("PseudoVSLIDEUP_VI_"#vti.LMul.MX)

llvm/test/CodeGen/RISCV/rvv/masked-vslide1down-rv32.ll

+; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
+; RUN: llc -mtriple=riscv32 -mattr=+v,+f -verify-machineinstrs \
+; RUN:   < %s | FileCheck %s
+
+declare <vscale x 1 x i64> @llvm.riscv.vslide1down.mask.nxv1i64.i64(
+  <vscale x 1 x i64>,
+  <vscale x 1 x i64>,
+  i64,
+  <vscale x 1 x i1>,
+  i32,
+  i32);
+
+define <vscale x 1 x i64> @intrinsic_vslide1down_mask_tumu_vx_nxv1i64_nxv1i64_i64(<vscale x 1 x i64> %0, <vscale x 1 x i64> %1, i64 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+; CHECK-LABEL: intrinsic_vslide1down_mask_tumu_vx_nxv1i64_nxv1i64_i64:
+; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    slli a3, a2, 1
+; CHECK-NEXT:    vsetvli zero, a3, e32, m1, ta, mu
+; CHECK-NEXT:    vslide1down.vx v9, v9, a0
+; CHECK-NEXT:    vslide1down.vx v9, v9, a1
+; CHECK-NEXT:    vsetvli zero, a2, e64, m1, tu, mu
+; CHECK-NEXT:    vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i64> @llvm.riscv.vslide1down.mask.nxv1i64.i64(
+    <vscale x 1 x i64> %0,
+    <vscale x 1 x i64> %1,
+    i64 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4, i32 0)
+
+  ret <vscale x 1 x i64> %a
+}
+
+define <vscale x 1 x i64> @intrinsic_vslide1down_mask_tamu_vx_nxv1i64_nxv1i64_i64(<vscale x 1 x i64> %0, <vscale x 1 x i64> %1, i64 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+; CHECK-LABEL: intrinsic_vslide1down_mask_tamu_vx_nxv1i64_nxv1i64_i64:
+; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    slli a3, a2, 1
+; CHECK-NEXT:    vsetvli zero, a3, e32, m1, ta, mu
+; CHECK-NEXT:    vslide1down.vx v9, v9, a0
+; CHECK-NEXT:    vslide1down.vx v9, v9, a1
+; CHECK-NEXT:    vsetvli zero, a2, e64, m1, ta, mu
+; CHECK-NEXT:    vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i64> @llvm.riscv.vslide1down.mask.nxv1i64.i64(
+    <vscale x 1 x i64> %0,
+    <vscale x 1 x i64> %1,
+    i64 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4, i32 1)
+
+  ret <vscale x 1 x i64> %a
+}
+
+
+; Fallback vslide1 to mask undisturbed untill InsertVSETVLI supports mask agnostic.
+define <vscale x 1 x i64> @intrinsic_vslide1down_mask_tuma_vx_nxv1i64_nxv1i64_i64(<vscale x 1 x i64> %0, <vscale x 1 x i64> %1, i64 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+; CHECK-LABEL: intrinsic_vslide1down_mask_tuma_vx_nxv1i64_nxv1i64_i64:
+; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    slli a3, a2, 1
+; CHECK-NEXT:    vsetvli zero, a3, e32, m1, ta, mu
+; CHECK-NEXT:    vslide1down.vx v9, v9, a0
+; CHECK-NEXT:    vslide1down.vx v9, v9, a1
+; CHECK-NEXT:    vsetvli zero, a2, e64, m1, tu, mu
+; CHECK-NEXT:    vmerge.vvm v8, v8, v9, v0
+; CHECK-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i64> @llvm.riscv.vslide1down.mask.nxv1i64.i64(
+    <vscale x 1 x i64> %0,
+    <vscale x 1 x i64> %1,
+    i64 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4, i32 2)
+
+  ret <vscale x 1 x i64> %a
+}
+
+; Fallback vslide1 to mask undisturbed untill InsertVSETVLI supports mask agnostic.
