[RISCV] Define vwadd/vwaddu/vwsub/vwsubu intrinsics.

Define vwadd/vwaddu/vwsub/vwsubu intrinsics and lower to V instructions.

Authored-by: Roger Ferrer Ibanez <rofirrim@gmail.com>
Co-Authored-by: Hsiangkai Wang <kai.wang@sifive.com>

Differential Revision: https://reviews.llvm.org/D93108

llvm/include/llvm/IR/IntrinsicsRISCV.td

@@ -96,13 +96,44 @@ let TargetPrefix = "riscv" in {
                    [IntrNoMem]>, RISCVVIntrinsic {
     let ExtendOperand = 3;
   }
+  // For destination vector type is NOT the same as first source vector.
+  // Input: (vector_in, vector_in/scalar_in, vl)
+  class RISCVBinaryABXNoMask
+        : Intrinsic<[llvm_anyvector_ty],
+                    [llvm_anyvector_ty, llvm_any_ty, llvm_anyint_ty],
+                    [IntrNoMem]>, RISCVVIntrinsic {
+    let ExtendOperand = 2;
+  }
+  // For destination vector type is NOT the same as first source vector (with mask).
+  // Input: (maskedoff, vector_in, vector_in/scalar_in, mask, vl)
+  class RISCVBinaryABXMask
+        : Intrinsic<[llvm_anyvector_ty],
+                    [LLVMMatchType<0>, llvm_anyvector_ty, llvm_any_ty,
+                     llvm_anyvector_ty, llvm_anyint_ty],
+                    [IntrNoMem]>, RISCVVIntrinsic {
+    let ExtendOperand = 3;
+  }
 
   multiclass RISCVBinaryAAX {
     def "int_riscv_" # NAME : RISCVBinaryAAXNoMask;
     def "int_riscv_" # NAME # "_mask" : RISCVBinaryAAXMask;
   }
+  multiclass RISCVBinaryABX {
+    def "int_riscv_" # NAME : RISCVBinaryABXNoMask;
+    def "int_riscv_" # NAME # "_mask" : RISCVBinaryABXMask;
+  }
 
   defm vadd : RISCVBinaryAAX;
   defm vsub : RISCVBinaryAAX;
   defm vrsub : RISCVBinaryAAX;
+
+  defm vwaddu : RISCVBinaryABX;
+  defm vwadd : RISCVBinaryABX;
+  defm vwaddu_w : RISCVBinaryAAX;
+  defm vwadd_w : RISCVBinaryAAX;
+  defm vwsubu : RISCVBinaryABX;
+  defm vwsub : RISCVBinaryABX;
+  defm vwsubu_w : RISCVBinaryAAX;
+  defm vwsub_w : RISCVBinaryAAX;
+
 } // TargetPrefix = "riscv"

llvm/lib/Target/RISCV/RISCVInstrInfoVPseudos.td

@@ -36,21 +36,22 @@ def NoX0 : SDNodeXForm<undef,
 //===----------------------------------------------------------------------===//
 
 // This class describes information associated to the LMUL.
-class LMULInfo<int lmul, VReg regclass, string mx> {
+class LMULInfo<int lmul, VReg regclass, VReg wregclass, string mx> {
   bits<3> value = lmul; // This is encoded as the vlmul field of vtype.
   VReg vrclass = regclass;
+  VReg wvrclass = wregclass;
   string MX = mx;
 }
 
 // Associate LMUL with tablegen records of register classes.
-def V_M1  : LMULInfo<0b000,   VR,  "M1">;
-def V_M2  : LMULInfo<0b001, VRM2,  "M2">;
-def V_M4  : LMULInfo<0b010, VRM4,  "M4">;
-def V_M8  : LMULInfo<0b011, VRM8,  "M8">;
+def V_M1  : LMULInfo<0b000,   VR,   VRM2, "M1">;
+def V_M2  : LMULInfo<0b001, VRM2,   VRM4, "M2">;
+def V_M4  : LMULInfo<0b010, VRM4,   VRM8, "M4">;
+def V_M8  : LMULInfo<0b011, VRM8, NoVReg, "M8">;
 
-def V_MF8 : LMULInfo<0b101,   VR, "MF8">;
-def V_MF4 : LMULInfo<0b110,   VR, "MF4">;
-def V_MF2 : LMULInfo<0b111,   VR, "MF2">;
+def V_MF8 : LMULInfo<0b101,   VR, VR, "MF8">;
+def V_MF4 : LMULInfo<0b110,   VR, VR, "MF4">;
+def V_MF2 : LMULInfo<0b111,   VR, VR, "MF2">;
 
 // Used to iterate over all possible LMULs.
 def MxList {
@@ -101,35 +102,61 @@ class GroupVTypeInfo<ValueType Vec, ValueType VecM1, ValueType Mas,
 
 defset list<VTypeInfo> AllVectors = {
   defset list<VTypeInfo> AllIntegerVectors = {
-    def : VTypeInfo<vint8mf8_t,  vbool64_t,  8, VR, V_MF8>;
-    def : VTypeInfo<vint8mf4_t,  vbool32_t,  8, VR, V_MF4>;
-    def : VTypeInfo<vint8mf2_t,  vbool16_t,  8, VR, V_MF2>;
-    def : VTypeInfo<vint8m1_t,   vbool8_t,   8, VR, V_M1>;
-    def : VTypeInfo<vint16mf4_t, vbool64_t, 16, VR, V_MF4>;
-    def : VTypeInfo<vint16mf2_t, vbool32_t, 16, VR, V_MF2>;
-    def : VTypeInfo<vint16m1_t,  vbool16_t, 16, VR, V_M1>;
-    def : VTypeInfo<vint32mf2_t, vbool64_t, 32, VR, V_MF2>;
-    def : VTypeInfo<vint32m1_t,  vbool32_t, 32, VR, V_M1>;
-    def : VTypeInfo<vint64m1_t,  vbool64_t, 64, VR, V_M1>;
+    def VI8MF8: VTypeInfo<vint8mf8_t,  vbool64_t,  8, VR, V_MF8>;
+    def VI8MF4: VTypeInfo<vint8mf4_t,  vbool32_t,  8, VR, V_MF4>;
+    def VI8MF2: VTypeInfo<vint8mf2_t,  vbool16_t,  8, VR, V_MF2>;
+    def VI8M1: VTypeInfo<vint8m1_t,   vbool8_t,   8, VR, V_M1>;
+    def VI16MF4: VTypeInfo<vint16mf4_t, vbool64_t, 16, VR, V_MF4>;
+    def VI16MF2: VTypeInfo<vint16mf2_t, vbool32_t, 16, VR, V_MF2>;
+    def VI16M1: VTypeInfo<vint16m1_t,  vbool16_t, 16, VR, V_M1>;
+    def VI32MF2: VTypeInfo<vint32mf2_t, vbool64_t, 32, VR, V_MF2>;
+    def VI32M1: VTypeInfo<vint32m1_t,  vbool32_t, 32, VR, V_M1>;
+    def VI64M1: VTypeInfo<vint64m1_t,  vbool64_t, 64, VR, V_M1>;
 
-    def : GroupVTypeInfo<vint8m2_t, vint8m1_t, vbool4_t, 8, VRM2, V_M2>;
-    def : GroupVTypeInfo<vint8m4_t, vint8m1_t, vbool2_t, 8, VRM4, V_M4>;
-    def : GroupVTypeInfo<vint8m8_t, vint8m1_t, vbool1_t, 8, VRM8, V_M8>;
+    def VI8M2: GroupVTypeInfo<vint8m2_t, vint8m1_t, vbool4_t, 8, VRM2, V_M2>;
+    def VI8M4: GroupVTypeInfo<vint8m4_t, vint8m1_t, vbool2_t, 8, VRM4, V_M4>;
+    def VI8M8: GroupVTypeInfo<vint8m8_t, vint8m1_t, vbool1_t, 8, VRM8, V_M8>;
 
-    def : GroupVTypeInfo<vint16m2_t,vint16m1_t,vbool8_t, 16,VRM2, V_M2>;
-    def : GroupVTypeInfo<vint16m4_t,vint16m1_t,vbool4_t, 16,VRM4, V_M4>;
-    def : GroupVTypeInfo<vint16m8_t,vint16m1_t,vbool2_t, 16,VRM8, V_M8>;
+    def VI16M2: GroupVTypeInfo<vint16m2_t,vint16m1_t,vbool8_t, 16,VRM2, V_M2>;
+    def VI16M4: GroupVTypeInfo<vint16m4_t,vint16m1_t,vbool4_t, 16,VRM4, V_M4>;
+    def VI16M8: GroupVTypeInfo<vint16m8_t,vint16m1_t,vbool2_t, 16,VRM8, V_M8>;
 
-    def : GroupVTypeInfo<vint32m2_t,vint32m1_t,vbool16_t,32,VRM2, V_M2>;
-    def : GroupVTypeInfo<vint32m4_t,vint32m1_t,vbool8_t, 32,VRM4, V_M4>;
-    def : GroupVTypeInfo<vint32m8_t,vint32m1_t,vbool4_t, 32,VRM8, V_M8>;
+    def VI32M2: GroupVTypeInfo<vint32m2_t,vint32m1_t,vbool16_t,32,VRM2, V_M2>;
+    def VI32M4: GroupVTypeInfo<vint32m4_t,vint32m1_t,vbool8_t, 32,VRM4, V_M4>;
+    def VI32M8: GroupVTypeInfo<vint32m8_t,vint32m1_t,vbool4_t, 32,VRM8, V_M8>;
 
-    def : GroupVTypeInfo<vint64m2_t,vint64m1_t,vbool32_t,64,VRM2, V_M2>;
-    def : GroupVTypeInfo<vint64m4_t,vint64m1_t,vbool16_t,64,VRM4, V_M4>;
-    def : GroupVTypeInfo<vint64m8_t,vint64m1_t,vbool8_t, 64,VRM8, V_M8>;
+    def VI64M2: GroupVTypeInfo<vint64m2_t,vint64m1_t,vbool32_t,64,VRM2, V_M2>;
+    def VI64M4: GroupVTypeInfo<vint64m4_t,vint64m1_t,vbool16_t,64,VRM4, V_M4>;
+    def VI64M8: GroupVTypeInfo<vint64m8_t,vint64m1_t,vbool8_t, 64,VRM8, V_M8>;
   }
 }
 
+class VTypeInfoToWide<VTypeInfo vti, VTypeInfo wti>
+{
+  VTypeInfo Vti = vti;
+  VTypeInfo Wti = wti;
+}
+
+defset list<VTypeInfoToWide> AllWidenableIntVectors = {
+  def : VTypeInfoToWide<VI8MF8,  VI16MF4>;
+  def : VTypeInfoToWide<VI8MF4,  VI16MF2>;
+  def : VTypeInfoToWide<VI8MF2,  VI16M1>;
+  def : VTypeInfoToWide<VI8M1,   VI16M2>;
+  def : VTypeInfoToWide<VI8M2,   VI16M4>;
+  def : VTypeInfoToWide<VI8M4,   VI16M8>;
+
+  def : VTypeInfoToWide<VI16MF4, VI32MF2>;
+  def : VTypeInfoToWide<VI16MF2, VI32M1>;
+  def : VTypeInfoToWide<VI16M1,  VI32M2>;
+  def : VTypeInfoToWide<VI16M2,  VI32M4>;
+  def : VTypeInfoToWide<VI16M4,  VI32M8>;
+
+  def : VTypeInfoToWide<VI32MF2, VI64M1>;
+  def : VTypeInfoToWide<VI32M1,  VI64M2>;
+  def : VTypeInfoToWide<VI32M2,  VI64M4>;
+  def : VTypeInfoToWide<VI32M4,  VI64M8>;
+}
+
 // This class holds the record of the RISCVVPseudoTable below.
 // This represents the information we need in codegen for each pseudo.
 // The definition should be consistent with `struct PseudoInfo` in
@@ -194,6 +221,16 @@ class GetVRegNoV0<VReg VRegClass> {
                  !eq(1, 1) : VRegClass);
 }
 
+// Join strings in list using separator and ignoring empty elements
+class Join<list<string> strings, string separator> {
+  string ret = !foldl(!head(strings), !tail(strings), a, b,
+                      !cond(
+                        !and(!empty(a), !empty(b)) : "",
+                        !empty(a) : b,
+                        !empty(b) : a,
+                        1 : a#separator#b));
+}
+
 class VPseudo<Instruction instr, LMULInfo m, dag outs, dag ins> :
       Pseudo<outs, ins, []>, RISCVVPseudo {
   let BaseInstr = instr;
@@ -202,7 +239,8 @@ class VPseudo<Instruction instr, LMULInfo m, dag outs, dag ins> :
 
 class VPseudoBinaryNoMask<VReg RetClass,
                           VReg Op1Class,
-                          DAGOperand Op2Class> :
+                          DAGOperand Op2Class,
+                          string Constraint> :
         Pseudo<(outs RetClass:$rd),
                (ins Op1Class:$rs2, Op2Class:$rs1, GPR:$vl, ixlenimm:$sew), []>,
         RISCVVPseudo {
@@ -210,6 +248,7 @@ class VPseudoBinaryNoMask<VReg RetClass,
   let mayStore = 0;
   let hasSideEffects = 0;
   let usesCustomInserter = 1;
+  let Constraints = Constraint;
   let Uses = [VL, VTYPE];
   let VLIndex = 3;
   let SEWIndex = 4;
@@ -219,7 +258,8 @@ class VPseudoBinaryNoMask<VReg RetClass,
 
 class VPseudoBinaryMask<VReg RetClass,
                         VReg Op1Class,
-                        DAGOperand Op2Class> :
+                        DAGOperand Op2Class,
+                        string Constraint> :
         Pseudo<(outs GetVRegNoV0<RetClass>.R:$rd),
                 (ins GetVRegNoV0<RetClass>.R:$merge,
                      Op1Class:$rs2, Op2Class:$rs1,
@@ -229,7 +269,7 @@ class VPseudoBinaryMask<VReg RetClass,
   let mayStore = 0;
   let hasSideEffects = 0;
   let usesCustomInserter = 1;
-  let Constraints = "$rd = $merge";
+  let Constraints = Join<[Constraint, "$rd = $merge"], ",">.ret;
   let Uses = [VL, VTYPE];
   let VLIndex = 5;
   let SEWIndex = 6;
@@ -240,10 +280,13 @@ class VPseudoBinaryMask<VReg RetClass,
 multiclass VPseudoBinary<VReg RetClass,
                          VReg Op1Class,
                          DAGOperand Op2Class,
-                         LMULInfo MInfo> {
+                         LMULInfo MInfo,
+                         string Constraint = ""> {
   let VLMul = MInfo.value in {
-    def "_" # MInfo.MX : VPseudoBinaryNoMask<RetClass, Op1Class, Op2Class>;
-    def "_" # MInfo.MX # "_MASK" : VPseudoBinaryMask<RetClass, Op1Class, Op2Class>;
+    def "_" # MInfo.MX : VPseudoBinaryNoMask<RetClass, Op1Class, Op2Class,
+                                             Constraint>;
+    def "_" # MInfo.MX # "_MASK" : VPseudoBinaryMask<RetClass, Op1Class, Op2Class,
+                                                     Constraint>;
   }
 }
 
@@ -262,6 +305,37 @@ multiclass VPseudoBinaryV_VI<Operand ImmType = simm5> {
     defm _VI : VPseudoBinary<m.vrclass, m.vrclass, ImmType, m>;
 }
 
+// We use earlyclobber here due to
+// * The destination EEW is smaller than the source EEW and the overlap is
+//   in the lowest-numbered part of the source register group is legal.
+//   Otherwise, it is illegal.
+// * The destination EEW is greater than the source EEW, the source EMUL is
+//   at least 1, and the overlap is in the highest-numbered part of the
+//   destination register group is legal. Otherwise, it is illegal.
+multiclass VPseudoBinaryW_VV {
+  foreach m = MxList.m[0-5] in
+    defm _VV : VPseudoBinary<m.wvrclass, m.vrclass, m.vrclass, m,
+                             "@earlyclobber $rd">;
+}
+
+multiclass VPseudoBinaryW_VX {
+  foreach m = MxList.m[0-5] in
+    defm _VX : VPseudoBinary<m.wvrclass, m.vrclass, GPR, m,
+                             "@earlyclobber $rd">;
+}
+
+multiclass VPseudoBinaryW_WV {
+  foreach m = MxList.m[0-5] in
+    defm _WV : VPseudoBinary<m.wvrclass, m.wvrclass, m.vrclass, m,
+                             "@earlyclobber $rd">;
+}
+
+multiclass VPseudoBinaryW_WX {
+  foreach m = MxList.m[0-5] in
+    defm _WX : VPseudoBinary<m.wvrclass, m.wvrclass, GPR, m,
+                             "@earlyclobber $rd">;
+}
+
 multiclass VPseudoBinaryV_VV_VX_VI<Operand ImmType = simm5> {
   defm "" : VPseudoBinaryV_VV;
   defm "" : VPseudoBinaryV_VX;
@@ -278,6 +352,16 @@ multiclass VPseudoBinaryV_VX_VI<Operand ImmType = simm5> {
   defm "" : VPseudoBinaryV_VI<ImmType>;
 }
 
+multiclass VPseudoBinaryW_VV_VX {
+  defm "" : VPseudoBinaryW_VV;
+  defm "" : VPseudoBinaryW_VX;
+}
+
+multiclass VPseudoBinaryW_WV_WX {
+  defm "" : VPseudoBinaryW_WV;
+  defm "" : VPseudoBinaryW_WX;
+}
+
 //===----------------------------------------------------------------------===//
 // Helpers to define the different patterns.
 //===----------------------------------------------------------------------===//
@@ -396,6 +480,50 @@ multiclass VPatBinaryV_VI<string intrinsic, string instruction,
                       vti.RegClass, imm_type>;
 }
 