+define <vscale x 1 x i64> @intrinsic_vslide1down_mask_tama_vx_nxv1i64_nxv1i64_i64(<vscale x 1 x i64> %0, i64 %1, <vscale x 1 x i1> %2, i32 %3) nounwind {
+; CHECK-LABEL: intrinsic_vslide1down_mask_tama_vx_nxv1i64_nxv1i64_i64:
+; CHECK:       # %bb.0: # %entry
+; CHECK-NEXT:    slli a2, a2, 1
+; CHECK-NEXT:    vsetvli zero, a2, e32, m1, ta, mu
+; CHECK-NEXT:    vslide1down.vx v8, v8, a0
+; CHECK-NEXT:    vslide1down.vx v8, v8, a1
+; CHECK-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i64> @llvm.riscv.vslide1down.mask.nxv1i64.i64(
+    <vscale x 1 x i64> undef,
+    <vscale x 1 x i64> %0,
+    i64 %1,
+    <vscale x 1 x i1> %2,
+    i32 %3, i32 3)
+
+  ret <vscale x 1 x i64> %a
+}

llvm/test/CodeGen/RISCV/rvv/unmasked-tu.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: sed 's/iXLen/i32/g' %s | llc -mtriple=riscv32 -mattr=+v \
-; RUN:   -verify-machineinstrs | FileCheck %s --check-prefix=RV32
-; RUN: sed 's/iXLen/i64/g' %s | llc -mtriple=riscv64 -mattr=+v \
-; RUN:   -verify-machineinstrs | FileCheck %s --check-prefix=RV64
+; RUN: sed 's/iXLen/i32/g' %s | llc -mtriple=riscv32 -mattr=+v,+zfh \
+; RUN:   -verify-machineinstrs -target-abi=ilp32d | FileCheck %s --check-prefix=RV32
+; RUN: sed 's/iXLen/i64/g' %s | llc -mtriple=riscv64 -mattr=+v,+zfh \
+; RUN:   -verify-machineinstrs -target-abi=lp64d | FileCheck %s --check-prefix=RV64
 
 declare <vscale x 1 x i8> @llvm.riscv.vle.nxv1i8(
   <vscale x 1 x i8>,
   <vscale x 1 x i8>*,
   iXLen);
 
 define <vscale x 1 x i8> @intrinsic_vle_v_tu_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8>* %1, iXLen %2) nounwind {
 ; RV32-LABEL: intrinsic_vle_v_tu_nxv1i8_nxv1i8:
entry:
   %a = call <vscale x 1 x i8> @llvm.riscv.vloxei.nxv1i8(
     <vscale x 1 x i8> %0,
     <vscale x 1 x i8>* %1,
     <vscale x 1 x i8> %2,
     iXLen %3)
 
   ret <vscale x 1 x i8> %a
 }
+
+declare <vscale x 1 x i8> @llvm.riscv.vaadd.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vaadd_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vaadd_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vaadd.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vaadd_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vaadd.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vaadd.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vaaddu.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vaaddu_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vaaddu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vaaddu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vaaddu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vaaddu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vaaddu.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vadd.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vadd_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vadd_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vadd.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vadd_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vadd.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vadd.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+declare <vscale x 1 x i8> @llvm.riscv.vand.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vand_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vand_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vand.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vand_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vand.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vand.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vasub.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vasub_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vasub_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vasub.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vasub_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vasub.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vasub.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vasubu.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vasubu_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vasubu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vasubu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vasubu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vasubu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vasubu.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vdiv.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vdiv_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vdiv_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vdiv.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vdiv_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vdiv.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vdiv.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vdivu.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vdivu_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vdivu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vdivu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vdivu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vdivu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vdivu.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfadd.nxv1f16.nxv1f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfadd_vv_nxv1f16_nxv1f16_nxv1f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfadd_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfadd.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfadd_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfadd.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfadd.nxv1f16.nxv1f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfdiv.nxv1f16.nxv1f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfdiv_vv_nxv1f16_nxv1f16_nxv1f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfdiv_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfdiv.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfdiv_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfdiv.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfdiv.nxv1f16.nxv1f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfmax.nxv1f16.nxv1f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfmax_vv_nxv1f16_nxv1f16_nxv1f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfmax_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfmax.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfmax_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfmax.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfmax.nxv1f16.nxv1f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfmin.nxv1f16.nxv1f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfmin_vv_nxv1f16_nxv1f16_nxv1f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfmin_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfmin.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfmin_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfmin.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfmin.nxv1f16.nxv1f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfmul.nxv1f16.nxv1f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfmul_vv_nxv1f16_nxv1f16_nxv1f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfmul_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfmul.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfmul_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfmul.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfmul.nxv1f16.nxv1f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfrdiv.nxv1f16.f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  half,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfrdiv_vf_nxv1f16_nxv1f16_f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, half %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfrdiv_vf_nxv1f16_nxv1f16_f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfrdiv.vf v8, v9, fa0
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfrdiv_vf_nxv1f16_nxv1f16_f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfrdiv.vf v8, v9, fa0
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfrdiv.nxv1f16.f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    half %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfsgnj.nxv1f16.nxv1f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfsgnj_vv_nxv1f16_nxv1f16_nxv1f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfsgnj_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfsgnj.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfsgnj_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfsgnj.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfsgnj.nxv1f16.nxv1f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfsgnjn.nxv1f16.nxv1f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfsgnjn_vv_nxv1f16_nxv1f16_nxv1f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfsgnjn_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfsgnjn.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfsgnjn_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfsgnjn.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfsgnjn.nxv1f16.nxv1f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfsgnjx.nxv1f16.nxv1f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfsgnjx_vv_nxv1f16_nxv1f16_nxv1f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfsgnjx_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfsgnjx.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfsgnjx_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfsgnjx.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfsgnjx.nxv1f16.nxv1f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfrsub.nxv1f16.f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  half,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfrsub_vf_nxv1f16_nxv1f16_f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, half %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfrsub_vf_nxv1f16_nxv1f16_f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfrsub.vf v8, v9, fa0
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfrsub_vf_nxv1f16_nxv1f16_f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfrsub.vf v8, v9, fa0
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfrsub.nxv1f16.f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    half %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfslide1down.nxv1f16.f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  half,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfslide1down_vf_nxv1f16_nxv1f16_f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, half %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfslide1down_vf_nxv1f16_nxv1f16_f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfslide1down.vf v8, v9, fa0
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfslide1down_vf_nxv1f16_nxv1f16_f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfslide1down.vf v8, v9, fa0
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfslide1down.nxv1f16.f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    half %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfslide1up.nxv1f16.f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  half,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfslide1up_vf_nxv1f16_nxv1f16_f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, half %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfslide1up_vf_nxv1f16_nxv1f16_f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfslide1up.vf v8, v9, fa0
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfslide1up_vf_nxv1f16_nxv1f16_f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfslide1up.vf v8, v9, fa0
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfslide1up.nxv1f16.f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    half %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+declare <vscale x 1 x float> @llvm.riscv.vfwsub.nxv1f32.nxv1f16.nxv1f16(
+  <vscale x 1 x float>,
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x float> @intrinsic_vfwsub_vv_nxv1f32_nxv1f16_nxv1f16(<vscale x 1 x float> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfwsub_vv_nxv1f32_nxv1f16_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfwsub.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfwsub_vv_nxv1f32_nxv1f16_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfwsub.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x float> @llvm.riscv.vfwsub.nxv1f32.nxv1f16.nxv1f16(
+    <vscale x 1 x float> %0,