+multiclass VPatBinaryW_VV<string intrinsic, string instruction> {
+  foreach VtiToWti = AllWidenableIntVectors in {
+    defvar Vti = VtiToWti.Vti;
+    defvar Wti = VtiToWti.Wti;
+    defm : VPatBinary<intrinsic, instruction, "VV",
+                      Wti.Vector, Vti.Vector, Vti.Vector, Vti.Mask,
+                      Vti.SEW, Vti.LMul, Wti.RegClass,
+                      Vti.RegClass, Vti.RegClass>;
+  }
+}
+
+multiclass VPatBinaryW_VX<string intrinsic, string instruction> {
+  foreach VtiToWti = AllWidenableIntVectors in {
+    defvar Vti = VtiToWti.Vti;
+    defvar Wti = VtiToWti.Wti;
+    defm : VPatBinary<intrinsic, instruction, "VX",
+                      Wti.Vector, Vti.Vector, XLenVT, Vti.Mask,
+                      Vti.SEW, Vti.LMul, Wti.RegClass,
+                      Vti.RegClass, GPR>;
+  }
+}
+
+multiclass VPatBinaryW_WV<string intrinsic, string instruction> {
+  foreach VtiToWti = AllWidenableIntVectors in {
+    defvar Vti = VtiToWti.Vti;
+    defvar Wti = VtiToWti.Wti;
+    defm : VPatBinary<intrinsic, instruction, "WV",
+                      Wti.Vector, Wti.Vector, Vti.Vector, Vti.Mask,
+                      Vti.SEW, Vti.LMul, Wti.RegClass,
+                      Wti.RegClass, Vti.RegClass>;
+  }
+}
+
+multiclass VPatBinaryW_WX<string intrinsic, string instruction> {
+  foreach VtiToWti = AllWidenableIntVectors in {
+    defvar Vti = VtiToWti.Vti;
+    defvar Wti = VtiToWti.Wti;
+    defm : VPatBinary<intrinsic, instruction, "WX",
+                      Wti.Vector, Wti.Vector, XLenVT, Vti.Mask,
+                      Vti.SEW, Vti.LMul, Wti.RegClass,
+                      Wti.RegClass, GPR>;
+  }
+}
+
 multiclass VPatBinaryV_VV_VX_VI<string intrinsic, string instruction,
                                 list<VTypeInfo> vtilist>
 {
@@ -418,6 +546,18 @@ multiclass VPatBinaryV_VX_VI<string intrinsic, string instruction,
   defm "" : VPatBinaryV_VI<intrinsic, instruction, vtilist, simm5>;
 }
 
+multiclass VPatBinaryW_VV_VX<string intrinsic, string instruction>
+{
+  defm "" : VPatBinaryW_VV<intrinsic, instruction>;
+  defm "" : VPatBinaryW_VX<intrinsic, instruction>;
+}
+
+multiclass VPatBinaryW_WV_WX<string intrinsic, string instruction>
+{
+  defm "" : VPatBinaryW_WV<intrinsic, instruction>;
+  defm "" : VPatBinaryW_WX<intrinsic, instruction>;
+}
+
 //===----------------------------------------------------------------------===//
 // Pseudo instructions and patterns.
 //===----------------------------------------------------------------------===//
@@ -542,6 +682,19 @@ defm PseudoVADD        : VPseudoBinaryV_VV_VX_VI;
 defm PseudoVSUB        : VPseudoBinaryV_VV_VX;
 defm PseudoVRSUB       : VPseudoBinaryV_VX_VI;
 
+//===----------------------------------------------------------------------===//
+// 12.2. Vector Widening Integer Add/Subtract
+//===----------------------------------------------------------------------===//
+defm PseudoVWADDU    : VPseudoBinaryW_VV_VX;
+defm PseudoVWSUBU    : VPseudoBinaryW_VV_VX;
+defm PseudoVWADD     : VPseudoBinaryW_VV_VX;
+defm PseudoVWSUB     : VPseudoBinaryW_VV_VX;
+defm PseudoVWADDU    : VPseudoBinaryW_WV_WX;
+defm PseudoVWSUBU    : VPseudoBinaryW_WV_WX;
+defm PseudoVWADD     : VPseudoBinaryW_WV_WX;
+defm PseudoVWSUB     : VPseudoBinaryW_WV_WX;
+
+
 //===----------------------------------------------------------------------===//
 // Patterns.
 //===----------------------------------------------------------------------===//
@@ -560,4 +713,16 @@ defm "" : VPatBinaryV_VV_VX_VI<"int_riscv_vadd", "PseudoVADD", AllIntegerVectors
 defm "" : VPatBinaryV_VV_VX<"int_riscv_vsub", "PseudoVSUB", AllIntegerVectors>;
 defm "" : VPatBinaryV_VX_VI<"int_riscv_vrsub", "PseudoVRSUB", AllIntegerVectors>;
 
+//===----------------------------------------------------------------------===//
+// 12.2. Vector Widening Integer Add/Subtract
+//===----------------------------------------------------------------------===//
+defm "" : VPatBinaryW_VV_VX<"int_riscv_vwaddu", "PseudoVWADDU">;
+defm "" : VPatBinaryW_VV_VX<"int_riscv_vwsubu", "PseudoVWSUBU">;
+defm "" : VPatBinaryW_VV_VX<"int_riscv_vwadd", "PseudoVWADD">;
+defm "" : VPatBinaryW_VV_VX<"int_riscv_vwsub", "PseudoVWSUB">;
+defm "" : VPatBinaryW_WV_WX<"int_riscv_vwaddu_w", "PseudoVWADDU">;
+defm "" : VPatBinaryW_WV_WX<"int_riscv_vwsubu_w", "PseudoVWSUBU">;
+defm "" : VPatBinaryW_WV_WX<"int_riscv_vwadd_w", "PseudoVWADD">;
+defm "" : VPatBinaryW_WV_WX<"int_riscv_vwsub_w", "PseudoVWSUB">;
+
 } // Predicates = [HasStdExtV]

llvm/lib/Target/RISCV/RISCVRegisterInfo.td

@@ -395,6 +395,9 @@ class VReg<list<ValueType> regTypes, dag regList, int Vlmul>
   int Size = !mul(Vlmul, 64); // FIXME: assuming ELEN=64
 }
 
+// Dummy V register class.
+def NoVReg : VReg<[vint8m1_t], (add V0), 0>;
+
 def VR : VReg<[vint8mf2_t, vint8mf4_t, vint8mf8_t,
                vint16mf2_t, vint16mf4_t, vint32mf2_t,
                vint8m1_t, vint16m1_t, vint32m1_t, vint64m1_t,

llvm/test/CodeGen/RISCV/rvv/vwadd-rv32.ll

@@ -0,0 +1,881 @@
+; RUN: llc -mtriple=riscv32 -mattr=+experimental-v -verify-machineinstrs \
+; RUN:   --riscv-no-aliases < %s | FileCheck %s
+declare <vscale x 1 x i16> @llvm.riscv.vwadd.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwadd_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vv_nxv1i16_nxv1i8_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwadd.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwadd.mask.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwadd_mask_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vv_nxv1i16_nxv1i8_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwadd.mask.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwadd.nxv2i16.nxv2i8.nxv2i8(
+  <vscale x 2 x i8>,
+  <vscale x 2 x i8>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwadd_vv_nxv2i16_nxv2i8_nxv2i8(<vscale x 2 x i8> %0, <vscale x 2 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vv_nxv2i16_nxv2i8_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwadd.nxv2i16.nxv2i8.nxv2i8(
+    <vscale x 2 x i8> %0,
+    <vscale x 2 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwadd.mask.nxv2i16.nxv2i8.nxv2i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  <vscale x 2 x i8>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwadd_mask_vv_nxv2i16_nxv2i8_nxv2i8(<vscale x 2 x i16> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vv_nxv2i16_nxv2i8_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwadd.mask.nxv2i16.nxv2i8.nxv2i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i8> %1,
+    <vscale x 2 x i8> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwadd.nxv4i16.nxv4i8.nxv4i8(
+  <vscale x 4 x i8>,
+  <vscale x 4 x i8>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwadd_vv_nxv4i16_nxv4i8_nxv4i8(<vscale x 4 x i8> %0, <vscale x 4 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vv_nxv4i16_nxv4i8_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwadd.nxv4i16.nxv4i8.nxv4i8(
+    <vscale x 4 x i8> %0,
+    <vscale x 4 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwadd.mask.nxv4i16.nxv4i8.nxv4i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  <vscale x 4 x i8>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwadd_mask_vv_nxv4i16_nxv4i8_nxv4i8(<vscale x 4 x i16> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vv_nxv4i16_nxv4i8_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwadd.mask.nxv4i16.nxv4i8.nxv4i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i8> %1,
+    <vscale x 4 x i8> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwadd.nxv8i16.nxv8i8.nxv8i8(
+  <vscale x 8 x i8>,
+  <vscale x 8 x i8>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwadd_vv_nxv8i16_nxv8i8_nxv8i8(<vscale x 8 x i8> %0, <vscale x 8 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vv_nxv8i16_nxv8i8_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwadd.nxv8i16.nxv8i8.nxv8i8(
+    <vscale x 8 x i8> %0,
+    <vscale x 8 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwadd.mask.nxv8i16.nxv8i8.nxv8i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  <vscale x 8 x i8>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwadd_mask_vv_nxv8i16_nxv8i8_nxv8i8(<vscale x 8 x i16> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vv_nxv8i16_nxv8i8_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwadd.mask.nxv8i16.nxv8i8.nxv8i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i8> %1,
+    <vscale x 8 x i8> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwadd.nxv16i16.nxv16i8.nxv16i8(
+  <vscale x 16 x i8>,
+  <vscale x 16 x i8>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwadd_vv_nxv16i16_nxv16i8_nxv16i8(<vscale x 16 x i8> %0, <vscale x 16 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vv_nxv16i16_nxv16i8_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwadd.nxv16i16.nxv16i8.nxv16i8(
+    <vscale x 16 x i8> %0,
+    <vscale x 16 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwadd.mask.nxv16i16.nxv16i8.nxv16i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  <vscale x 16 x i8>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwadd_mask_vv_nxv16i16_nxv16i8_nxv16i8(<vscale x 16 x i16> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vv_nxv16i16_nxv16i8_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwadd.mask.nxv16i16.nxv16i8.nxv16i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i8> %1,
+    <vscale x 16 x i8> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwadd.nxv32i16.nxv32i8.nxv32i8(
+  <vscale x 32 x i8>,
+  <vscale x 32 x i8>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwadd_vv_nxv32i16_nxv32i8_nxv32i8(<vscale x 32 x i8> %0, <vscale x 32 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vv_nxv32i16_nxv32i8_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwadd.nxv32i16.nxv32i8.nxv32i8(
+    <vscale x 32 x i8> %0,
+    <vscale x 32 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwadd.mask.nxv32i16.nxv32i8.nxv32i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  <vscale x 32 x i8>,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwadd_mask_vv_nxv32i16_nxv32i8_nxv32i8(<vscale x 32 x i16> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vv_nxv32i16_nxv32i8_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwadd.mask.nxv32i16.nxv32i8.nxv32i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i8> %1,
+    <vscale x 32 x i8> %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwadd.nxv1i32.nxv1i16.nxv1i16(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwadd_vv_nxv1i32_nxv1i16_nxv1i16(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vv_nxv1i32_nxv1i16_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwadd.nxv1i32.nxv1i16.nxv1i16(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwadd.mask.nxv1i32.nxv1i16.nxv1i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwadd_mask_vv_nxv1i32_nxv1i16_nxv1i16(<vscale x 1 x i32> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vv_nxv1i32_nxv1i16_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwadd.mask.nxv1i32.nxv1i16.nxv1i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i16> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwadd.nxv2i32.nxv2i16.nxv2i16(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwadd_vv_nxv2i32_nxv2i16_nxv2i16(<vscale x 2 x i16> %0, <vscale x 2 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vv_nxv2i32_nxv2i16_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwadd.nxv2i32.nxv2i16.nxv2i16(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwadd.mask.nxv2i32.nxv2i16.nxv2i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwadd_mask_vv_nxv2i32_nxv2i16_nxv2i16(<vscale x 2 x i32> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vv_nxv2i32_nxv2i16_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwadd.mask.nxv2i32.nxv2i16.nxv2i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i16> %1,
+    <vscale x 2 x i16> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwadd.nxv4i32.nxv4i16.nxv4i16(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwadd_vv_nxv4i32_nxv4i16_nxv4i16(<vscale x 4 x i16> %0, <vscale x 4 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vv_nxv4i32_nxv4i16_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwadd.nxv4i32.nxv4i16.nxv4i16(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwadd.mask.nxv4i32.nxv4i16.nxv4i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwadd_mask_vv_nxv4i32_nxv4i16_nxv4i16(<vscale x 4 x i32> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vv_nxv4i32_nxv4i16_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwadd.mask.nxv4i32.nxv4i16.nxv4i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i16> %1,
+    <vscale x 4 x i16> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwadd.nxv8i32.nxv8i16.nxv8i16(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwadd_vv_nxv8i32_nxv8i16_nxv8i16(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vv_nxv8i32_nxv8i16_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwadd.nxv8i32.nxv8i16.nxv8i16(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwadd.mask.nxv8i32.nxv8i16.nxv8i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwadd_mask_vv_nxv8i32_nxv8i16_nxv8i16(<vscale x 8 x i32> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vv_nxv8i32_nxv8i16_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwadd.mask.nxv8i32.nxv8i16.nxv8i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i16> %1,
+    <vscale x 8 x i16> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwadd.nxv16i32.nxv16i16.nxv16i16(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwadd_vv_nxv16i32_nxv16i16_nxv16i16(<vscale x 16 x i16> %0, <vscale x 16 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vv_nxv16i32_nxv16i16_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwadd.nxv16i32.nxv16i16.nxv16i16(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwadd.mask.nxv16i32.nxv16i16.nxv16i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwadd_mask_vv_nxv16i32_nxv16i16_nxv16i16(<vscale x 16 x i32> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vv_nxv16i32_nxv16i16_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwadd.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwadd.mask.nxv16i32.nxv16i16.nxv16i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i16> %1,
+    <vscale x 16 x i16> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwadd.nxv1i16.nxv1i8.i8(
+  <vscale x 1 x i8>,
+  i8,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwadd_vx_nxv1i16_nxv1i8_i8(<vscale x 1 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vx_nxv1i16_nxv1i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwadd.nxv1i16.nxv1i8.i8(
+    <vscale x 1 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwadd.mask.nxv1i16.nxv1i8.i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  i8,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwadd_mask_vx_nxv1i16_nxv1i8_i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vx_nxv1i16_nxv1i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwadd.mask.nxv1i16.nxv1i8.i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    i8 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwadd.nxv2i16.nxv2i8.i8(
+  <vscale x 2 x i8>,
+  i8,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwadd_vx_nxv2i16_nxv2i8_i8(<vscale x 2 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vx_nxv2i16_nxv2i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwadd.nxv2i16.nxv2i8.i8(
+    <vscale x 2 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwadd.mask.nxv2i16.nxv2i8.i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  i8,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwadd_mask_vx_nxv2i16_nxv2i8_i8(<vscale x 2 x i16> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vx_nxv2i16_nxv2i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwadd.mask.nxv2i16.nxv2i8.i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i8> %1,
+    i8 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwadd.nxv4i16.nxv4i8.i8(
+  <vscale x 4 x i8>,
+  i8,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwadd_vx_nxv4i16_nxv4i8_i8(<vscale x 4 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vx_nxv4i16_nxv4i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwadd.nxv4i16.nxv4i8.i8(
+    <vscale x 4 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwadd.mask.nxv4i16.nxv4i8.i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  i8,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwadd_mask_vx_nxv4i16_nxv4i8_i8(<vscale x 4 x i16> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vx_nxv4i16_nxv4i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwadd.mask.nxv4i16.nxv4i8.i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i8> %1,
+    i8 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwadd.nxv8i16.nxv8i8.i8(
+  <vscale x 8 x i8>,
+  i8,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwadd_vx_nxv8i16_nxv8i8_i8(<vscale x 8 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vx_nxv8i16_nxv8i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwadd.nxv8i16.nxv8i8.i8(
+    <vscale x 8 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwadd.mask.nxv8i16.nxv8i8.i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  i8,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwadd_mask_vx_nxv8i16_nxv8i8_i8(<vscale x 8 x i16> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vx_nxv8i16_nxv8i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwadd.mask.nxv8i16.nxv8i8.i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i8> %1,
+    i8 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwadd.nxv16i16.nxv16i8.i8(
+  <vscale x 16 x i8>,
+  i8,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwadd_vx_nxv16i16_nxv16i8_i8(<vscale x 16 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vx_nxv16i16_nxv16i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwadd.nxv16i16.nxv16i8.i8(
+    <vscale x 16 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwadd.mask.nxv16i16.nxv16i8.i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  i8,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwadd_mask_vx_nxv16i16_nxv16i8_i8(<vscale x 16 x i16> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vx_nxv16i16_nxv16i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwadd.mask.nxv16i16.nxv16i8.i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i8> %1,
+    i8 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwadd.nxv32i16.nxv32i8.i8(
+  <vscale x 32 x i8>,
+  i8,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwadd_vx_nxv32i16_nxv32i8_i8(<vscale x 32 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vx_nxv32i16_nxv32i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwadd.nxv32i16.nxv32i8.i8(
+    <vscale x 32 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwadd.mask.nxv32i16.nxv32i8.i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  i8,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwadd_mask_vx_nxv32i16_nxv32i8_i8(<vscale x 32 x i16> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vx_nxv32i16_nxv32i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwadd.mask.nxv32i16.nxv32i8.i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i8> %1,
+    i8 %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwadd.nxv1i32.nxv1i16.i16(
+  <vscale x 1 x i16>,
+  i16,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwadd_vx_nxv1i32_nxv1i16_i16(<vscale x 1 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vx_nxv1i32_nxv1i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwadd.nxv1i32.nxv1i16.i16(
+    <vscale x 1 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwadd.mask.nxv1i32.nxv1i16.i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  i16,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwadd_mask_vx_nxv1i32_nxv1i16_i16(<vscale x 1 x i32> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vx_nxv1i32_nxv1i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwadd.mask.nxv1i32.nxv1i16.i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i16> %1,
+    i16 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwadd.nxv2i32.nxv2i16.i16(
+  <vscale x 2 x i16>,
+  i16,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwadd_vx_nxv2i32_nxv2i16_i16(<vscale x 2 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vx_nxv2i32_nxv2i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwadd.nxv2i32.nxv2i16.i16(
+    <vscale x 2 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwadd.mask.nxv2i32.nxv2i16.i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  i16,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwadd_mask_vx_nxv2i32_nxv2i16_i16(<vscale x 2 x i32> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vx_nxv2i32_nxv2i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwadd.mask.nxv2i32.nxv2i16.i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i16> %1,
+    i16 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwadd.nxv4i32.nxv4i16.i16(
+  <vscale x 4 x i16>,
+  i16,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwadd_vx_nxv4i32_nxv4i16_i16(<vscale x 4 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vx_nxv4i32_nxv4i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwadd.nxv4i32.nxv4i16.i16(
+    <vscale x 4 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwadd.mask.nxv4i32.nxv4i16.i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  i16,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwadd_mask_vx_nxv4i32_nxv4i16_i16(<vscale x 4 x i32> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vx_nxv4i32_nxv4i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwadd.mask.nxv4i32.nxv4i16.i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i16> %1,
+    i16 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwadd.nxv8i32.nxv8i16.i16(
+  <vscale x 8 x i16>,
+  i16,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwadd_vx_nxv8i32_nxv8i16_i16(<vscale x 8 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vx_nxv8i32_nxv8i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwadd.nxv8i32.nxv8i16.i16(
+    <vscale x 8 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwadd.mask.nxv8i32.nxv8i16.i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  i16,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwadd_mask_vx_nxv8i32_nxv8i16_i16(<vscale x 8 x i32> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vx_nxv8i32_nxv8i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwadd.mask.nxv8i32.nxv8i16.i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i16> %1,
+    i16 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwadd.nxv16i32.nxv16i16.i16(
+  <vscale x 16 x i16>,
+  i16,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwadd_vx_nxv16i32_nxv16i16_i16(<vscale x 16 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_vx_nxv16i32_nxv16i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwadd.nxv16i32.nxv16i16.i16(
+    <vscale x 16 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwadd.mask.nxv16i32.nxv16i16.i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  i16,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwadd_mask_vx_nxv16i32_nxv16i16_i16(<vscale x 16 x i32> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd_mask_vx_nxv16i32_nxv16i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwadd.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwadd.mask.nxv16i32.nxv16i16.i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i16> %1,
+    i16 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}

llvm/test/CodeGen/RISCV/rvv/vwadd.w-rv32.ll

@@ -0,0 +1,881 @@
+; RUN: llc -mtriple=riscv32 -mattr=+experimental-v -verify-machineinstrs \
+; RUN:   --riscv-no-aliases < %s | FileCheck %s
+declare <vscale x 1 x i16> @llvm.riscv.vwadd.w.nxv1i16.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwadd.w_wv_nxv1i16_nxv1i16_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wv_nxv1i16_nxv1i16_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwadd.w.nxv1i16.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwadd.w.mask.nxv1i16.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwadd.w_mask_wv_nxv1i16_nxv1i16_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, <vscale x 1 x i8> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wv_nxv1i16_nxv1i16_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwadd.w.mask.nxv1i16.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i8> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwadd.w.nxv2i16.nxv2i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwadd.w_wv_nxv2i16_nxv2i16_nxv2i8(<vscale x 2 x i16> %0, <vscale x 2 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wv_nxv2i16_nxv2i16_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwadd.w.nxv2i16.nxv2i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwadd.w.mask.nxv2i16.nxv2i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwadd.w_mask_wv_nxv2i16_nxv2i16_nxv2i8(<vscale x 2 x i16> %0, <vscale x 2 x i16> %1, <vscale x 2 x i8> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wv_nxv2i16_nxv2i16_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwadd.w.mask.nxv2i16.nxv2i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i16> %1,
+    <vscale x 2 x i8> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwadd.w.nxv4i16.nxv4i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwadd.w_wv_nxv4i16_nxv4i16_nxv4i8(<vscale x 4 x i16> %0, <vscale x 4 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wv_nxv4i16_nxv4i16_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwadd.w.nxv4i16.nxv4i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwadd.w.mask.nxv4i16.nxv4i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwadd.w_mask_wv_nxv4i16_nxv4i16_nxv4i8(<vscale x 4 x i16> %0, <vscale x 4 x i16> %1, <vscale x 4 x i8> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wv_nxv4i16_nxv4i16_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwadd.w.mask.nxv4i16.nxv4i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i16> %1,
+    <vscale x 4 x i8> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwadd.w.nxv8i16.nxv8i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwadd.w_wv_nxv8i16_nxv8i16_nxv8i8(<vscale x 8 x i16> %0, <vscale x 8 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wv_nxv8i16_nxv8i16_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwadd.w.nxv8i16.nxv8i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwadd.w.mask.nxv8i16.nxv8i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwadd.w_mask_wv_nxv8i16_nxv8i16_nxv8i8(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, <vscale x 8 x i8> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wv_nxv8i16_nxv8i16_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwadd.w.mask.nxv8i16.nxv8i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i16> %1,
+    <vscale x 8 x i8> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwadd.w.nxv16i16.nxv16i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwadd.w_wv_nxv16i16_nxv16i16_nxv16i8(<vscale x 16 x i16> %0, <vscale x 16 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wv_nxv16i16_nxv16i16_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwadd.w.nxv16i16.nxv16i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwadd.w.mask.nxv16i16.nxv16i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwadd.w_mask_wv_nxv16i16_nxv16i16_nxv16i8(<vscale x 16 x i16> %0, <vscale x 16 x i16> %1, <vscale x 16 x i8> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wv_nxv16i16_nxv16i16_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwadd.w.mask.nxv16i16.nxv16i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i16> %1,
+    <vscale x 16 x i8> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwadd.w.nxv32i16.nxv32i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwadd.w_wv_nxv32i16_nxv32i16_nxv32i8(<vscale x 32 x i16> %0, <vscale x 32 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wv_nxv32i16_nxv32i16_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwadd.w.nxv32i16.nxv32i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwadd.w.mask.nxv32i16.nxv32i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwadd.w_mask_wv_nxv32i16_nxv32i16_nxv32i8(<vscale x 32 x i16> %0, <vscale x 32 x i16> %1, <vscale x 32 x i8> %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wv_nxv32i16_nxv32i16_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwadd.w.mask.nxv32i16.nxv32i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i16> %1,
+    <vscale x 32 x i8> %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwadd.w.nxv1i32.nxv1i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwadd.w_wv_nxv1i32_nxv1i32_nxv1i16(<vscale x 1 x i32> %0, <vscale x 1 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wv_nxv1i32_nxv1i32_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwadd.w.nxv1i32.nxv1i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwadd.w.mask.nxv1i32.nxv1i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwadd.w_mask_wv_nxv1i32_nxv1i32_nxv1i16(<vscale x 1 x i32> %0, <vscale x 1 x i32> %1, <vscale x 1 x i16> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wv_nxv1i32_nxv1i32_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwadd.w.mask.nxv1i32.nxv1i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i32> %1,
+    <vscale x 1 x i16> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwadd.w.nxv2i32.nxv2i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwadd.w_wv_nxv2i32_nxv2i32_nxv2i16(<vscale x 2 x i32> %0, <vscale x 2 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wv_nxv2i32_nxv2i32_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwadd.w.nxv2i32.nxv2i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwadd.w.mask.nxv2i32.nxv2i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwadd.w_mask_wv_nxv2i32_nxv2i32_nxv2i16(<vscale x 2 x i32> %0, <vscale x 2 x i32> %1, <vscale x 2 x i16> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wv_nxv2i32_nxv2i32_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwadd.w.mask.nxv2i32.nxv2i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i32> %1,
+    <vscale x 2 x i16> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwadd.w.nxv4i32.nxv4i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwadd.w_wv_nxv4i32_nxv4i32_nxv4i16(<vscale x 4 x i32> %0, <vscale x 4 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wv_nxv4i32_nxv4i32_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwadd.w.nxv4i32.nxv4i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwadd.w.mask.nxv4i32.nxv4i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwadd.w_mask_wv_nxv4i32_nxv4i32_nxv4i16(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, <vscale x 4 x i16> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wv_nxv4i32_nxv4i32_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwadd.w.mask.nxv4i32.nxv4i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i32> %1,
+    <vscale x 4 x i16> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwadd.w.nxv8i32.nxv8i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwadd.w_wv_nxv8i32_nxv8i32_nxv8i16(<vscale x 8 x i32> %0, <vscale x 8 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wv_nxv8i32_nxv8i32_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwadd.w.nxv8i32.nxv8i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwadd.w.mask.nxv8i32.nxv8i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwadd.w_mask_wv_nxv8i32_nxv8i32_nxv8i16(<vscale x 8 x i32> %0, <vscale x 8 x i32> %1, <vscale x 8 x i16> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wv_nxv8i32_nxv8i32_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwadd.w.mask.nxv8i32.nxv8i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i32> %1,
+    <vscale x 8 x i16> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwadd.w.nxv16i32.nxv16i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwadd.w_wv_nxv16i32_nxv16i32_nxv16i16(<vscale x 16 x i32> %0, <vscale x 16 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wv_nxv16i32_nxv16i32_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwadd.w.nxv16i32.nxv16i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwadd.w.mask.nxv16i32.nxv16i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwadd.w_mask_wv_nxv16i32_nxv16i32_nxv16i16(<vscale x 16 x i32> %0, <vscale x 16 x i32> %1, <vscale x 16 x i16> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wv_nxv16i32_nxv16i32_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwadd.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwadd.w.mask.nxv16i32.nxv16i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i32> %1,
+    <vscale x 16 x i16> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwadd.w.nxv1i16.i8(
+  <vscale x 1 x i16>,
+  i8,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwadd.w_wx_nxv1i16_nxv1i16_i8(<vscale x 1 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wx_nxv1i16_nxv1i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwadd.w.nxv1i16.i8(
+    <vscale x 1 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwadd.w.mask.nxv1i16.i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  i8,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwadd.w_mask_wx_nxv1i16_nxv1i16_i8(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, i8 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wx_nxv1i16_nxv1i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwadd.w.mask.nxv1i16.i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    i8 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwadd.w.nxv2i16.i8(
+  <vscale x 2 x i16>,
+  i8,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwadd.w_wx_nxv2i16_nxv2i16_i8(<vscale x 2 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wx_nxv2i16_nxv2i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwadd.w.nxv2i16.i8(
+    <vscale x 2 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwadd.w.mask.nxv2i16.i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  i8,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwadd.w_mask_wx_nxv2i16_nxv2i16_i8(<vscale x 2 x i16> %0, <vscale x 2 x i16> %1, i8 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wx_nxv2i16_nxv2i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwadd.w.mask.nxv2i16.i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i16> %1,
+    i8 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwadd.w.nxv4i16.i8(
+  <vscale x 4 x i16>,
+  i8,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwadd.w_wx_nxv4i16_nxv4i16_i8(<vscale x 4 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wx_nxv4i16_nxv4i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwadd.w.nxv4i16.i8(
+    <vscale x 4 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwadd.w.mask.nxv4i16.i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  i8,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwadd.w_mask_wx_nxv4i16_nxv4i16_i8(<vscale x 4 x i16> %0, <vscale x 4 x i16> %1, i8 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wx_nxv4i16_nxv4i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwadd.w.mask.nxv4i16.i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i16> %1,
+    i8 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwadd.w.nxv8i16.i8(
+  <vscale x 8 x i16>,
+  i8,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwadd.w_wx_nxv8i16_nxv8i16_i8(<vscale x 8 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wx_nxv8i16_nxv8i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwadd.w.nxv8i16.i8(
+    <vscale x 8 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwadd.w.mask.nxv8i16.i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  i8,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwadd.w_mask_wx_nxv8i16_nxv8i16_i8(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, i8 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wx_nxv8i16_nxv8i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwadd.w.mask.nxv8i16.i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i16> %1,
+    i8 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwadd.w.nxv16i16.i8(
+  <vscale x 16 x i16>,
+  i8,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwadd.w_wx_nxv16i16_nxv16i16_i8(<vscale x 16 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wx_nxv16i16_nxv16i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwadd.w.nxv16i16.i8(
+    <vscale x 16 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwadd.w.mask.nxv16i16.i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  i8,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwadd.w_mask_wx_nxv16i16_nxv16i16_i8(<vscale x 16 x i16> %0, <vscale x 16 x i16> %1, i8 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wx_nxv16i16_nxv16i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwadd.w.mask.nxv16i16.i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i16> %1,
+    i8 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwadd.w.nxv32i16.i8(
+  <vscale x 32 x i16>,
+  i8,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwadd.w_wx_nxv32i16_nxv32i16_i8(<vscale x 32 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wx_nxv32i16_nxv32i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwadd.w.nxv32i16.i8(
+    <vscale x 32 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwadd.w.mask.nxv32i16.i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i16>,
+  i8,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwadd.w_mask_wx_nxv32i16_nxv32i16_i8(<vscale x 32 x i16> %0, <vscale x 32 x i16> %1, i8 %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wx_nxv32i16_nxv32i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwadd.w.mask.nxv32i16.i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i16> %1,
+    i8 %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwadd.w.nxv1i32.i16(
+  <vscale x 1 x i32>,
+  i16,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwadd.w_wx_nxv1i32_nxv1i32_i16(<vscale x 1 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wx_nxv1i32_nxv1i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwadd.w.nxv1i32.i16(
+    <vscale x 1 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwadd.w.mask.nxv1i32.i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i32>,
+  i16,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwadd.w_mask_wx_nxv1i32_nxv1i32_i16(<vscale x 1 x i32> %0, <vscale x 1 x i32> %1, i16 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wx_nxv1i32_nxv1i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwadd.w.mask.nxv1i32.i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i32> %1,
+    i16 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwadd.w.nxv2i32.i16(
+  <vscale x 2 x i32>,
+  i16,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwadd.w_wx_nxv2i32_nxv2i32_i16(<vscale x 2 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wx_nxv2i32_nxv2i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwadd.w.nxv2i32.i16(
+    <vscale x 2 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwadd.w.mask.nxv2i32.i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i32>,
+  i16,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwadd.w_mask_wx_nxv2i32_nxv2i32_i16(<vscale x 2 x i32> %0, <vscale x 2 x i32> %1, i16 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wx_nxv2i32_nxv2i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwadd.w.mask.nxv2i32.i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i32> %1,
+    i16 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwadd.w.nxv4i32.i16(
+  <vscale x 4 x i32>,
+  i16,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwadd.w_wx_nxv4i32_nxv4i32_i16(<vscale x 4 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wx_nxv4i32_nxv4i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwadd.w.nxv4i32.i16(
+    <vscale x 4 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwadd.w.mask.nxv4i32.i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i32>,
+  i16,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwadd.w_mask_wx_nxv4i32_nxv4i32_i16(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, i16 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wx_nxv4i32_nxv4i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwadd.w.mask.nxv4i32.i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i32> %1,
+    i16 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwadd.w.nxv8i32.i16(
+  <vscale x 8 x i32>,
+  i16,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwadd.w_wx_nxv8i32_nxv8i32_i16(<vscale x 8 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wx_nxv8i32_nxv8i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwadd.w.nxv8i32.i16(
+    <vscale x 8 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwadd.w.mask.nxv8i32.i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i32>,
+  i16,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwadd.w_mask_wx_nxv8i32_nxv8i32_i16(<vscale x 8 x i32> %0, <vscale x 8 x i32> %1, i16 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wx_nxv8i32_nxv8i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwadd.w.mask.nxv8i32.i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i32> %1,
+    i16 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwadd.w.nxv16i32.i16(
+  <vscale x 16 x i32>,
+  i16,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwadd.w_wx_nxv16i32_nxv16i32_i16(<vscale x 16 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_wx_nxv16i32_nxv16i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwadd.w.nxv16i32.i16(
+    <vscale x 16 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwadd.w.mask.nxv16i32.i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i32>,
+  i16,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwadd.w_mask_wx_nxv16i32_nxv16i32_i16(<vscale x 16 x i32> %0, <vscale x 16 x i32> %1, i16 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwadd.w_mask_wx_nxv16i32_nxv16i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwadd.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwadd.w.mask.nxv16i32.i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i32> %1,
+    i16 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}