+    <vscale x 1 x half> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x float> %a
+}
+
+declare <vscale x 1 x float> @llvm.riscv.vfwsub.w.nxv1f32.nxv1f16(
+  <vscale x 1 x float>,
+  <vscale x 1 x float>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x float> @intrinsic_vfwsub.w_wv_nxv1f32_nxv1f32_nxv1f16(<vscale x 1 x float> %0, <vscale x 1 x float> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfwsub.w_wv_nxv1f32_nxv1f32_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfwsub.wv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfwsub.w_wv_nxv1f32_nxv1f32_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfwsub.wv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x float> @llvm.riscv.vfwsub.w.nxv1f32.nxv1f16(
+    <vscale x 1 x float> %0,
+    <vscale x 1 x float> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x float> %a
+}
+
+declare <vscale x 16 x float> @llvm.riscv.vfwsub.w.nxv16f32.nxv16f16(
+  <vscale x 16 x float>,
+  <vscale x 16 x float>,
+  <vscale x 16 x half>,
+  iXLen);
+
+define <vscale x 16 x float> @intrinsic_vfwsub.w_wv_nxv16f32_nxv16f32_nxv16f16(<vscale x 16 x float> %0, <vscale x 16 x float> %1, <vscale x 16 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfwsub.w_wv_nxv16f32_nxv16f32_nxv16f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vl4re16.v v24, (a0)
+; RV32-NEXT:    vsetvli zero, a1, e16, m4, tu, mu
+; RV32-NEXT:    vfwsub.wv v8, v16, v24
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfwsub.w_wv_nxv16f32_nxv16f32_nxv16f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vl4re16.v v24, (a0)
+; RV64-NEXT:    vsetvli zero, a1, e16, m4, tu, mu
+; RV64-NEXT:    vfwsub.wv v8, v16, v24
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 16 x float> @llvm.riscv.vfwsub.w.nxv16f32.nxv16f16(
+    <vscale x 16 x float> %0,
+    <vscale x 16 x float> %1,
+    <vscale x 16 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 16 x float> %a
+}
+
+declare <vscale x 1 x float> @llvm.riscv.vfwmul.nxv1f32.nxv1f16.nxv1f16(
+  <vscale x 1 x float>,
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x float> @intrinsic_vfwmul_vv_nxv1f32_nxv1f16_nxv1f16(<vscale x 1 x float> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfwmul_vv_nxv1f32_nxv1f16_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfwmul.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfwmul_vv_nxv1f32_nxv1f16_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfwmul.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x float> @llvm.riscv.vfwmul.nxv1f32.nxv1f16.nxv1f16(
+    <vscale x 1 x float> %0,
+    <vscale x 1 x half> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x float> %a
+}
+
+declare <vscale x 1 x float> @llvm.riscv.vfwadd.w.nxv1f32.nxv1f16(
+  <vscale x 1 x float>,
+  <vscale x 1 x float>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x float> @intrinsic_vfwadd.w_wv_nxv1f32_nxv1f32_nxv1f16(<vscale x 1 x float> %0, <vscale x 1 x float> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfwadd.w_wv_nxv1f32_nxv1f32_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfwadd.wv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfwadd.w_wv_nxv1f32_nxv1f32_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfwadd.wv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x float> @llvm.riscv.vfwadd.w.nxv1f32.nxv1f16(
+    <vscale x 1 x float> %0,
+    <vscale x 1 x float> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x float> %a
+}
+
+declare <vscale x 1 x float> @llvm.riscv.vfwadd.nxv1f32.nxv1f16.nxv1f16(
+  <vscale x 1 x float>,
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x float> @intrinsic_vfwadd_vv_nxv1f32_nxv1f16_nxv1f16(<vscale x 1 x float> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfwadd_vv_nxv1f32_nxv1f16_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfwadd.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfwadd_vv_nxv1f32_nxv1f16_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfwadd.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x float> @llvm.riscv.vfwadd.nxv1f32.nxv1f16.nxv1f16(
+    <vscale x 1 x float> %0,
+    <vscale x 1 x half> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x float> %a
+}
+
+declare <vscale x 1 x half> @llvm.riscv.vfsub.nxv1f16.nxv1f16(
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  <vscale x 1 x half>,
+  iXLen);
+
+define <vscale x 1 x half> @intrinsic_vfsub_vv_nxv1f16_nxv1f16_nxv1f16(<vscale x 1 x half> %0, <vscale x 1 x half> %1, <vscale x 1 x half> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vfsub_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV32-NEXT:    vfsub.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vfsub_vv_nxv1f16_nxv1f16_nxv1f16:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e16, mf4, tu, mu
+; RV64-NEXT:    vfsub.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x half> @llvm.riscv.vfsub.nxv1f16.nxv1f16(
+    <vscale x 1 x half> %0,
+    <vscale x 1 x half> %1,
+    <vscale x 1 x half> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x half> %a
+}
+
+
+declare <vscale x 1 x i64> @llvm.riscv.vslide1down.nxv1i64(
+  <vscale x 1 x i64>,
+  <vscale x 1 x i64>,
+  i64,
+  iXLen);
+
+define <vscale x 1 x i64> @intrinsic_vslide1down_vx_nxv1i64_nxv1i64_i64(<vscale x 1 x i64> %0, <vscale x 1 x i64> %1, i64 %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vslide1down_vx_nxv1i64_nxv1i64_i64:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    slli a2, a2, 1
+; RV32-NEXT:    vsetvli zero, a2, e32, m1, tu, mu
+; RV32-NEXT:    vmv1r.v v10, v8
+; RV32-NEXT:    vslide1down.vx v10, v9, a0
+; RV32-NEXT:    vslide1down.vx v8, v10, a1
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vslide1down_vx_nxv1i64_nxv1i64_i64:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, tu, mu
+; RV64-NEXT:    vslide1down.vx v8, v9, a0
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i64> @llvm.riscv.vslide1down.nxv1i64(
+    <vscale x 1 x i64> %0,
+    <vscale x 1 x i64> %1,
+    i64 %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i64> %a
+}
+
+declare <vscale x 1 x i64> @llvm.riscv.vslide1up.nxv1i64.i64(
+  <vscale x 1 x i64>,
+  <vscale x 1 x i64>,
+  i64,
+  iXLen);
+