llvm/test/CodeGen/RISCV/rvv/vwaddu-rv32.ll

@@ -0,0 +1,881 @@
+; RUN: llc -mtriple=riscv32 -mattr=+experimental-v -verify-machineinstrs \
+; RUN:   --riscv-no-aliases < %s | FileCheck %s
+declare <vscale x 1 x i16> @llvm.riscv.vwaddu.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwaddu_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vv_nxv1i16_nxv1i8_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwaddu.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwaddu.mask.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwaddu_mask_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vv_nxv1i16_nxv1i8_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwaddu.mask.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwaddu.nxv2i16.nxv2i8.nxv2i8(
+  <vscale x 2 x i8>,
+  <vscale x 2 x i8>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwaddu_vv_nxv2i16_nxv2i8_nxv2i8(<vscale x 2 x i8> %0, <vscale x 2 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vv_nxv2i16_nxv2i8_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwaddu.nxv2i16.nxv2i8.nxv2i8(
+    <vscale x 2 x i8> %0,
+    <vscale x 2 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwaddu.mask.nxv2i16.nxv2i8.nxv2i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  <vscale x 2 x i8>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwaddu_mask_vv_nxv2i16_nxv2i8_nxv2i8(<vscale x 2 x i16> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vv_nxv2i16_nxv2i8_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwaddu.mask.nxv2i16.nxv2i8.nxv2i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i8> %1,
+    <vscale x 2 x i8> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwaddu.nxv4i16.nxv4i8.nxv4i8(
+  <vscale x 4 x i8>,
+  <vscale x 4 x i8>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwaddu_vv_nxv4i16_nxv4i8_nxv4i8(<vscale x 4 x i8> %0, <vscale x 4 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vv_nxv4i16_nxv4i8_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwaddu.nxv4i16.nxv4i8.nxv4i8(
+    <vscale x 4 x i8> %0,
+    <vscale x 4 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwaddu.mask.nxv4i16.nxv4i8.nxv4i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  <vscale x 4 x i8>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwaddu_mask_vv_nxv4i16_nxv4i8_nxv4i8(<vscale x 4 x i16> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vv_nxv4i16_nxv4i8_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwaddu.mask.nxv4i16.nxv4i8.nxv4i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i8> %1,
+    <vscale x 4 x i8> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwaddu.nxv8i16.nxv8i8.nxv8i8(
+  <vscale x 8 x i8>,
+  <vscale x 8 x i8>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwaddu_vv_nxv8i16_nxv8i8_nxv8i8(<vscale x 8 x i8> %0, <vscale x 8 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vv_nxv8i16_nxv8i8_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwaddu.nxv8i16.nxv8i8.nxv8i8(
+    <vscale x 8 x i8> %0,
+    <vscale x 8 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwaddu.mask.nxv8i16.nxv8i8.nxv8i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  <vscale x 8 x i8>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwaddu_mask_vv_nxv8i16_nxv8i8_nxv8i8(<vscale x 8 x i16> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vv_nxv8i16_nxv8i8_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwaddu.mask.nxv8i16.nxv8i8.nxv8i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i8> %1,
+    <vscale x 8 x i8> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwaddu.nxv16i16.nxv16i8.nxv16i8(
+  <vscale x 16 x i8>,
+  <vscale x 16 x i8>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwaddu_vv_nxv16i16_nxv16i8_nxv16i8(<vscale x 16 x i8> %0, <vscale x 16 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vv_nxv16i16_nxv16i8_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwaddu.nxv16i16.nxv16i8.nxv16i8(
+    <vscale x 16 x i8> %0,
+    <vscale x 16 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwaddu.mask.nxv16i16.nxv16i8.nxv16i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  <vscale x 16 x i8>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwaddu_mask_vv_nxv16i16_nxv16i8_nxv16i8(<vscale x 16 x i16> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vv_nxv16i16_nxv16i8_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwaddu.mask.nxv16i16.nxv16i8.nxv16i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i8> %1,
+    <vscale x 16 x i8> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwaddu.nxv32i16.nxv32i8.nxv32i8(
+  <vscale x 32 x i8>,
+  <vscale x 32 x i8>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwaddu_vv_nxv32i16_nxv32i8_nxv32i8(<vscale x 32 x i8> %0, <vscale x 32 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vv_nxv32i16_nxv32i8_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwaddu.nxv32i16.nxv32i8.nxv32i8(
+    <vscale x 32 x i8> %0,
+    <vscale x 32 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwaddu.mask.nxv32i16.nxv32i8.nxv32i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  <vscale x 32 x i8>,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwaddu_mask_vv_nxv32i16_nxv32i8_nxv32i8(<vscale x 32 x i16> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vv_nxv32i16_nxv32i8_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwaddu.mask.nxv32i16.nxv32i8.nxv32i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i8> %1,
+    <vscale x 32 x i8> %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwaddu.nxv1i32.nxv1i16.nxv1i16(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwaddu_vv_nxv1i32_nxv1i16_nxv1i16(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vv_nxv1i32_nxv1i16_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwaddu.nxv1i32.nxv1i16.nxv1i16(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwaddu.mask.nxv1i32.nxv1i16.nxv1i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwaddu_mask_vv_nxv1i32_nxv1i16_nxv1i16(<vscale x 1 x i32> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vv_nxv1i32_nxv1i16_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwaddu.mask.nxv1i32.nxv1i16.nxv1i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i16> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwaddu.nxv2i32.nxv2i16.nxv2i16(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwaddu_vv_nxv2i32_nxv2i16_nxv2i16(<vscale x 2 x i16> %0, <vscale x 2 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vv_nxv2i32_nxv2i16_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwaddu.nxv2i32.nxv2i16.nxv2i16(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwaddu.mask.nxv2i32.nxv2i16.nxv2i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwaddu_mask_vv_nxv2i32_nxv2i16_nxv2i16(<vscale x 2 x i32> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vv_nxv2i32_nxv2i16_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwaddu.mask.nxv2i32.nxv2i16.nxv2i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i16> %1,
+    <vscale x 2 x i16> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwaddu.nxv4i32.nxv4i16.nxv4i16(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwaddu_vv_nxv4i32_nxv4i16_nxv4i16(<vscale x 4 x i16> %0, <vscale x 4 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vv_nxv4i32_nxv4i16_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwaddu.nxv4i32.nxv4i16.nxv4i16(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwaddu.mask.nxv4i32.nxv4i16.nxv4i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwaddu_mask_vv_nxv4i32_nxv4i16_nxv4i16(<vscale x 4 x i32> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vv_nxv4i32_nxv4i16_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwaddu.mask.nxv4i32.nxv4i16.nxv4i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i16> %1,
+    <vscale x 4 x i16> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwaddu.nxv8i32.nxv8i16.nxv8i16(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwaddu_vv_nxv8i32_nxv8i16_nxv8i16(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vv_nxv8i32_nxv8i16_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwaddu.nxv8i32.nxv8i16.nxv8i16(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwaddu.mask.nxv8i32.nxv8i16.nxv8i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwaddu_mask_vv_nxv8i32_nxv8i16_nxv8i16(<vscale x 8 x i32> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vv_nxv8i32_nxv8i16_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwaddu.mask.nxv8i32.nxv8i16.nxv8i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i16> %1,
+    <vscale x 8 x i16> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwaddu.nxv16i32.nxv16i16.nxv16i16(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwaddu_vv_nxv16i32_nxv16i16_nxv16i16(<vscale x 16 x i16> %0, <vscale x 16 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vv_nxv16i32_nxv16i16_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwaddu.nxv16i32.nxv16i16.nxv16i16(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwaddu.mask.nxv16i32.nxv16i16.nxv16i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwaddu_mask_vv_nxv16i32_nxv16i16_nxv16i16(<vscale x 16 x i32> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vv_nxv16i32_nxv16i16_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwaddu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwaddu.mask.nxv16i32.nxv16i16.nxv16i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i16> %1,
+    <vscale x 16 x i16> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwaddu.nxv1i16.nxv1i8.i8(
+  <vscale x 1 x i8>,
+  i8,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwaddu_vx_nxv1i16_nxv1i8_i8(<vscale x 1 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vx_nxv1i16_nxv1i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwaddu.nxv1i16.nxv1i8.i8(
+    <vscale x 1 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwaddu.mask.nxv1i16.nxv1i8.i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  i8,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwaddu_mask_vx_nxv1i16_nxv1i8_i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vx_nxv1i16_nxv1i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwaddu.mask.nxv1i16.nxv1i8.i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    i8 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwaddu.nxv2i16.nxv2i8.i8(
+  <vscale x 2 x i8>,
+  i8,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwaddu_vx_nxv2i16_nxv2i8_i8(<vscale x 2 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vx_nxv2i16_nxv2i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwaddu.nxv2i16.nxv2i8.i8(
+    <vscale x 2 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwaddu.mask.nxv2i16.nxv2i8.i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  i8,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwaddu_mask_vx_nxv2i16_nxv2i8_i8(<vscale x 2 x i16> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vx_nxv2i16_nxv2i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwaddu.mask.nxv2i16.nxv2i8.i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i8> %1,
+    i8 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwaddu.nxv4i16.nxv4i8.i8(
+  <vscale x 4 x i8>,
+  i8,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwaddu_vx_nxv4i16_nxv4i8_i8(<vscale x 4 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vx_nxv4i16_nxv4i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwaddu.nxv4i16.nxv4i8.i8(
+    <vscale x 4 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwaddu.mask.nxv4i16.nxv4i8.i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  i8,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwaddu_mask_vx_nxv4i16_nxv4i8_i8(<vscale x 4 x i16> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vx_nxv4i16_nxv4i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwaddu.mask.nxv4i16.nxv4i8.i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i8> %1,
+    i8 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwaddu.nxv8i16.nxv8i8.i8(
+  <vscale x 8 x i8>,
+  i8,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwaddu_vx_nxv8i16_nxv8i8_i8(<vscale x 8 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vx_nxv8i16_nxv8i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwaddu.nxv8i16.nxv8i8.i8(
+    <vscale x 8 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwaddu.mask.nxv8i16.nxv8i8.i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  i8,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwaddu_mask_vx_nxv8i16_nxv8i8_i8(<vscale x 8 x i16> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vx_nxv8i16_nxv8i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwaddu.mask.nxv8i16.nxv8i8.i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i8> %1,
+    i8 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwaddu.nxv16i16.nxv16i8.i8(
+  <vscale x 16 x i8>,
+  i8,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwaddu_vx_nxv16i16_nxv16i8_i8(<vscale x 16 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vx_nxv16i16_nxv16i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwaddu.nxv16i16.nxv16i8.i8(
+    <vscale x 16 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwaddu.mask.nxv16i16.nxv16i8.i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  i8,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwaddu_mask_vx_nxv16i16_nxv16i8_i8(<vscale x 16 x i16> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vx_nxv16i16_nxv16i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwaddu.mask.nxv16i16.nxv16i8.i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i8> %1,
+    i8 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwaddu.nxv32i16.nxv32i8.i8(
+  <vscale x 32 x i8>,
+  i8,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwaddu_vx_nxv32i16_nxv32i8_i8(<vscale x 32 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vx_nxv32i16_nxv32i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwaddu.nxv32i16.nxv32i8.i8(
+    <vscale x 32 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwaddu.mask.nxv32i16.nxv32i8.i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  i8,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwaddu_mask_vx_nxv32i16_nxv32i8_i8(<vscale x 32 x i16> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vx_nxv32i16_nxv32i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwaddu.mask.nxv32i16.nxv32i8.i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i8> %1,
+    i8 %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwaddu.nxv1i32.nxv1i16.i16(
+  <vscale x 1 x i16>,
+  i16,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwaddu_vx_nxv1i32_nxv1i16_i16(<vscale x 1 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vx_nxv1i32_nxv1i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwaddu.nxv1i32.nxv1i16.i16(
+    <vscale x 1 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwaddu.mask.nxv1i32.nxv1i16.i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  i16,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwaddu_mask_vx_nxv1i32_nxv1i16_i16(<vscale x 1 x i32> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vx_nxv1i32_nxv1i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwaddu.mask.nxv1i32.nxv1i16.i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i16> %1,
+    i16 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwaddu.nxv2i32.nxv2i16.i16(
+  <vscale x 2 x i16>,
+  i16,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwaddu_vx_nxv2i32_nxv2i16_i16(<vscale x 2 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vx_nxv2i32_nxv2i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwaddu.nxv2i32.nxv2i16.i16(
+    <vscale x 2 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwaddu.mask.nxv2i32.nxv2i16.i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  i16,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwaddu_mask_vx_nxv2i32_nxv2i16_i16(<vscale x 2 x i32> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vx_nxv2i32_nxv2i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwaddu.mask.nxv2i32.nxv2i16.i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i16> %1,
+    i16 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwaddu.nxv4i32.nxv4i16.i16(
+  <vscale x 4 x i16>,
+  i16,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwaddu_vx_nxv4i32_nxv4i16_i16(<vscale x 4 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vx_nxv4i32_nxv4i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwaddu.nxv4i32.nxv4i16.i16(
+    <vscale x 4 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwaddu.mask.nxv4i32.nxv4i16.i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  i16,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwaddu_mask_vx_nxv4i32_nxv4i16_i16(<vscale x 4 x i32> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vx_nxv4i32_nxv4i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwaddu.mask.nxv4i32.nxv4i16.i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i16> %1,
+    i16 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwaddu.nxv8i32.nxv8i16.i16(
+  <vscale x 8 x i16>,
+  i16,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwaddu_vx_nxv8i32_nxv8i16_i16(<vscale x 8 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vx_nxv8i32_nxv8i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwaddu.nxv8i32.nxv8i16.i16(
+    <vscale x 8 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwaddu.mask.nxv8i32.nxv8i16.i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  i16,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwaddu_mask_vx_nxv8i32_nxv8i16_i16(<vscale x 8 x i32> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vx_nxv8i32_nxv8i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwaddu.mask.nxv8i32.nxv8i16.i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i16> %1,
+    i16 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwaddu.nxv16i32.nxv16i16.i16(
+  <vscale x 16 x i16>,
+  i16,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwaddu_vx_nxv16i32_nxv16i16_i16(<vscale x 16 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_vx_nxv16i32_nxv16i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwaddu.nxv16i32.nxv16i16.i16(
+    <vscale x 16 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwaddu.mask.nxv16i32.nxv16i16.i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  i16,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwaddu_mask_vx_nxv16i32_nxv16i16_i16(<vscale x 16 x i32> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu_mask_vx_nxv16i32_nxv16i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwaddu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwaddu.mask.nxv16i32.nxv16i16.i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i16> %1,
+    i16 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}