+define <vscale x 1 x i64> @intrinsic_vslide1up_vx_nxv1i64_nxv1i64_i64(<vscale x 1 x i64> %0, <vscale x 1 x i64> %1, i64 %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vslide1up_vx_nxv1i64_nxv1i64_i64:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    slli a2, a2, 1
+; RV32-NEXT:    vsetvli zero, a2, e32, m1, tu, mu
+; RV32-NEXT:    vmv1r.v v10, v8
+; RV32-NEXT:    vslide1up.vx v10, v9, a1
+; RV32-NEXT:    vslide1up.vx v8, v10, a0
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vslide1up_vx_nxv1i64_nxv1i64_i64:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, tu, mu
+; RV64-NEXT:    vslide1up.vx v8, v9, a0
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i64> @llvm.riscv.vslide1up.nxv1i64.i64(
+    <vscale x 1 x i64> %0,
+    <vscale x 1 x i64> %1,
+    i64 %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i64> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vmax.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vmax_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vmax_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vmax.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vmax_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vmax.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vmax.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vmaxu.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vmaxu_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vmaxu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vmaxu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vmaxu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vmaxu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vmaxu.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vmin.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vmin_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vmin_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vmin.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vmin_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vmin.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vmin.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vminu.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vminu_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vminu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vminu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vminu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vminu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vminu.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vmul.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vmul_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vmul_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vmul.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vmul_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vmul.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vmul.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vmulh.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vmulh_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vmulh_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vmulh.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vmulh_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vmulh.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vmulh.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vmulhsu.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vmulhsu_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vmulhsu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vmulhsu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vmulhsu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vmulhsu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vmulhsu.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vmulhu.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vmulhu_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vmulhu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vmulhu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vmulhu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vmulhu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vmulhu.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vnclip.nxv1i8.nxv1i16.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vnclip_wv_nxv1i8_nxv1i16_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i16> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vnclip_wv_nxv1i8_nxv1i16_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vnclip.wv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vnclip_wv_nxv1i8_nxv1i16_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vnclip.wv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vnclip.nxv1i8.nxv1i16.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vnclipu.nxv1i8.nxv1i16.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vnclipu_wv_nxv1i8_nxv1i16_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i16> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vnclipu_wv_nxv1i8_nxv1i16_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vnclipu.wv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vnclipu_wv_nxv1i8_nxv1i16_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vnclipu.wv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vnclipu.nxv1i8.nxv1i16.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vnsra.nxv1i8.nxv1i16.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vnsra_wv_nxv1i8_nxv1i16_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i16> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vnsra_wv_nxv1i8_nxv1i16_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vnsra.wv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vnsra_wv_nxv1i8_nxv1i16_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vnsra.wv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vnsra.nxv1i8.nxv1i16.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vnsrl.nxv1i8.nxv1i16.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vnsrl_wv_nxv1i8_nxv1i16_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i16> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vnsrl_wv_nxv1i8_nxv1i16_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vnsrl.wv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vnsrl_wv_nxv1i8_nxv1i16_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vnsrl.wv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vnsrl.nxv1i8.nxv1i16.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vor.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vor_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vor_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vor.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vor_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vor.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vor.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vrem.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vrem_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vrem_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vrem.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vrem_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vrem.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vrem.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vremu.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+declare <vscale x 1 x i8> @llvm.riscv.vrgather.vv.nxv1i8.i32(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vrgather_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vrgather_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vrgather.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vrgather_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vrgather.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vrgather.vv.nxv1i8.i32(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vrgather.vx.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vrgather_vx_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, iXLen %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vrgather_vx_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a1, e8, mf8, tu, mu
+; RV32-NEXT:    vrgather.vx v8, v9, a0
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vrgather_vx_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e8, mf8, tu, mu
+; RV64-NEXT:    vrgather.vx v8, v9, a0
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vrgather.vx.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    iXLen %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vrgatherei16.vv.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i16>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vrgatherei16_vv_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i16> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vrgatherei16_vv_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vrgatherei16.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vrgatherei16_vv_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vrgatherei16.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vrgatherei16.vv.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i16> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i64> @llvm.riscv.vrsub.nxv1i64.i64(
+  <vscale x 1 x i64>,
+  <vscale x 1 x i64>,
+  i64,
+  iXLen);
+
+define <vscale x 1 x i64> @intrinsic_vrsub_vx_nxv1i64_nxv1i64_i64(<vscale x 1 x i64> %0, <vscale x 1 x i64> %1, i64 %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vrsub_vx_nxv1i64_nxv1i64_i64:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    addi sp, sp, -16
+; RV32-NEXT:    sw a1, 12(sp)
+; RV32-NEXT:    sw a0, 8(sp)
+; RV32-NEXT:    vsetvli zero, a2, e64, m1, ta, mu
+; RV32-NEXT:    addi a0, sp, 8
+; RV32-NEXT:    vlse64.v v10, (a0), zero
+; RV32-NEXT:    vsetvli zero, zero, e64, m1, tu, mu
+; RV32-NEXT:    vsub.vv v8, v10, v9
+; RV32-NEXT:    addi sp, sp, 16
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vrsub_vx_nxv1i64_nxv1i64_i64:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, tu, mu
+; RV64-NEXT:    vrsub.vx v8, v9, a0
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i64> @llvm.riscv.vrsub.nxv1i64.i64(
+    <vscale x 1 x i64> %0,
+    <vscale x 1 x i64> %1,
+    i64 %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i64> %a
+}
+
+declare <vscale x 1 x i64> @llvm.riscv.vsadd.nxv1i64.i64(
+  <vscale x 1 x i64>,
+  <vscale x 1 x i64>,
+  i64,
+  iXLen);
+
+define <vscale x 1 x i64> @intrinsic_vsadd_vx_nxv1i64_nxv1i64_i64(<vscale x 1 x i64> %0, <vscale x 1 x i64> %1, i64 %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vsadd_vx_nxv1i64_nxv1i64_i64:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    addi sp, sp, -16
+; RV32-NEXT:    sw a1, 12(sp)
+; RV32-NEXT:    sw a0, 8(sp)
+; RV32-NEXT:    vsetvli zero, a2, e64, m1, ta, mu
+; RV32-NEXT:    addi a0, sp, 8
+; RV32-NEXT:    vlse64.v v10, (a0), zero
+; RV32-NEXT:    vsetvli zero, zero, e64, m1, tu, mu
+; RV32-NEXT:    vsadd.vv v8, v9, v10
+; RV32-NEXT:    addi sp, sp, 16
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vsadd_vx_nxv1i64_nxv1i64_i64:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, tu, mu
+; RV64-NEXT:    vsadd.vx v8, v9, a0
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i64> @llvm.riscv.vsadd.nxv1i64.i64(
+    <vscale x 1 x i64> %0,
+    <vscale x 1 x i64> %1,
+    i64 %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i64> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vsaddu.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vsaddu_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vsaddu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vsaddu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vsaddu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vsaddu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vsaddu.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vsll.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vsll_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vsll_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vsll.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vsll_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vsll.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vsll.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vsmul.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vsmul_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vsmul_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vsmul.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vsmul_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vsmul.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vsmul.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i64> @llvm.riscv.vsmul.nxv1i64.i64(