llvm/test/CodeGen/RISCV/rvv/vwaddu.w-rv32.ll

@@ -0,0 +1,881 @@
+; RUN: llc -mtriple=riscv32 -mattr=+experimental-v -verify-machineinstrs \
+; RUN:   --riscv-no-aliases < %s | FileCheck %s
+declare <vscale x 1 x i16> @llvm.riscv.vwaddu.w.nxv1i16.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwaddu.w_wv_nxv1i16_nxv1i16_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wv_nxv1i16_nxv1i16_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwaddu.w.nxv1i16.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwaddu.w.mask.nxv1i16.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwaddu.w_mask_wv_nxv1i16_nxv1i16_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, <vscale x 1 x i8> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wv_nxv1i16_nxv1i16_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwaddu.w.mask.nxv1i16.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i8> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwaddu.w.nxv2i16.nxv2i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwaddu.w_wv_nxv2i16_nxv2i16_nxv2i8(<vscale x 2 x i16> %0, <vscale x 2 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wv_nxv2i16_nxv2i16_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwaddu.w.nxv2i16.nxv2i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwaddu.w.mask.nxv2i16.nxv2i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwaddu.w_mask_wv_nxv2i16_nxv2i16_nxv2i8(<vscale x 2 x i16> %0, <vscale x 2 x i16> %1, <vscale x 2 x i8> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wv_nxv2i16_nxv2i16_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwaddu.w.mask.nxv2i16.nxv2i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i16> %1,
+    <vscale x 2 x i8> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwaddu.w.nxv4i16.nxv4i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwaddu.w_wv_nxv4i16_nxv4i16_nxv4i8(<vscale x 4 x i16> %0, <vscale x 4 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wv_nxv4i16_nxv4i16_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwaddu.w.nxv4i16.nxv4i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwaddu.w.mask.nxv4i16.nxv4i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwaddu.w_mask_wv_nxv4i16_nxv4i16_nxv4i8(<vscale x 4 x i16> %0, <vscale x 4 x i16> %1, <vscale x 4 x i8> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wv_nxv4i16_nxv4i16_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwaddu.w.mask.nxv4i16.nxv4i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i16> %1,
+    <vscale x 4 x i8> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwaddu.w.nxv8i16.nxv8i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwaddu.w_wv_nxv8i16_nxv8i16_nxv8i8(<vscale x 8 x i16> %0, <vscale x 8 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wv_nxv8i16_nxv8i16_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwaddu.w.nxv8i16.nxv8i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwaddu.w.mask.nxv8i16.nxv8i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwaddu.w_mask_wv_nxv8i16_nxv8i16_nxv8i8(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, <vscale x 8 x i8> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wv_nxv8i16_nxv8i16_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwaddu.w.mask.nxv8i16.nxv8i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i16> %1,
+    <vscale x 8 x i8> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwaddu.w.nxv16i16.nxv16i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwaddu.w_wv_nxv16i16_nxv16i16_nxv16i8(<vscale x 16 x i16> %0, <vscale x 16 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wv_nxv16i16_nxv16i16_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwaddu.w.nxv16i16.nxv16i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwaddu.w.mask.nxv16i16.nxv16i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwaddu.w_mask_wv_nxv16i16_nxv16i16_nxv16i8(<vscale x 16 x i16> %0, <vscale x 16 x i16> %1, <vscale x 16 x i8> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wv_nxv16i16_nxv16i16_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwaddu.w.mask.nxv16i16.nxv16i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i16> %1,
+    <vscale x 16 x i8> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwaddu.w.nxv32i16.nxv32i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwaddu.w_wv_nxv32i16_nxv32i16_nxv32i8(<vscale x 32 x i16> %0, <vscale x 32 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wv_nxv32i16_nxv32i16_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwaddu.w.nxv32i16.nxv32i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwaddu.w.mask.nxv32i16.nxv32i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwaddu.w_mask_wv_nxv32i16_nxv32i16_nxv32i8(<vscale x 32 x i16> %0, <vscale x 32 x i16> %1, <vscale x 32 x i8> %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wv_nxv32i16_nxv32i16_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwaddu.w.mask.nxv32i16.nxv32i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i16> %1,
+    <vscale x 32 x i8> %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwaddu.w.nxv1i32.nxv1i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwaddu.w_wv_nxv1i32_nxv1i32_nxv1i16(<vscale x 1 x i32> %0, <vscale x 1 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wv_nxv1i32_nxv1i32_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwaddu.w.nxv1i32.nxv1i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwaddu.w.mask.nxv1i32.nxv1i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwaddu.w_mask_wv_nxv1i32_nxv1i32_nxv1i16(<vscale x 1 x i32> %0, <vscale x 1 x i32> %1, <vscale x 1 x i16> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wv_nxv1i32_nxv1i32_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwaddu.w.mask.nxv1i32.nxv1i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i32> %1,
+    <vscale x 1 x i16> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwaddu.w.nxv2i32.nxv2i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwaddu.w_wv_nxv2i32_nxv2i32_nxv2i16(<vscale x 2 x i32> %0, <vscale x 2 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wv_nxv2i32_nxv2i32_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwaddu.w.nxv2i32.nxv2i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwaddu.w.mask.nxv2i32.nxv2i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwaddu.w_mask_wv_nxv2i32_nxv2i32_nxv2i16(<vscale x 2 x i32> %0, <vscale x 2 x i32> %1, <vscale x 2 x i16> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wv_nxv2i32_nxv2i32_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwaddu.w.mask.nxv2i32.nxv2i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i32> %1,
+    <vscale x 2 x i16> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwaddu.w.nxv4i32.nxv4i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwaddu.w_wv_nxv4i32_nxv4i32_nxv4i16(<vscale x 4 x i32> %0, <vscale x 4 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wv_nxv4i32_nxv4i32_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwaddu.w.nxv4i32.nxv4i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwaddu.w.mask.nxv4i32.nxv4i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwaddu.w_mask_wv_nxv4i32_nxv4i32_nxv4i16(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, <vscale x 4 x i16> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wv_nxv4i32_nxv4i32_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwaddu.w.mask.nxv4i32.nxv4i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i32> %1,
+    <vscale x 4 x i16> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwaddu.w.nxv8i32.nxv8i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwaddu.w_wv_nxv8i32_nxv8i32_nxv8i16(<vscale x 8 x i32> %0, <vscale x 8 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wv_nxv8i32_nxv8i32_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwaddu.w.nxv8i32.nxv8i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwaddu.w.mask.nxv8i32.nxv8i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwaddu.w_mask_wv_nxv8i32_nxv8i32_nxv8i16(<vscale x 8 x i32> %0, <vscale x 8 x i32> %1, <vscale x 8 x i16> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wv_nxv8i32_nxv8i32_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwaddu.w.mask.nxv8i32.nxv8i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i32> %1,
+    <vscale x 8 x i16> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwaddu.w.nxv16i32.nxv16i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwaddu.w_wv_nxv16i32_nxv16i32_nxv16i16(<vscale x 16 x i32> %0, <vscale x 16 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wv_nxv16i32_nxv16i32_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwaddu.w.nxv16i32.nxv16i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwaddu.w.mask.nxv16i32.nxv16i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwaddu.w_mask_wv_nxv16i32_nxv16i32_nxv16i16(<vscale x 16 x i32> %0, <vscale x 16 x i32> %1, <vscale x 16 x i16> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wv_nxv16i32_nxv16i32_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwaddu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwaddu.w.mask.nxv16i32.nxv16i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i32> %1,
+    <vscale x 16 x i16> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwaddu.w.nxv1i16.i8(
+  <vscale x 1 x i16>,
+  i8,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwaddu.w_wx_nxv1i16_nxv1i16_i8(<vscale x 1 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wx_nxv1i16_nxv1i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwaddu.w.nxv1i16.i8(
+    <vscale x 1 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwaddu.w.mask.nxv1i16.i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  i8,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwaddu.w_mask_wx_nxv1i16_nxv1i16_i8(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, i8 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wx_nxv1i16_nxv1i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwaddu.w.mask.nxv1i16.i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    i8 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwaddu.w.nxv2i16.i8(
+  <vscale x 2 x i16>,
+  i8,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwaddu.w_wx_nxv2i16_nxv2i16_i8(<vscale x 2 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wx_nxv2i16_nxv2i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwaddu.w.nxv2i16.i8(
+    <vscale x 2 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwaddu.w.mask.nxv2i16.i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  i8,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwaddu.w_mask_wx_nxv2i16_nxv2i16_i8(<vscale x 2 x i16> %0, <vscale x 2 x i16> %1, i8 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wx_nxv2i16_nxv2i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwaddu.w.mask.nxv2i16.i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i16> %1,
+    i8 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwaddu.w.nxv4i16.i8(
+  <vscale x 4 x i16>,
+  i8,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwaddu.w_wx_nxv4i16_nxv4i16_i8(<vscale x 4 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wx_nxv4i16_nxv4i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwaddu.w.nxv4i16.i8(
+    <vscale x 4 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwaddu.w.mask.nxv4i16.i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  i8,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwaddu.w_mask_wx_nxv4i16_nxv4i16_i8(<vscale x 4 x i16> %0, <vscale x 4 x i16> %1, i8 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wx_nxv4i16_nxv4i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwaddu.w.mask.nxv4i16.i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i16> %1,
+    i8 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwaddu.w.nxv8i16.i8(
+  <vscale x 8 x i16>,
+  i8,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwaddu.w_wx_nxv8i16_nxv8i16_i8(<vscale x 8 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wx_nxv8i16_nxv8i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwaddu.w.nxv8i16.i8(
+    <vscale x 8 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwaddu.w.mask.nxv8i16.i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  i8,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwaddu.w_mask_wx_nxv8i16_nxv8i16_i8(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, i8 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wx_nxv8i16_nxv8i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwaddu.w.mask.nxv8i16.i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i16> %1,
+    i8 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwaddu.w.nxv16i16.i8(
+  <vscale x 16 x i16>,
+  i8,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwaddu.w_wx_nxv16i16_nxv16i16_i8(<vscale x 16 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wx_nxv16i16_nxv16i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwaddu.w.nxv16i16.i8(
+    <vscale x 16 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwaddu.w.mask.nxv16i16.i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  i8,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwaddu.w_mask_wx_nxv16i16_nxv16i16_i8(<vscale x 16 x i16> %0, <vscale x 16 x i16> %1, i8 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wx_nxv16i16_nxv16i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwaddu.w.mask.nxv16i16.i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i16> %1,
+    i8 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwaddu.w.nxv32i16.i8(
+  <vscale x 32 x i16>,
+  i8,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwaddu.w_wx_nxv32i16_nxv32i16_i8(<vscale x 32 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wx_nxv32i16_nxv32i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwaddu.w.nxv32i16.i8(
+    <vscale x 32 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwaddu.w.mask.nxv32i16.i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i16>,
+  i8,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwaddu.w_mask_wx_nxv32i16_nxv32i16_i8(<vscale x 32 x i16> %0, <vscale x 32 x i16> %1, i8 %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wx_nxv32i16_nxv32i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwaddu.w.mask.nxv32i16.i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i16> %1,
+    i8 %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwaddu.w.nxv1i32.i16(
+  <vscale x 1 x i32>,
+  i16,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwaddu.w_wx_nxv1i32_nxv1i32_i16(<vscale x 1 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wx_nxv1i32_nxv1i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwaddu.w.nxv1i32.i16(
+    <vscale x 1 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwaddu.w.mask.nxv1i32.i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i32>,
+  i16,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwaddu.w_mask_wx_nxv1i32_nxv1i32_i16(<vscale x 1 x i32> %0, <vscale x 1 x i32> %1, i16 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wx_nxv1i32_nxv1i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwaddu.w.mask.nxv1i32.i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i32> %1,
+    i16 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwaddu.w.nxv2i32.i16(
+  <vscale x 2 x i32>,
+  i16,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwaddu.w_wx_nxv2i32_nxv2i32_i16(<vscale x 2 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wx_nxv2i32_nxv2i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwaddu.w.nxv2i32.i16(
+    <vscale x 2 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwaddu.w.mask.nxv2i32.i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i32>,
+  i16,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwaddu.w_mask_wx_nxv2i32_nxv2i32_i16(<vscale x 2 x i32> %0, <vscale x 2 x i32> %1, i16 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wx_nxv2i32_nxv2i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwaddu.w.mask.nxv2i32.i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i32> %1,
+    i16 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwaddu.w.nxv4i32.i16(
+  <vscale x 4 x i32>,
+  i16,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwaddu.w_wx_nxv4i32_nxv4i32_i16(<vscale x 4 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wx_nxv4i32_nxv4i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwaddu.w.nxv4i32.i16(
+    <vscale x 4 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwaddu.w.mask.nxv4i32.i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i32>,
+  i16,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwaddu.w_mask_wx_nxv4i32_nxv4i32_i16(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, i16 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wx_nxv4i32_nxv4i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwaddu.w.mask.nxv4i32.i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i32> %1,
+    i16 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwaddu.w.nxv8i32.i16(
+  <vscale x 8 x i32>,
+  i16,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwaddu.w_wx_nxv8i32_nxv8i32_i16(<vscale x 8 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wx_nxv8i32_nxv8i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwaddu.w.nxv8i32.i16(
+    <vscale x 8 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwaddu.w.mask.nxv8i32.i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i32>,
+  i16,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwaddu.w_mask_wx_nxv8i32_nxv8i32_i16(<vscale x 8 x i32> %0, <vscale x 8 x i32> %1, i16 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wx_nxv8i32_nxv8i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwaddu.w.mask.nxv8i32.i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i32> %1,
+    i16 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwaddu.w.nxv16i32.i16(
+  <vscale x 16 x i32>,
+  i16,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwaddu.w_wx_nxv16i32_nxv16i32_i16(<vscale x 16 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_wx_nxv16i32_nxv16i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwaddu.w.nxv16i32.i16(
+    <vscale x 16 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwaddu.w.mask.nxv16i32.i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i32>,
+  i16,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwaddu.w_mask_wx_nxv16i32_nxv16i32_i16(<vscale x 16 x i32> %0, <vscale x 16 x i32> %1, i16 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwaddu.w_mask_wx_nxv16i32_nxv16i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwaddu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwaddu.w.mask.nxv16i32.i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i32> %1,
+    i16 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}

llvm/test/CodeGen/RISCV/rvv/vwsub-rv32.ll

@@ -0,0 +1,881 @@
+; RUN: llc -mtriple=riscv32 -mattr=+experimental-v -verify-machineinstrs \
+; RUN:   --riscv-no-aliases < %s | FileCheck %s
+declare <vscale x 1 x i16> @llvm.riscv.vwsub.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsub_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vv_nxv1i16_nxv1i8_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsub.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsub.mask.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsub_mask_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vv_nxv1i16_nxv1i8_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsub.mask.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsub.nxv2i16.nxv2i8.nxv2i8(
+  <vscale x 2 x i8>,
+  <vscale x 2 x i8>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsub_vv_nxv2i16_nxv2i8_nxv2i8(<vscale x 2 x i8> %0, <vscale x 2 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vv_nxv2i16_nxv2i8_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsub.nxv2i16.nxv2i8.nxv2i8(
+    <vscale x 2 x i8> %0,
+    <vscale x 2 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsub.mask.nxv2i16.nxv2i8.nxv2i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  <vscale x 2 x i8>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsub_mask_vv_nxv2i16_nxv2i8_nxv2i8(<vscale x 2 x i16> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vv_nxv2i16_nxv2i8_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsub.mask.nxv2i16.nxv2i8.nxv2i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i8> %1,
+    <vscale x 2 x i8> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsub.nxv4i16.nxv4i8.nxv4i8(
+  <vscale x 4 x i8>,
+  <vscale x 4 x i8>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsub_vv_nxv4i16_nxv4i8_nxv4i8(<vscale x 4 x i8> %0, <vscale x 4 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vv_nxv4i16_nxv4i8_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsub.nxv4i16.nxv4i8.nxv4i8(
+    <vscale x 4 x i8> %0,
+    <vscale x 4 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsub.mask.nxv4i16.nxv4i8.nxv4i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  <vscale x 4 x i8>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsub_mask_vv_nxv4i16_nxv4i8_nxv4i8(<vscale x 4 x i16> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vv_nxv4i16_nxv4i8_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsub.mask.nxv4i16.nxv4i8.nxv4i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i8> %1,
+    <vscale x 4 x i8> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsub.nxv8i16.nxv8i8.nxv8i8(
+  <vscale x 8 x i8>,
+  <vscale x 8 x i8>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsub_vv_nxv8i16_nxv8i8_nxv8i8(<vscale x 8 x i8> %0, <vscale x 8 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vv_nxv8i16_nxv8i8_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsub.nxv8i16.nxv8i8.nxv8i8(
+    <vscale x 8 x i8> %0,
+    <vscale x 8 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsub.mask.nxv8i16.nxv8i8.nxv8i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  <vscale x 8 x i8>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsub_mask_vv_nxv8i16_nxv8i8_nxv8i8(<vscale x 8 x i16> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vv_nxv8i16_nxv8i8_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsub.mask.nxv8i16.nxv8i8.nxv8i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i8> %1,
+    <vscale x 8 x i8> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsub.nxv16i16.nxv16i8.nxv16i8(
+  <vscale x 16 x i8>,
+  <vscale x 16 x i8>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsub_vv_nxv16i16_nxv16i8_nxv16i8(<vscale x 16 x i8> %0, <vscale x 16 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vv_nxv16i16_nxv16i8_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsub.nxv16i16.nxv16i8.nxv16i8(
+    <vscale x 16 x i8> %0,
+    <vscale x 16 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsub.mask.nxv16i16.nxv16i8.nxv16i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  <vscale x 16 x i8>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsub_mask_vv_nxv16i16_nxv16i8_nxv16i8(<vscale x 16 x i16> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vv_nxv16i16_nxv16i8_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsub.mask.nxv16i16.nxv16i8.nxv16i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i8> %1,
+    <vscale x 16 x i8> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsub.nxv32i16.nxv32i8.nxv32i8(
+  <vscale x 32 x i8>,
+  <vscale x 32 x i8>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsub_vv_nxv32i16_nxv32i8_nxv32i8(<vscale x 32 x i8> %0, <vscale x 32 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vv_nxv32i16_nxv32i8_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsub.nxv32i16.nxv32i8.nxv32i8(
+    <vscale x 32 x i8> %0,
+    <vscale x 32 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsub.mask.nxv32i16.nxv32i8.nxv32i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  <vscale x 32 x i8>,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsub_mask_vv_nxv32i16_nxv32i8_nxv32i8(<vscale x 32 x i16> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vv_nxv32i16_nxv32i8_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsub.mask.nxv32i16.nxv32i8.nxv32i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i8> %1,
+    <vscale x 32 x i8> %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsub.nxv1i32.nxv1i16.nxv1i16(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsub_vv_nxv1i32_nxv1i16_nxv1i16(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vv_nxv1i32_nxv1i16_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsub.nxv1i32.nxv1i16.nxv1i16(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsub.mask.nxv1i32.nxv1i16.nxv1i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsub_mask_vv_nxv1i32_nxv1i16_nxv1i16(<vscale x 1 x i32> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vv_nxv1i32_nxv1i16_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsub.mask.nxv1i32.nxv1i16.nxv1i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i16> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsub.nxv2i32.nxv2i16.nxv2i16(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsub_vv_nxv2i32_nxv2i16_nxv2i16(<vscale x 2 x i16> %0, <vscale x 2 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vv_nxv2i32_nxv2i16_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsub.nxv2i32.nxv2i16.nxv2i16(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsub.mask.nxv2i32.nxv2i16.nxv2i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsub_mask_vv_nxv2i32_nxv2i16_nxv2i16(<vscale x 2 x i32> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vv_nxv2i32_nxv2i16_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsub.mask.nxv2i32.nxv2i16.nxv2i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i16> %1,
+    <vscale x 2 x i16> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsub.nxv4i32.nxv4i16.nxv4i16(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsub_vv_nxv4i32_nxv4i16_nxv4i16(<vscale x 4 x i16> %0, <vscale x 4 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vv_nxv4i32_nxv4i16_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsub.nxv4i32.nxv4i16.nxv4i16(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsub.mask.nxv4i32.nxv4i16.nxv4i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsub_mask_vv_nxv4i32_nxv4i16_nxv4i16(<vscale x 4 x i32> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vv_nxv4i32_nxv4i16_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsub.mask.nxv4i32.nxv4i16.nxv4i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i16> %1,
+    <vscale x 4 x i16> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsub.nxv8i32.nxv8i16.nxv8i16(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsub_vv_nxv8i32_nxv8i16_nxv8i16(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vv_nxv8i32_nxv8i16_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsub.nxv8i32.nxv8i16.nxv8i16(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsub.mask.nxv8i32.nxv8i16.nxv8i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsub_mask_vv_nxv8i32_nxv8i16_nxv8i16(<vscale x 8 x i32> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vv_nxv8i32_nxv8i16_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsub.mask.nxv8i32.nxv8i16.nxv8i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i16> %1,
+    <vscale x 8 x i16> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsub.nxv16i32.nxv16i16.nxv16i16(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsub_vv_nxv16i32_nxv16i16_nxv16i16(<vscale x 16 x i16> %0, <vscale x 16 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vv_nxv16i32_nxv16i16_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsub.nxv16i32.nxv16i16.nxv16i16(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsub.mask.nxv16i32.nxv16i16.nxv16i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsub_mask_vv_nxv16i32_nxv16i16_nxv16i16(<vscale x 16 x i32> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vv_nxv16i32_nxv16i16_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsub.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsub.mask.nxv16i32.nxv16i16.nxv16i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i16> %1,
+    <vscale x 16 x i16> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsub.nxv1i16.nxv1i8.i8(
+  <vscale x 1 x i8>,
+  i8,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsub_vx_nxv1i16_nxv1i8_i8(<vscale x 1 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vx_nxv1i16_nxv1i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsub.nxv1i16.nxv1i8.i8(
+    <vscale x 1 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsub.mask.nxv1i16.nxv1i8.i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  i8,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsub_mask_vx_nxv1i16_nxv1i8_i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vx_nxv1i16_nxv1i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsub.mask.nxv1i16.nxv1i8.i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    i8 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsub.nxv2i16.nxv2i8.i8(
+  <vscale x 2 x i8>,
+  i8,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsub_vx_nxv2i16_nxv2i8_i8(<vscale x 2 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vx_nxv2i16_nxv2i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsub.nxv2i16.nxv2i8.i8(
+    <vscale x 2 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsub.mask.nxv2i16.nxv2i8.i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  i8,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsub_mask_vx_nxv2i16_nxv2i8_i8(<vscale x 2 x i16> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vx_nxv2i16_nxv2i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsub.mask.nxv2i16.nxv2i8.i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i8> %1,
+    i8 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsub.nxv4i16.nxv4i8.i8(
+  <vscale x 4 x i8>,
+  i8,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsub_vx_nxv4i16_nxv4i8_i8(<vscale x 4 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vx_nxv4i16_nxv4i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsub.nxv4i16.nxv4i8.i8(
+    <vscale x 4 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsub.mask.nxv4i16.nxv4i8.i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  i8,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsub_mask_vx_nxv4i16_nxv4i8_i8(<vscale x 4 x i16> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vx_nxv4i16_nxv4i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsub.mask.nxv4i16.nxv4i8.i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i8> %1,
+    i8 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsub.nxv8i16.nxv8i8.i8(
+  <vscale x 8 x i8>,
+  i8,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsub_vx_nxv8i16_nxv8i8_i8(<vscale x 8 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vx_nxv8i16_nxv8i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsub.nxv8i16.nxv8i8.i8(
+    <vscale x 8 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsub.mask.nxv8i16.nxv8i8.i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  i8,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsub_mask_vx_nxv8i16_nxv8i8_i8(<vscale x 8 x i16> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vx_nxv8i16_nxv8i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsub.mask.nxv8i16.nxv8i8.i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i8> %1,
+    i8 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsub.nxv16i16.nxv16i8.i8(
+  <vscale x 16 x i8>,
+  i8,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsub_vx_nxv16i16_nxv16i8_i8(<vscale x 16 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vx_nxv16i16_nxv16i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsub.nxv16i16.nxv16i8.i8(
+    <vscale x 16 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsub.mask.nxv16i16.nxv16i8.i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  i8,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsub_mask_vx_nxv16i16_nxv16i8_i8(<vscale x 16 x i16> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vx_nxv16i16_nxv16i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsub.mask.nxv16i16.nxv16i8.i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i8> %1,
+    i8 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsub.nxv32i16.nxv32i8.i8(
+  <vscale x 32 x i8>,
+  i8,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsub_vx_nxv32i16_nxv32i8_i8(<vscale x 32 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vx_nxv32i16_nxv32i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsub.nxv32i16.nxv32i8.i8(
+    <vscale x 32 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsub.mask.nxv32i16.nxv32i8.i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  i8,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsub_mask_vx_nxv32i16_nxv32i8_i8(<vscale x 32 x i16> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vx_nxv32i16_nxv32i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsub.mask.nxv32i16.nxv32i8.i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i8> %1,
+    i8 %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsub.nxv1i32.nxv1i16.i16(
+  <vscale x 1 x i16>,
+  i16,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsub_vx_nxv1i32_nxv1i16_i16(<vscale x 1 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vx_nxv1i32_nxv1i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsub.nxv1i32.nxv1i16.i16(
+    <vscale x 1 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsub.mask.nxv1i32.nxv1i16.i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  i16,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsub_mask_vx_nxv1i32_nxv1i16_i16(<vscale x 1 x i32> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vx_nxv1i32_nxv1i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsub.mask.nxv1i32.nxv1i16.i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i16> %1,
+    i16 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsub.nxv2i32.nxv2i16.i16(
+  <vscale x 2 x i16>,
+  i16,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsub_vx_nxv2i32_nxv2i16_i16(<vscale x 2 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vx_nxv2i32_nxv2i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsub.nxv2i32.nxv2i16.i16(
+    <vscale x 2 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsub.mask.nxv2i32.nxv2i16.i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  i16,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsub_mask_vx_nxv2i32_nxv2i16_i16(<vscale x 2 x i32> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vx_nxv2i32_nxv2i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsub.mask.nxv2i32.nxv2i16.i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i16> %1,
+    i16 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsub.nxv4i32.nxv4i16.i16(
+  <vscale x 4 x i16>,
+  i16,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsub_vx_nxv4i32_nxv4i16_i16(<vscale x 4 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vx_nxv4i32_nxv4i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsub.nxv4i32.nxv4i16.i16(
+    <vscale x 4 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsub.mask.nxv4i32.nxv4i16.i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  i16,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsub_mask_vx_nxv4i32_nxv4i16_i16(<vscale x 4 x i32> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vx_nxv4i32_nxv4i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsub.mask.nxv4i32.nxv4i16.i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i16> %1,
+    i16 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsub.nxv8i32.nxv8i16.i16(
+  <vscale x 8 x i16>,
+  i16,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsub_vx_nxv8i32_nxv8i16_i16(<vscale x 8 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vx_nxv8i32_nxv8i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsub.nxv8i32.nxv8i16.i16(
+    <vscale x 8 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsub.mask.nxv8i32.nxv8i16.i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  i16,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsub_mask_vx_nxv8i32_nxv8i16_i16(<vscale x 8 x i32> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vx_nxv8i32_nxv8i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsub.mask.nxv8i32.nxv8i16.i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i16> %1,
+    i16 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsub.nxv16i32.nxv16i16.i16(
+  <vscale x 16 x i16>,
+  i16,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsub_vx_nxv16i32_nxv16i16_i16(<vscale x 16 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_vx_nxv16i32_nxv16i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsub.nxv16i32.nxv16i16.i16(
+    <vscale x 16 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsub.mask.nxv16i32.nxv16i16.i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  i16,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsub_mask_vx_nxv16i32_nxv16i16_i16(<vscale x 16 x i32> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub_mask_vx_nxv16i32_nxv16i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsub.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsub.mask.nxv16i32.nxv16i16.i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i16> %1,
+    i16 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}