+  <vscale x 1 x i64>,
+  <vscale x 1 x i64>,
+  i64,
+  iXLen);
+
+define <vscale x 1 x i64> @intrinsic_vsmul_vx_nxv1i64_nxv1i64_i64(<vscale x 1 x i64> %0, <vscale x 1 x i64> %1, i64 %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vsmul_vx_nxv1i64_nxv1i64_i64:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    addi sp, sp, -16
+; RV32-NEXT:    sw a1, 12(sp)
+; RV32-NEXT:    sw a0, 8(sp)
+; RV32-NEXT:    vsetvli zero, a2, e64, m1, ta, mu
+; RV32-NEXT:    addi a0, sp, 8
+; RV32-NEXT:    vlse64.v v10, (a0), zero
+; RV32-NEXT:    vsetvli zero, zero, e64, m1, tu, mu
+; RV32-NEXT:    vsmul.vv v8, v9, v10
+; RV32-NEXT:    addi sp, sp, 16
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vsmul_vx_nxv1i64_nxv1i64_i64:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, tu, mu
+; RV64-NEXT:    vsmul.vx v8, v9, a0
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i64> @llvm.riscv.vsmul.nxv1i64.i64(
+    <vscale x 1 x i64> %0,
+    <vscale x 1 x i64> %1,
+    i64 %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i64> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vsra.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vsra_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vsra_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vsra.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vsra_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vsra.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vsra.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+declare <vscale x 1 x i8> @llvm.riscv.vsrl.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vsrl_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vsrl_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vsrl.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vsrl_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vsrl.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vsrl.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vssra.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vssra_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vssra_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vssra.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vssra_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vssra.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vssra.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vssrl.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vssrl_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vssrl_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vssrl.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vssrl_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vssrl.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vssrl.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vssub.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vssub_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vssub_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vssub.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vssub_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vssub.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vssub.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vssubu.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vssubu_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vssubu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vssubu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vssubu_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vssubu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vssubu.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i64> @llvm.riscv.vssub.nxv1i64.i64(
+  <vscale x 1 x i64>,
+  <vscale x 1 x i64>,
+  i64,
+  iXLen);
+
+define <vscale x 1 x i64> @intrinsic_vssub_vx_nxv1i64_nxv1i64_i64(<vscale x 1 x i64> %0, <vscale x 1 x i64> %1, i64 %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vssub_vx_nxv1i64_nxv1i64_i64:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    addi sp, sp, -16
+; RV32-NEXT:    sw a1, 12(sp)
+; RV32-NEXT:    sw a0, 8(sp)
+; RV32-NEXT:    vsetvli zero, a2, e64, m1, ta, mu
+; RV32-NEXT:    addi a0, sp, 8
+; RV32-NEXT:    vlse64.v v10, (a0), zero
+; RV32-NEXT:    vsetvli zero, zero, e64, m1, tu, mu
+; RV32-NEXT:    vssub.vv v8, v9, v10
+; RV32-NEXT:    addi sp, sp, 16
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vssub_vx_nxv1i64_nxv1i64_i64:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, tu, mu
+; RV64-NEXT:    vssub.vx v8, v9, a0
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i64> @llvm.riscv.vssub.nxv1i64.i64(
+    <vscale x 1 x i64> %0,
+    <vscale x 1 x i64> %1,
+    i64 %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i64> %a
+}
+
+declare <vscale x 1 x i64> @llvm.riscv.vssubu.nxv1i64.i64(
+  <vscale x 1 x i64>,
+  <vscale x 1 x i64>,
+  i64,
+  iXLen);
+
+define <vscale x 1 x i64> @intrinsic_vssubu_vx_nxv1i64_nxv1i64_i64(<vscale x 1 x i64> %0, <vscale x 1 x i64> %1, i64 %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vssubu_vx_nxv1i64_nxv1i64_i64:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    addi sp, sp, -16
+; RV32-NEXT:    sw a1, 12(sp)
+; RV32-NEXT:    sw a0, 8(sp)
+; RV32-NEXT:    vsetvli zero, a2, e64, m1, ta, mu
+; RV32-NEXT:    addi a0, sp, 8
+; RV32-NEXT:    vlse64.v v10, (a0), zero
+; RV32-NEXT:    vsetvli zero, zero, e64, m1, tu, mu
+; RV32-NEXT:    vssubu.vv v8, v9, v10
+; RV32-NEXT:    addi sp, sp, 16
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vssubu_vx_nxv1i64_nxv1i64_i64:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a1, e64, m1, tu, mu
+; RV64-NEXT:    vssubu.vx v8, v9, a0
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i64> @llvm.riscv.vssubu.nxv1i64.i64(
+    <vscale x 1 x i64> %0,
+    <vscale x 1 x i64> %1,