llvm/test/CodeGen/RISCV/rvv/vwsub.w-rv32.ll

@@ -0,0 +1,881 @@
+; RUN: llc -mtriple=riscv32 -mattr=+experimental-v -verify-machineinstrs \
+; RUN:   --riscv-no-aliases < %s | FileCheck %s
+declare <vscale x 1 x i16> @llvm.riscv.vwsub.w.nxv1i16.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsub.w_wv_nxv1i16_nxv1i16_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wv_nxv1i16_nxv1i16_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsub.w.nxv1i16.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsub.w.mask.nxv1i16.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsub.w_mask_wv_nxv1i16_nxv1i16_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, <vscale x 1 x i8> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wv_nxv1i16_nxv1i16_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsub.w.mask.nxv1i16.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i8> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsub.w.nxv2i16.nxv2i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsub.w_wv_nxv2i16_nxv2i16_nxv2i8(<vscale x 2 x i16> %0, <vscale x 2 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wv_nxv2i16_nxv2i16_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsub.w.nxv2i16.nxv2i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsub.w.mask.nxv2i16.nxv2i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsub.w_mask_wv_nxv2i16_nxv2i16_nxv2i8(<vscale x 2 x i16> %0, <vscale x 2 x i16> %1, <vscale x 2 x i8> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wv_nxv2i16_nxv2i16_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsub.w.mask.nxv2i16.nxv2i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i16> %1,
+    <vscale x 2 x i8> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsub.w.nxv4i16.nxv4i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsub.w_wv_nxv4i16_nxv4i16_nxv4i8(<vscale x 4 x i16> %0, <vscale x 4 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wv_nxv4i16_nxv4i16_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsub.w.nxv4i16.nxv4i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsub.w.mask.nxv4i16.nxv4i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsub.w_mask_wv_nxv4i16_nxv4i16_nxv4i8(<vscale x 4 x i16> %0, <vscale x 4 x i16> %1, <vscale x 4 x i8> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wv_nxv4i16_nxv4i16_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsub.w.mask.nxv4i16.nxv4i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i16> %1,
+    <vscale x 4 x i8> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsub.w.nxv8i16.nxv8i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsub.w_wv_nxv8i16_nxv8i16_nxv8i8(<vscale x 8 x i16> %0, <vscale x 8 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wv_nxv8i16_nxv8i16_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsub.w.nxv8i16.nxv8i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsub.w.mask.nxv8i16.nxv8i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsub.w_mask_wv_nxv8i16_nxv8i16_nxv8i8(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, <vscale x 8 x i8> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wv_nxv8i16_nxv8i16_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsub.w.mask.nxv8i16.nxv8i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i16> %1,
+    <vscale x 8 x i8> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsub.w.nxv16i16.nxv16i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsub.w_wv_nxv16i16_nxv16i16_nxv16i8(<vscale x 16 x i16> %0, <vscale x 16 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wv_nxv16i16_nxv16i16_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsub.w.nxv16i16.nxv16i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsub.w.mask.nxv16i16.nxv16i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsub.w_mask_wv_nxv16i16_nxv16i16_nxv16i8(<vscale x 16 x i16> %0, <vscale x 16 x i16> %1, <vscale x 16 x i8> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wv_nxv16i16_nxv16i16_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsub.w.mask.nxv16i16.nxv16i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i16> %1,
+    <vscale x 16 x i8> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsub.w.nxv32i16.nxv32i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsub.w_wv_nxv32i16_nxv32i16_nxv32i8(<vscale x 32 x i16> %0, <vscale x 32 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wv_nxv32i16_nxv32i16_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsub.w.nxv32i16.nxv32i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsub.w.mask.nxv32i16.nxv32i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsub.w_mask_wv_nxv32i16_nxv32i16_nxv32i8(<vscale x 32 x i16> %0, <vscale x 32 x i16> %1, <vscale x 32 x i8> %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wv_nxv32i16_nxv32i16_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsub.w.mask.nxv32i16.nxv32i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i16> %1,
+    <vscale x 32 x i8> %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsub.w.nxv1i32.nxv1i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsub.w_wv_nxv1i32_nxv1i32_nxv1i16(<vscale x 1 x i32> %0, <vscale x 1 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wv_nxv1i32_nxv1i32_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsub.w.nxv1i32.nxv1i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsub.w.mask.nxv1i32.nxv1i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsub.w_mask_wv_nxv1i32_nxv1i32_nxv1i16(<vscale x 1 x i32> %0, <vscale x 1 x i32> %1, <vscale x 1 x i16> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wv_nxv1i32_nxv1i32_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsub.w.mask.nxv1i32.nxv1i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i32> %1,
+    <vscale x 1 x i16> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsub.w.nxv2i32.nxv2i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsub.w_wv_nxv2i32_nxv2i32_nxv2i16(<vscale x 2 x i32> %0, <vscale x 2 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wv_nxv2i32_nxv2i32_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsub.w.nxv2i32.nxv2i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsub.w.mask.nxv2i32.nxv2i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsub.w_mask_wv_nxv2i32_nxv2i32_nxv2i16(<vscale x 2 x i32> %0, <vscale x 2 x i32> %1, <vscale x 2 x i16> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wv_nxv2i32_nxv2i32_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsub.w.mask.nxv2i32.nxv2i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i32> %1,
+    <vscale x 2 x i16> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsub.w.nxv4i32.nxv4i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsub.w_wv_nxv4i32_nxv4i32_nxv4i16(<vscale x 4 x i32> %0, <vscale x 4 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wv_nxv4i32_nxv4i32_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsub.w.nxv4i32.nxv4i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsub.w.mask.nxv4i32.nxv4i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsub.w_mask_wv_nxv4i32_nxv4i32_nxv4i16(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, <vscale x 4 x i16> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wv_nxv4i32_nxv4i32_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsub.w.mask.nxv4i32.nxv4i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i32> %1,
+    <vscale x 4 x i16> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsub.w.nxv8i32.nxv8i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsub.w_wv_nxv8i32_nxv8i32_nxv8i16(<vscale x 8 x i32> %0, <vscale x 8 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wv_nxv8i32_nxv8i32_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsub.w.nxv8i32.nxv8i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsub.w.mask.nxv8i32.nxv8i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsub.w_mask_wv_nxv8i32_nxv8i32_nxv8i16(<vscale x 8 x i32> %0, <vscale x 8 x i32> %1, <vscale x 8 x i16> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wv_nxv8i32_nxv8i32_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsub.w.mask.nxv8i32.nxv8i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i32> %1,
+    <vscale x 8 x i16> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsub.w.nxv16i32.nxv16i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsub.w_wv_nxv16i32_nxv16i32_nxv16i16(<vscale x 16 x i32> %0, <vscale x 16 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wv_nxv16i32_nxv16i32_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsub.w.nxv16i32.nxv16i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsub.w.mask.nxv16i32.nxv16i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsub.w_mask_wv_nxv16i32_nxv16i32_nxv16i16(<vscale x 16 x i32> %0, <vscale x 16 x i32> %1, <vscale x 16 x i16> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wv_nxv16i32_nxv16i32_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsub.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsub.w.mask.nxv16i32.nxv16i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i32> %1,
+    <vscale x 16 x i16> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsub.w.nxv1i16.i8(
+  <vscale x 1 x i16>,
+  i8,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsub.w_wx_nxv1i16_nxv1i16_i8(<vscale x 1 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wx_nxv1i16_nxv1i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsub.w.nxv1i16.i8(
+    <vscale x 1 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsub.w.mask.nxv1i16.i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  i8,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsub.w_mask_wx_nxv1i16_nxv1i16_i8(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, i8 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wx_nxv1i16_nxv1i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsub.w.mask.nxv1i16.i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    i8 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsub.w.nxv2i16.i8(
+  <vscale x 2 x i16>,
+  i8,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsub.w_wx_nxv2i16_nxv2i16_i8(<vscale x 2 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wx_nxv2i16_nxv2i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsub.w.nxv2i16.i8(
+    <vscale x 2 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsub.w.mask.nxv2i16.i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  i8,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsub.w_mask_wx_nxv2i16_nxv2i16_i8(<vscale x 2 x i16> %0, <vscale x 2 x i16> %1, i8 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wx_nxv2i16_nxv2i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsub.w.mask.nxv2i16.i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i16> %1,
+    i8 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsub.w.nxv4i16.i8(
+  <vscale x 4 x i16>,
+  i8,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsub.w_wx_nxv4i16_nxv4i16_i8(<vscale x 4 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wx_nxv4i16_nxv4i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsub.w.nxv4i16.i8(
+    <vscale x 4 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsub.w.mask.nxv4i16.i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  i8,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsub.w_mask_wx_nxv4i16_nxv4i16_i8(<vscale x 4 x i16> %0, <vscale x 4 x i16> %1, i8 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wx_nxv4i16_nxv4i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsub.w.mask.nxv4i16.i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i16> %1,
+    i8 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsub.w.nxv8i16.i8(
+  <vscale x 8 x i16>,
+  i8,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsub.w_wx_nxv8i16_nxv8i16_i8(<vscale x 8 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wx_nxv8i16_nxv8i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsub.w.nxv8i16.i8(
+    <vscale x 8 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsub.w.mask.nxv8i16.i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  i8,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsub.w_mask_wx_nxv8i16_nxv8i16_i8(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, i8 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wx_nxv8i16_nxv8i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsub.w.mask.nxv8i16.i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i16> %1,
+    i8 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsub.w.nxv16i16.i8(
+  <vscale x 16 x i16>,
+  i8,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsub.w_wx_nxv16i16_nxv16i16_i8(<vscale x 16 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wx_nxv16i16_nxv16i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsub.w.nxv16i16.i8(
+    <vscale x 16 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsub.w.mask.nxv16i16.i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  i8,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsub.w_mask_wx_nxv16i16_nxv16i16_i8(<vscale x 16 x i16> %0, <vscale x 16 x i16> %1, i8 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wx_nxv16i16_nxv16i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsub.w.mask.nxv16i16.i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i16> %1,
+    i8 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsub.w.nxv32i16.i8(
+  <vscale x 32 x i16>,
+  i8,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsub.w_wx_nxv32i16_nxv32i16_i8(<vscale x 32 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wx_nxv32i16_nxv32i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsub.w.nxv32i16.i8(
+    <vscale x 32 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsub.w.mask.nxv32i16.i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i16>,
+  i8,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsub.w_mask_wx_nxv32i16_nxv32i16_i8(<vscale x 32 x i16> %0, <vscale x 32 x i16> %1, i8 %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wx_nxv32i16_nxv32i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsub.w.mask.nxv32i16.i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i16> %1,
+    i8 %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsub.w.nxv1i32.i16(
+  <vscale x 1 x i32>,
+  i16,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsub.w_wx_nxv1i32_nxv1i32_i16(<vscale x 1 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wx_nxv1i32_nxv1i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsub.w.nxv1i32.i16(
+    <vscale x 1 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsub.w.mask.nxv1i32.i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i32>,
+  i16,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsub.w_mask_wx_nxv1i32_nxv1i32_i16(<vscale x 1 x i32> %0, <vscale x 1 x i32> %1, i16 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wx_nxv1i32_nxv1i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsub.w.mask.nxv1i32.i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i32> %1,
+    i16 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsub.w.nxv2i32.i16(
+  <vscale x 2 x i32>,
+  i16,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsub.w_wx_nxv2i32_nxv2i32_i16(<vscale x 2 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wx_nxv2i32_nxv2i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsub.w.nxv2i32.i16(
+    <vscale x 2 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsub.w.mask.nxv2i32.i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i32>,
+  i16,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsub.w_mask_wx_nxv2i32_nxv2i32_i16(<vscale x 2 x i32> %0, <vscale x 2 x i32> %1, i16 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wx_nxv2i32_nxv2i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsub.w.mask.nxv2i32.i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i32> %1,
+    i16 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsub.w.nxv4i32.i16(
+  <vscale x 4 x i32>,
+  i16,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsub.w_wx_nxv4i32_nxv4i32_i16(<vscale x 4 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wx_nxv4i32_nxv4i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsub.w.nxv4i32.i16(
+    <vscale x 4 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsub.w.mask.nxv4i32.i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i32>,
+  i16,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsub.w_mask_wx_nxv4i32_nxv4i32_i16(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, i16 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wx_nxv4i32_nxv4i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsub.w.mask.nxv4i32.i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i32> %1,
+    i16 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsub.w.nxv8i32.i16(
+  <vscale x 8 x i32>,
+  i16,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsub.w_wx_nxv8i32_nxv8i32_i16(<vscale x 8 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wx_nxv8i32_nxv8i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsub.w.nxv8i32.i16(
+    <vscale x 8 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsub.w.mask.nxv8i32.i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i32>,
+  i16,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsub.w_mask_wx_nxv8i32_nxv8i32_i16(<vscale x 8 x i32> %0, <vscale x 8 x i32> %1, i16 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wx_nxv8i32_nxv8i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsub.w.mask.nxv8i32.i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i32> %1,
+    i16 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsub.w.nxv16i32.i16(
+  <vscale x 16 x i32>,
+  i16,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsub.w_wx_nxv16i32_nxv16i32_i16(<vscale x 16 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_wx_nxv16i32_nxv16i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsub.w.nxv16i32.i16(
+    <vscale x 16 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsub.w.mask.nxv16i32.i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i32>,
+  i16,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsub.w_mask_wx_nxv16i32_nxv16i32_i16(<vscale x 16 x i32> %0, <vscale x 16 x i32> %1, i16 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsub.w_mask_wx_nxv16i32_nxv16i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsub.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsub.w.mask.nxv16i32.i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i32> %1,
+    i16 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}

llvm/test/CodeGen/RISCV/rvv/vwsubu-rv32.ll

@@ -0,0 +1,881 @@
+; RUN: llc -mtriple=riscv32 -mattr=+experimental-v -verify-machineinstrs \
+; RUN:   --riscv-no-aliases < %s | FileCheck %s
+declare <vscale x 1 x i16> @llvm.riscv.vwsubu.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsubu_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i8> %0, <vscale x 1 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vv_nxv1i16_nxv1i8_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsubu.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i8> %0,
+    <vscale x 1 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsubu.mask.nxv1i16.nxv1i8.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsubu_mask_vv_nxv1i16_nxv1i8_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, <vscale x 1 x i8> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vv_nxv1i16_nxv1i8_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsubu.mask.nxv1i16.nxv1i8.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    <vscale x 1 x i8> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsubu.nxv2i16.nxv2i8.nxv2i8(
+  <vscale x 2 x i8>,
+  <vscale x 2 x i8>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsubu_vv_nxv2i16_nxv2i8_nxv2i8(<vscale x 2 x i8> %0, <vscale x 2 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vv_nxv2i16_nxv2i8_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsubu.nxv2i16.nxv2i8.nxv2i8(
+    <vscale x 2 x i8> %0,
+    <vscale x 2 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsubu.mask.nxv2i16.nxv2i8.nxv2i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  <vscale x 2 x i8>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsubu_mask_vv_nxv2i16_nxv2i8_nxv2i8(<vscale x 2 x i16> %0, <vscale x 2 x i8> %1, <vscale x 2 x i8> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vv_nxv2i16_nxv2i8_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsubu.mask.nxv2i16.nxv2i8.nxv2i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i8> %1,
+    <vscale x 2 x i8> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsubu.nxv4i16.nxv4i8.nxv4i8(
+  <vscale x 4 x i8>,
+  <vscale x 4 x i8>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsubu_vv_nxv4i16_nxv4i8_nxv4i8(<vscale x 4 x i8> %0, <vscale x 4 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vv_nxv4i16_nxv4i8_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsubu.nxv4i16.nxv4i8.nxv4i8(
+    <vscale x 4 x i8> %0,
+    <vscale x 4 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsubu.mask.nxv4i16.nxv4i8.nxv4i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  <vscale x 4 x i8>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsubu_mask_vv_nxv4i16_nxv4i8_nxv4i8(<vscale x 4 x i16> %0, <vscale x 4 x i8> %1, <vscale x 4 x i8> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vv_nxv4i16_nxv4i8_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsubu.mask.nxv4i16.nxv4i8.nxv4i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i8> %1,
+    <vscale x 4 x i8> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsubu.nxv8i16.nxv8i8.nxv8i8(
+  <vscale x 8 x i8>,
+  <vscale x 8 x i8>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsubu_vv_nxv8i16_nxv8i8_nxv8i8(<vscale x 8 x i8> %0, <vscale x 8 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vv_nxv8i16_nxv8i8_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsubu.nxv8i16.nxv8i8.nxv8i8(
+    <vscale x 8 x i8> %0,
+    <vscale x 8 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsubu.mask.nxv8i16.nxv8i8.nxv8i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  <vscale x 8 x i8>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsubu_mask_vv_nxv8i16_nxv8i8_nxv8i8(<vscale x 8 x i16> %0, <vscale x 8 x i8> %1, <vscale x 8 x i8> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vv_nxv8i16_nxv8i8_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsubu.mask.nxv8i16.nxv8i8.nxv8i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i8> %1,
+    <vscale x 8 x i8> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsubu.nxv16i16.nxv16i8.nxv16i8(
+  <vscale x 16 x i8>,
+  <vscale x 16 x i8>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsubu_vv_nxv16i16_nxv16i8_nxv16i8(<vscale x 16 x i8> %0, <vscale x 16 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vv_nxv16i16_nxv16i8_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsubu.nxv16i16.nxv16i8.nxv16i8(
+    <vscale x 16 x i8> %0,
+    <vscale x 16 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsubu.mask.nxv16i16.nxv16i8.nxv16i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  <vscale x 16 x i8>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsubu_mask_vv_nxv16i16_nxv16i8_nxv16i8(<vscale x 16 x i16> %0, <vscale x 16 x i8> %1, <vscale x 16 x i8> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vv_nxv16i16_nxv16i8_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsubu.mask.nxv16i16.nxv16i8.nxv16i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i8> %1,
+    <vscale x 16 x i8> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsubu.nxv32i16.nxv32i8.nxv32i8(
+  <vscale x 32 x i8>,
+  <vscale x 32 x i8>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsubu_vv_nxv32i16_nxv32i8_nxv32i8(<vscale x 32 x i8> %0, <vscale x 32 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vv_nxv32i16_nxv32i8_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsubu.nxv32i16.nxv32i8.nxv32i8(
+    <vscale x 32 x i8> %0,
+    <vscale x 32 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsubu.mask.nxv32i16.nxv32i8.nxv32i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  <vscale x 32 x i8>,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsubu_mask_vv_nxv32i16_nxv32i8_nxv32i8(<vscale x 32 x i16> %0, <vscale x 32 x i8> %1, <vscale x 32 x i8> %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vv_nxv32i16_nxv32i8_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsubu.mask.nxv32i16.nxv32i8.nxv32i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i8> %1,
+    <vscale x 32 x i8> %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsubu.nxv1i32.nxv1i16.nxv1i16(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsubu_vv_nxv1i32_nxv1i16_nxv1i16(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vv_nxv1i32_nxv1i16_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsubu.nxv1i32.nxv1i16.nxv1i16(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsubu.mask.nxv1i32.nxv1i16.nxv1i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsubu_mask_vv_nxv1i32_nxv1i16_nxv1i16(<vscale x 1 x i32> %0, <vscale x 1 x i16> %1, <vscale x 1 x i16> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vv_nxv1i32_nxv1i16_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsubu.mask.nxv1i32.nxv1i16.nxv1i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i16> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsubu.nxv2i32.nxv2i16.nxv2i16(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsubu_vv_nxv2i32_nxv2i16_nxv2i16(<vscale x 2 x i16> %0, <vscale x 2 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vv_nxv2i32_nxv2i16_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsubu.nxv2i32.nxv2i16.nxv2i16(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsubu.mask.nxv2i32.nxv2i16.nxv2i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsubu_mask_vv_nxv2i32_nxv2i16_nxv2i16(<vscale x 2 x i32> %0, <vscale x 2 x i16> %1, <vscale x 2 x i16> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vv_nxv2i32_nxv2i16_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsubu.mask.nxv2i32.nxv2i16.nxv2i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i16> %1,
+    <vscale x 2 x i16> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsubu.nxv4i32.nxv4i16.nxv4i16(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsubu_vv_nxv4i32_nxv4i16_nxv4i16(<vscale x 4 x i16> %0, <vscale x 4 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vv_nxv4i32_nxv4i16_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsubu.nxv4i32.nxv4i16.nxv4i16(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsubu.mask.nxv4i32.nxv4i16.nxv4i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsubu_mask_vv_nxv4i32_nxv4i16_nxv4i16(<vscale x 4 x i32> %0, <vscale x 4 x i16> %1, <vscale x 4 x i16> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vv_nxv4i32_nxv4i16_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsubu.mask.nxv4i32.nxv4i16.nxv4i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i16> %1,
+    <vscale x 4 x i16> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsubu.nxv8i32.nxv8i16.nxv8i16(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsubu_vv_nxv8i32_nxv8i16_nxv8i16(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vv_nxv8i32_nxv8i16_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsubu.nxv8i32.nxv8i16.nxv8i16(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsubu.mask.nxv8i32.nxv8i16.nxv8i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsubu_mask_vv_nxv8i32_nxv8i16_nxv8i16(<vscale x 8 x i32> %0, <vscale x 8 x i16> %1, <vscale x 8 x i16> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vv_nxv8i32_nxv8i16_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsubu.mask.nxv8i32.nxv8i16.nxv8i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i16> %1,
+    <vscale x 8 x i16> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsubu.nxv16i32.nxv16i16.nxv16i16(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsubu_vv_nxv16i32_nxv16i16_nxv16i16(<vscale x 16 x i16> %0, <vscale x 16 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vv_nxv16i32_nxv16i16_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsubu.nxv16i32.nxv16i16.nxv16i16(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsubu.mask.nxv16i32.nxv16i16.nxv16i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsubu_mask_vv_nxv16i32_nxv16i16_nxv16i16(<vscale x 16 x i32> %0, <vscale x 16 x i16> %1, <vscale x 16 x i16> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vv_nxv16i32_nxv16i16_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsubu.vv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsubu.mask.nxv16i32.nxv16i16.nxv16i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i16> %1,
+    <vscale x 16 x i16> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsubu.nxv1i16.nxv1i8.i8(
+  <vscale x 1 x i8>,
+  i8,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsubu_vx_nxv1i16_nxv1i8_i8(<vscale x 1 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vx_nxv1i16_nxv1i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsubu.nxv1i16.nxv1i8.i8(
+    <vscale x 1 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsubu.mask.nxv1i16.nxv1i8.i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  i8,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsubu_mask_vx_nxv1i16_nxv1i8_i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, i8 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vx_nxv1i16_nxv1i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsubu.mask.nxv1i16.nxv1i8.i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    i8 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsubu.nxv2i16.nxv2i8.i8(
+  <vscale x 2 x i8>,
+  i8,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsubu_vx_nxv2i16_nxv2i8_i8(<vscale x 2 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vx_nxv2i16_nxv2i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsubu.nxv2i16.nxv2i8.i8(
+    <vscale x 2 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsubu.mask.nxv2i16.nxv2i8.i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  i8,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsubu_mask_vx_nxv2i16_nxv2i8_i8(<vscale x 2 x i16> %0, <vscale x 2 x i8> %1, i8 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vx_nxv2i16_nxv2i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsubu.mask.nxv2i16.nxv2i8.i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i8> %1,
+    i8 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsubu.nxv4i16.nxv4i8.i8(
+  <vscale x 4 x i8>,
+  i8,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsubu_vx_nxv4i16_nxv4i8_i8(<vscale x 4 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vx_nxv4i16_nxv4i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsubu.nxv4i16.nxv4i8.i8(
+    <vscale x 4 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsubu.mask.nxv4i16.nxv4i8.i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  i8,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsubu_mask_vx_nxv4i16_nxv4i8_i8(<vscale x 4 x i16> %0, <vscale x 4 x i8> %1, i8 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vx_nxv4i16_nxv4i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsubu.mask.nxv4i16.nxv4i8.i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i8> %1,
+    i8 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsubu.nxv8i16.nxv8i8.i8(
+  <vscale x 8 x i8>,
+  i8,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsubu_vx_nxv8i16_nxv8i8_i8(<vscale x 8 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vx_nxv8i16_nxv8i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsubu.nxv8i16.nxv8i8.i8(
+    <vscale x 8 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsubu.mask.nxv8i16.nxv8i8.i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  i8,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsubu_mask_vx_nxv8i16_nxv8i8_i8(<vscale x 8 x i16> %0, <vscale x 8 x i8> %1, i8 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vx_nxv8i16_nxv8i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsubu.mask.nxv8i16.nxv8i8.i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i8> %1,
+    i8 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsubu.nxv16i16.nxv16i8.i8(
+  <vscale x 16 x i8>,
+  i8,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsubu_vx_nxv16i16_nxv16i8_i8(<vscale x 16 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vx_nxv16i16_nxv16i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsubu.nxv16i16.nxv16i8.i8(
+    <vscale x 16 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsubu.mask.nxv16i16.nxv16i8.i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  i8,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsubu_mask_vx_nxv16i16_nxv16i8_i8(<vscale x 16 x i16> %0, <vscale x 16 x i8> %1, i8 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vx_nxv16i16_nxv16i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsubu.mask.nxv16i16.nxv16i8.i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i8> %1,
+    i8 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsubu.nxv32i16.nxv32i8.i8(
+  <vscale x 32 x i8>,
+  i8,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsubu_vx_nxv32i16_nxv32i8_i8(<vscale x 32 x i8> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vx_nxv32i16_nxv32i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsubu.nxv32i16.nxv32i8.i8(
+    <vscale x 32 x i8> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsubu.mask.nxv32i16.nxv32i8.i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  i8,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsubu_mask_vx_nxv32i16_nxv32i8_i8(<vscale x 32 x i16> %0, <vscale x 32 x i8> %1, i8 %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vx_nxv32i16_nxv32i8_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsubu.mask.nxv32i16.nxv32i8.i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i8> %1,
+    i8 %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsubu.nxv1i32.nxv1i16.i16(
+  <vscale x 1 x i16>,
+  i16,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsubu_vx_nxv1i32_nxv1i16_i16(<vscale x 1 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vx_nxv1i32_nxv1i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsubu.nxv1i32.nxv1i16.i16(
+    <vscale x 1 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsubu.mask.nxv1i32.nxv1i16.i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  i16,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsubu_mask_vx_nxv1i32_nxv1i16_i16(<vscale x 1 x i32> %0, <vscale x 1 x i16> %1, i16 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vx_nxv1i32_nxv1i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsubu.mask.nxv1i32.nxv1i16.i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i16> %1,
+    i16 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsubu.nxv2i32.nxv2i16.i16(
+  <vscale x 2 x i16>,
+  i16,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsubu_vx_nxv2i32_nxv2i16_i16(<vscale x 2 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vx_nxv2i32_nxv2i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsubu.nxv2i32.nxv2i16.i16(
+    <vscale x 2 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsubu.mask.nxv2i32.nxv2i16.i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  i16,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsubu_mask_vx_nxv2i32_nxv2i16_i16(<vscale x 2 x i32> %0, <vscale x 2 x i16> %1, i16 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vx_nxv2i32_nxv2i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsubu.mask.nxv2i32.nxv2i16.i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i16> %1,
+    i16 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsubu.nxv4i32.nxv4i16.i16(
+  <vscale x 4 x i16>,
+  i16,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsubu_vx_nxv4i32_nxv4i16_i16(<vscale x 4 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vx_nxv4i32_nxv4i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsubu.nxv4i32.nxv4i16.i16(
+    <vscale x 4 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsubu.mask.nxv4i32.nxv4i16.i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  i16,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsubu_mask_vx_nxv4i32_nxv4i16_i16(<vscale x 4 x i32> %0, <vscale x 4 x i16> %1, i16 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vx_nxv4i32_nxv4i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsubu.mask.nxv4i32.nxv4i16.i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i16> %1,
+    i16 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsubu.nxv8i32.nxv8i16.i16(
+  <vscale x 8 x i16>,
+  i16,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsubu_vx_nxv8i32_nxv8i16_i16(<vscale x 8 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vx_nxv8i32_nxv8i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsubu.nxv8i32.nxv8i16.i16(
+    <vscale x 8 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsubu.mask.nxv8i32.nxv8i16.i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  i16,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsubu_mask_vx_nxv8i32_nxv8i16_i16(<vscale x 8 x i32> %0, <vscale x 8 x i16> %1, i16 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vx_nxv8i32_nxv8i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsubu.mask.nxv8i32.nxv8i16.i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i16> %1,
+    i16 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsubu.nxv16i32.nxv16i16.i16(
+  <vscale x 16 x i16>,
+  i16,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsubu_vx_nxv16i32_nxv16i16_i16(<vscale x 16 x i16> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_vx_nxv16i32_nxv16i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsubu.nxv16i32.nxv16i16.i16(
+    <vscale x 16 x i16> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsubu.mask.nxv16i32.nxv16i16.i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  i16,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsubu_mask_vx_nxv16i32_nxv16i16_i16(<vscale x 16 x i32> %0, <vscale x 16 x i16> %1, i16 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu_mask_vx_nxv16i32_nxv16i16_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsubu.vx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsubu.mask.nxv16i32.nxv16i16.i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i16> %1,
+    i16 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}