+    i64 %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i64> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vsub.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vsub_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vsub_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vsub.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vsub_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vsub.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vsub.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwadd.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i16> @intrinsic_vwadd_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vwadd_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vwadd.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vwadd_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vwadd.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwadd.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwadd.w.nxv1i16.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i16> @intrinsic_vwadd.w_wv_nxv1i16_nxv1i16_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vwadd.w_wv_nxv1i16_nxv1i16_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vwadd.wv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vwadd.w_wv_nxv1i16_nxv1i16_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vwadd.wv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwadd.w.nxv1i16.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwaddu.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i16> @intrinsic_vwaddu_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vwaddu_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vwaddu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vwaddu_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vwaddu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwaddu.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwmul.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i16> @intrinsic_vwmul_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vwmul_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vwmul.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vwmul_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vwmul.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwmul.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwmulu.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i16> @intrinsic_vwmulu_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vwmulu_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vwmulu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vwmulu_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vwmulu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwmulu.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwmulsu.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i16> @intrinsic_vwmulsu_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vwmulsu_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vwmulsu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vwmulsu_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vwmulsu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwmulsu.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsub.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i16> @intrinsic_vwsub_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vwsub_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vwsub.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vwsub_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vwsub.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsub.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsub.w.nxv1i16.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i16> @intrinsic_vwsub.w_wv_nxv1i16_nxv1i16_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vwsub.w_wv_nxv1i16_nxv1i16_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vwsub.wv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vwsub.w_wv_nxv1i16_nxv1i16_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vwsub.wv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsub.w.nxv1i16.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i16> %a
+}
+
+define <vscale x 1 x i16> @intrinsic_vwsub.w_wv_nxv1i16_nxv1i16_nxv1i8_tied(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, iXLen %2) nounwind {
+; RV32-LABEL: intrinsic_vwsub.w_wv_nxv1i16_nxv1i16_nxv1i8_tied:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vwsub.wv v8, v8, v9
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vwsub.w_wv_nxv1i16_nxv1i16_nxv1i8_tied:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vwsub.wv v8, v8, v9
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsub.w.nxv1i16.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    iXLen %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsubu.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i16> @intrinsic_vwsubu_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vwsubu_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vwsubu.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vwsubu_vv_nxv1i16_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vwsubu.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsubu.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsubu.w.nxv1i16.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i16> @intrinsic_vwsubu.w_wv_nxv1i16_nxv1i16_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vwsubu.w_wv_nxv1i16_nxv1i16_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vwsubu.wv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vwsubu.w_wv_nxv1i16_nxv1i16_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vwsubu.wv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsubu.w.nxv1i16.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i8> @llvm.riscv.vxor.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  iXLen);
+
+define <vscale x 1 x i8> @intrinsic_vxor_vv_nxv1i8_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, iXLen %3) nounwind {
+; RV32-LABEL: intrinsic_vxor_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV32:       # %bb.0: # %entry
+; RV32-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV32-NEXT:    vxor.vv v8, v9, v10
+; RV32-NEXT:    ret
+;
+; RV64-LABEL: intrinsic_vxor_vv_nxv1i8_nxv1i8_nxv1i8:
+; RV64:       # %bb.0: # %entry
+; RV64-NEXT:    vsetvli zero, a0, e8, mf8, tu, mu
+; RV64-NEXT:    vxor.vv v8, v9, v10
+; RV64-NEXT:    ret
+entry:
+  %a = call <vscale x 1 x i8> @llvm.riscv.vxor.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    iXLen %3)
+
+  ret <vscale x 1 x i8> %a
+}