llvm/test/CodeGen/RISCV/rvv/vwsubu.w-rv32.ll

@@ -0,0 +1,881 @@
+; RUN: llc -mtriple=riscv32 -mattr=+experimental-v -verify-machineinstrs \
+; RUN:   --riscv-no-aliases < %s | FileCheck %s
+declare <vscale x 1 x i16> @llvm.riscv.vwsubu.w.nxv1i16.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsubu.w_wv_nxv1i16_nxv1i16_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wv_nxv1i16_nxv1i16_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsubu.w.nxv1i16.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsubu.w.mask.nxv1i16.nxv1i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i8>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsubu.w_mask_wv_nxv1i16_nxv1i16_nxv1i8(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, <vscale x 1 x i8> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wv_nxv1i16_nxv1i16_nxv1i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsubu.w.mask.nxv1i16.nxv1i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    <vscale x 1 x i8> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsubu.w.nxv2i16.nxv2i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsubu.w_wv_nxv2i16_nxv2i16_nxv2i8(<vscale x 2 x i16> %0, <vscale x 2 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wv_nxv2i16_nxv2i16_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsubu.w.nxv2i16.nxv2i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsubu.w.mask.nxv2i16.nxv2i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i8>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsubu.w_mask_wv_nxv2i16_nxv2i16_nxv2i8(<vscale x 2 x i16> %0, <vscale x 2 x i16> %1, <vscale x 2 x i8> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wv_nxv2i16_nxv2i16_nxv2i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsubu.w.mask.nxv2i16.nxv2i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i16> %1,
+    <vscale x 2 x i8> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsubu.w.nxv4i16.nxv4i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsubu.w_wv_nxv4i16_nxv4i16_nxv4i8(<vscale x 4 x i16> %0, <vscale x 4 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wv_nxv4i16_nxv4i16_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsubu.w.nxv4i16.nxv4i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsubu.w.mask.nxv4i16.nxv4i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i8>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsubu.w_mask_wv_nxv4i16_nxv4i16_nxv4i8(<vscale x 4 x i16> %0, <vscale x 4 x i16> %1, <vscale x 4 x i8> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wv_nxv4i16_nxv4i16_nxv4i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsubu.w.mask.nxv4i16.nxv4i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i16> %1,
+    <vscale x 4 x i8> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsubu.w.nxv8i16.nxv8i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsubu.w_wv_nxv8i16_nxv8i16_nxv8i8(<vscale x 8 x i16> %0, <vscale x 8 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wv_nxv8i16_nxv8i16_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsubu.w.nxv8i16.nxv8i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsubu.w.mask.nxv8i16.nxv8i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i8>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsubu.w_mask_wv_nxv8i16_nxv8i16_nxv8i8(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, <vscale x 8 x i8> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wv_nxv8i16_nxv8i16_nxv8i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsubu.w.mask.nxv8i16.nxv8i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i16> %1,
+    <vscale x 8 x i8> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsubu.w.nxv16i16.nxv16i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsubu.w_wv_nxv16i16_nxv16i16_nxv16i8(<vscale x 16 x i16> %0, <vscale x 16 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wv_nxv16i16_nxv16i16_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsubu.w.nxv16i16.nxv16i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsubu.w.mask.nxv16i16.nxv16i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i8>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsubu.w_mask_wv_nxv16i16_nxv16i16_nxv16i8(<vscale x 16 x i16> %0, <vscale x 16 x i16> %1, <vscale x 16 x i8> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wv_nxv16i16_nxv16i16_nxv16i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsubu.w.mask.nxv16i16.nxv16i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i16> %1,
+    <vscale x 16 x i8> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsubu.w.nxv32i16.nxv32i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsubu.w_wv_nxv32i16_nxv32i16_nxv32i8(<vscale x 32 x i16> %0, <vscale x 32 x i8> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wv_nxv32i16_nxv32i16_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsubu.w.nxv32i16.nxv32i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i8> %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsubu.w.mask.nxv32i16.nxv32i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i16>,
+  <vscale x 32 x i8>,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsubu.w_mask_wv_nxv32i16_nxv32i16_nxv32i8(<vscale x 32 x i16> %0, <vscale x 32 x i16> %1, <vscale x 32 x i8> %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wv_nxv32i16_nxv32i16_nxv32i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsubu.w.mask.nxv32i16.nxv32i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i16> %1,
+    <vscale x 32 x i8> %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsubu.w.nxv1i32.nxv1i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsubu.w_wv_nxv1i32_nxv1i32_nxv1i16(<vscale x 1 x i32> %0, <vscale x 1 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wv_nxv1i32_nxv1i32_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsubu.w.nxv1i32.nxv1i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsubu.w.mask.nxv1i32.nxv1i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i32>,
+  <vscale x 1 x i16>,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsubu.w_mask_wv_nxv1i32_nxv1i32_nxv1i16(<vscale x 1 x i32> %0, <vscale x 1 x i32> %1, <vscale x 1 x i16> %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wv_nxv1i32_nxv1i32_nxv1i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsubu.w.mask.nxv1i32.nxv1i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i32> %1,
+    <vscale x 1 x i16> %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsubu.w.nxv2i32.nxv2i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsubu.w_wv_nxv2i32_nxv2i32_nxv2i16(<vscale x 2 x i32> %0, <vscale x 2 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wv_nxv2i32_nxv2i32_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsubu.w.nxv2i32.nxv2i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsubu.w.mask.nxv2i32.nxv2i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i32>,
+  <vscale x 2 x i16>,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsubu.w_mask_wv_nxv2i32_nxv2i32_nxv2i16(<vscale x 2 x i32> %0, <vscale x 2 x i32> %1, <vscale x 2 x i16> %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wv_nxv2i32_nxv2i32_nxv2i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsubu.w.mask.nxv2i32.nxv2i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i32> %1,
+    <vscale x 2 x i16> %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsubu.w.nxv4i32.nxv4i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsubu.w_wv_nxv4i32_nxv4i32_nxv4i16(<vscale x 4 x i32> %0, <vscale x 4 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wv_nxv4i32_nxv4i32_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsubu.w.nxv4i32.nxv4i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsubu.w.mask.nxv4i32.nxv4i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i32>,
+  <vscale x 4 x i16>,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsubu.w_mask_wv_nxv4i32_nxv4i32_nxv4i16(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, <vscale x 4 x i16> %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wv_nxv4i32_nxv4i32_nxv4i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsubu.w.mask.nxv4i32.nxv4i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i32> %1,
+    <vscale x 4 x i16> %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsubu.w.nxv8i32.nxv8i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsubu.w_wv_nxv8i32_nxv8i32_nxv8i16(<vscale x 8 x i32> %0, <vscale x 8 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wv_nxv8i32_nxv8i32_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsubu.w.nxv8i32.nxv8i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsubu.w.mask.nxv8i32.nxv8i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i32>,
+  <vscale x 8 x i16>,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsubu.w_mask_wv_nxv8i32_nxv8i32_nxv8i16(<vscale x 8 x i32> %0, <vscale x 8 x i32> %1, <vscale x 8 x i16> %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wv_nxv8i32_nxv8i32_nxv8i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsubu.w.mask.nxv8i32.nxv8i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i32> %1,
+    <vscale x 8 x i16> %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsubu.w.nxv16i32.nxv16i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsubu.w_wv_nxv16i32_nxv16i32_nxv16i16(<vscale x 16 x i32> %0, <vscale x 16 x i16> %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wv_nxv16i32_nxv16i32_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsubu.w.nxv16i32.nxv16i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i16> %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsubu.w.mask.nxv16i32.nxv16i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i32>,
+  <vscale x 16 x i16>,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsubu.w_mask_wv_nxv16i32_nxv16i32_nxv16i16(<vscale x 16 x i32> %0, <vscale x 16 x i32> %1, <vscale x 16 x i16> %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wv_nxv16i32_nxv16i32_nxv16i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsubu.wv {{v[0-9]+}}, {{v[0-9]+}}, {{v[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsubu.w.mask.nxv16i32.nxv16i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i32> %1,
+    <vscale x 16 x i16> %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsubu.w.nxv1i16.i8(
+  <vscale x 1 x i16>,
+  i8,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsubu.w_wx_nxv1i16_nxv1i16_i8(<vscale x 1 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wx_nxv1i16_nxv1i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsubu.w.nxv1i16.i8(
+    <vscale x 1 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 1 x i16> @llvm.riscv.vwsubu.w.mask.nxv1i16.i8(
+  <vscale x 1 x i16>,
+  <vscale x 1 x i16>,
+  i8,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i16> @intrinsic_vwsubu.w_mask_wx_nxv1i16_nxv1i16_i8(<vscale x 1 x i16> %0, <vscale x 1 x i16> %1, i8 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wx_nxv1i16_nxv1i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf8,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i16> @llvm.riscv.vwsubu.w.mask.nxv1i16.i8(
+    <vscale x 1 x i16> %0,
+    <vscale x 1 x i16> %1,
+    i8 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsubu.w.nxv2i16.i8(
+  <vscale x 2 x i16>,
+  i8,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsubu.w_wx_nxv2i16_nxv2i16_i8(<vscale x 2 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wx_nxv2i16_nxv2i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsubu.w.nxv2i16.i8(
+    <vscale x 2 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 2 x i16> @llvm.riscv.vwsubu.w.mask.nxv2i16.i8(
+  <vscale x 2 x i16>,
+  <vscale x 2 x i16>,
+  i8,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i16> @intrinsic_vwsubu.w_mask_wx_nxv2i16_nxv2i16_i8(<vscale x 2 x i16> %0, <vscale x 2 x i16> %1, i8 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wx_nxv2i16_nxv2i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf4,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i16> @llvm.riscv.vwsubu.w.mask.nxv2i16.i8(
+    <vscale x 2 x i16> %0,
+    <vscale x 2 x i16> %1,
+    i8 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsubu.w.nxv4i16.i8(
+  <vscale x 4 x i16>,
+  i8,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsubu.w_wx_nxv4i16_nxv4i16_i8(<vscale x 4 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wx_nxv4i16_nxv4i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsubu.w.nxv4i16.i8(
+    <vscale x 4 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 4 x i16> @llvm.riscv.vwsubu.w.mask.nxv4i16.i8(
+  <vscale x 4 x i16>,
+  <vscale x 4 x i16>,
+  i8,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i16> @intrinsic_vwsubu.w_mask_wx_nxv4i16_nxv4i16_i8(<vscale x 4 x i16> %0, <vscale x 4 x i16> %1, i8 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wx_nxv4i16_nxv4i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,mf2,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i16> @llvm.riscv.vwsubu.w.mask.nxv4i16.i8(
+    <vscale x 4 x i16> %0,
+    <vscale x 4 x i16> %1,
+    i8 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsubu.w.nxv8i16.i8(
+  <vscale x 8 x i16>,
+  i8,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsubu.w_wx_nxv8i16_nxv8i16_i8(<vscale x 8 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wx_nxv8i16_nxv8i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsubu.w.nxv8i16.i8(
+    <vscale x 8 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 8 x i16> @llvm.riscv.vwsubu.w.mask.nxv8i16.i8(
+  <vscale x 8 x i16>,
+  <vscale x 8 x i16>,
+  i8,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i16> @intrinsic_vwsubu.w_mask_wx_nxv8i16_nxv8i16_i8(<vscale x 8 x i16> %0, <vscale x 8 x i16> %1, i8 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wx_nxv8i16_nxv8i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m1,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i16> @llvm.riscv.vwsubu.w.mask.nxv8i16.i8(
+    <vscale x 8 x i16> %0,
+    <vscale x 8 x i16> %1,
+    i8 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsubu.w.nxv16i16.i8(
+  <vscale x 16 x i16>,
+  i8,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsubu.w_wx_nxv16i16_nxv16i16_i8(<vscale x 16 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wx_nxv16i16_nxv16i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsubu.w.nxv16i16.i8(
+    <vscale x 16 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 16 x i16> @llvm.riscv.vwsubu.w.mask.nxv16i16.i8(
+  <vscale x 16 x i16>,
+  <vscale x 16 x i16>,
+  i8,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i16> @intrinsic_vwsubu.w_mask_wx_nxv16i16_nxv16i16_i8(<vscale x 16 x i16> %0, <vscale x 16 x i16> %1, i8 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wx_nxv16i16_nxv16i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m2,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i16> @llvm.riscv.vwsubu.w.mask.nxv16i16.i8(
+    <vscale x 16 x i16> %0,
+    <vscale x 16 x i16> %1,
+    i8 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsubu.w.nxv32i16.i8(
+  <vscale x 32 x i16>,
+  i8,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsubu.w_wx_nxv32i16_nxv32i16_i8(<vscale x 32 x i16> %0, i8 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wx_nxv32i16_nxv32i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsubu.w.nxv32i16.i8(
+    <vscale x 32 x i16> %0,
+    i8 %1,
+    i32 %2)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 32 x i16> @llvm.riscv.vwsubu.w.mask.nxv32i16.i8(
+  <vscale x 32 x i16>,
+  <vscale x 32 x i16>,
+  i8,
+  <vscale x 32 x i1>,
+  i32);
+
+define <vscale x 32 x i16> @intrinsic_vwsubu.w_mask_wx_nxv32i16_nxv32i16_i8(<vscale x 32 x i16> %0, <vscale x 32 x i16> %1, i8 %2, <vscale x 32 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wx_nxv32i16_nxv32i16_i8
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e8,m4,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 32 x i16> @llvm.riscv.vwsubu.w.mask.nxv32i16.i8(
+    <vscale x 32 x i16> %0,
+    <vscale x 32 x i16> %1,
+    i8 %2,
+    <vscale x 32 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 32 x i16> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsubu.w.nxv1i32.i16(
+  <vscale x 1 x i32>,
+  i16,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsubu.w_wx_nxv1i32_nxv1i32_i16(<vscale x 1 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wx_nxv1i32_nxv1i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsubu.w.nxv1i32.i16(
+    <vscale x 1 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 1 x i32> @llvm.riscv.vwsubu.w.mask.nxv1i32.i16(
+  <vscale x 1 x i32>,
+  <vscale x 1 x i32>,
+  i16,
+  <vscale x 1 x i1>,
+  i32);
+
+define <vscale x 1 x i32> @intrinsic_vwsubu.w_mask_wx_nxv1i32_nxv1i32_i16(<vscale x 1 x i32> %0, <vscale x 1 x i32> %1, i16 %2, <vscale x 1 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wx_nxv1i32_nxv1i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf4,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 1 x i32> @llvm.riscv.vwsubu.w.mask.nxv1i32.i16(
+    <vscale x 1 x i32> %0,
+    <vscale x 1 x i32> %1,
+    i16 %2,
+    <vscale x 1 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 1 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsubu.w.nxv2i32.i16(
+  <vscale x 2 x i32>,
+  i16,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsubu.w_wx_nxv2i32_nxv2i32_i16(<vscale x 2 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wx_nxv2i32_nxv2i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsubu.w.nxv2i32.i16(
+    <vscale x 2 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 2 x i32> @llvm.riscv.vwsubu.w.mask.nxv2i32.i16(
+  <vscale x 2 x i32>,
+  <vscale x 2 x i32>,
+  i16,
+  <vscale x 2 x i1>,
+  i32);
+
+define <vscale x 2 x i32> @intrinsic_vwsubu.w_mask_wx_nxv2i32_nxv2i32_i16(<vscale x 2 x i32> %0, <vscale x 2 x i32> %1, i16 %2, <vscale x 2 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wx_nxv2i32_nxv2i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,mf2,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 2 x i32> @llvm.riscv.vwsubu.w.mask.nxv2i32.i16(
+    <vscale x 2 x i32> %0,
+    <vscale x 2 x i32> %1,
+    i16 %2,
+    <vscale x 2 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 2 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsubu.w.nxv4i32.i16(
+  <vscale x 4 x i32>,
+  i16,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsubu.w_wx_nxv4i32_nxv4i32_i16(<vscale x 4 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wx_nxv4i32_nxv4i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsubu.w.nxv4i32.i16(
+    <vscale x 4 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 4 x i32> @llvm.riscv.vwsubu.w.mask.nxv4i32.i16(
+  <vscale x 4 x i32>,
+  <vscale x 4 x i32>,
+  i16,
+  <vscale x 4 x i1>,
+  i32);
+
+define <vscale x 4 x i32> @intrinsic_vwsubu.w_mask_wx_nxv4i32_nxv4i32_i16(<vscale x 4 x i32> %0, <vscale x 4 x i32> %1, i16 %2, <vscale x 4 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wx_nxv4i32_nxv4i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m1,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 4 x i32> @llvm.riscv.vwsubu.w.mask.nxv4i32.i16(
+    <vscale x 4 x i32> %0,
+    <vscale x 4 x i32> %1,
+    i16 %2,
+    <vscale x 4 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 4 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsubu.w.nxv8i32.i16(
+  <vscale x 8 x i32>,
+  i16,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsubu.w_wx_nxv8i32_nxv8i32_i16(<vscale x 8 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wx_nxv8i32_nxv8i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsubu.w.nxv8i32.i16(
+    <vscale x 8 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 8 x i32> @llvm.riscv.vwsubu.w.mask.nxv8i32.i16(
+  <vscale x 8 x i32>,
+  <vscale x 8 x i32>,
+  i16,
+  <vscale x 8 x i1>,
+  i32);
+
+define <vscale x 8 x i32> @intrinsic_vwsubu.w_mask_wx_nxv8i32_nxv8i32_i16(<vscale x 8 x i32> %0, <vscale x 8 x i32> %1, i16 %2, <vscale x 8 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wx_nxv8i32_nxv8i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m2,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 8 x i32> @llvm.riscv.vwsubu.w.mask.nxv8i32.i16(
+    <vscale x 8 x i32> %0,
+    <vscale x 8 x i32> %1,
+    i16 %2,
+    <vscale x 8 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 8 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsubu.w.nxv16i32.i16(
+  <vscale x 16 x i32>,
+  i16,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsubu.w_wx_nxv16i32_nxv16i32_i16(<vscale x 16 x i32> %0, i16 %1, i32 %2) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_wx_nxv16i32_nxv16i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsubu.w.nxv16i32.i16(
+    <vscale x 16 x i32> %0,
+    i16 %1,
+    i32 %2)
+
+  ret <vscale x 16 x i32> %a
+}
+
+declare <vscale x 16 x i32> @llvm.riscv.vwsubu.w.mask.nxv16i32.i16(
+  <vscale x 16 x i32>,
+  <vscale x 16 x i32>,
+  i16,
+  <vscale x 16 x i1>,
+  i32);
+
+define <vscale x 16 x i32> @intrinsic_vwsubu.w_mask_wx_nxv16i32_nxv16i32_i16(<vscale x 16 x i32> %0, <vscale x 16 x i32> %1, i16 %2, <vscale x 16 x i1> %3, i32 %4) nounwind {
+entry:
+; CHECK-LABEL: intrinsic_vwsubu.w_mask_wx_nxv16i32_nxv16i32_i16
+; CHECK:       vsetvli {{.*}}, {{a[0-9]+}}, e16,m4,ta,mu
+; CHECK:       vwsubu.wx {{v[0-9]+}}, {{v[0-9]+}}, {{a[0-9]+}}, v0.t
+  %a = call <vscale x 16 x i32> @llvm.riscv.vwsubu.w.mask.nxv16i32.i16(
+    <vscale x 16 x i32> %0,
+    <vscale x 16 x i32> %1,
+    i16 %2,
+    <vscale x 16 x i1> %3,
+    i32 %4)
+
+  ret <vscale x 16 x i32> %a
+}