[RISCV] Remove legacy TA/TU pseudo distinction for VID

[reames]

Summary

This is a follow on to D152740. The focus of this patch is on actually removing the old TA (unsuffixed) version. I realized we already had plumbing for combined TA/TU pseudos - used by some of the ternary instructions. As such, we can go ahead and fully remove the old TA, and rename the _TU variant to be unsuffixed. (The rename must happen in this patch for the table structure to work out as expected.)

The scheduling difference comes from an omission in D152740. If we selected a _MASK variant - either from manual ISEL or instrincs - we were going through doPeepholeMaskedRVV and still getting the TA variant. The use of the IsCombined flag in the MaskedPseudo table causes us to use the TU (now unsuffixed) variant instead.

[reames]

Ignore for now, uploaded from patch and have to fix once back from running an errand.

[reames]

Reopen with corrected patch

[craig.topper]

llvm/lib/Target/RISCV/RISCVInstrInfoVPseudos.td

477–489

psuedu.-> pseudo

485

Why does the second sentence start with "If"? Do some instructions in this category not have a policy operand?

1821

Can we wrap this closer to 80 columns?

4311

I think the undef in this pattern will become IMPLICIT_DEF on its own. Do we need this pattern?

[craig.topper]

llvm/lib/Target/RISCV/RISCVInstrInfoVPseudos.td

930

Line this up with RegClass:$merge on the previous line. That's usually how I write multiline operand lists.

[reames]

Address review comments

[craig.topper]

llvm/lib/Target/RISCV/RISCVInstrInfoVPseudos.td

477–489

pseudu -> pseudo

4321

Didn't we just add a TU_MU defvar? I thought we were moving to shorter name.

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fmf.ll

12 ↗

(On Diff #533800)

Is this change related?

[reames]

llvm/lib/Target/RISCV/RISCVInstrInfoVPseudos.td

477–489

I really don't seem to be able to spell in this changeset...

4321

We did. I'd just written this code in a different order than I'd posted patches and forgot to update.

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-fmf.ll

12 ↗

(On Diff #533800)

Nope, spurious regen change. Separate and landed.

[reames]

Address review comments.

[craig.topper]

LGTM

llvm/lib/Target/RISCV/RISCVInstrInfoVPseudos.td

 def RISCVVIntrinsicsTable : GenericTable {
   let FilterClass = "RISCVVIntrinsic";
   let CppTypeName = "RISCVVIntrinsicInfo";
   let Fields = ["IntrinsicID", "ScalarOperand", "VLOperand"];
   let PrimaryKey = ["IntrinsicID"];
   let PrimaryKeyName = "getRISCVVIntrinsicInfo";
 }
 
-class RISCVMaskedPseudo<bits<4> MaskIdx, bit HasTU = true, bit IsTernary = false> {
+// Describes the relation of a masked psuedu to the unmasked variants.
+// (HasTU = true, IsCombined = false)
+//    Has both a TA (unsuffixed) and _TU variant defined.  TU variant
+//    may (or may not) have a vector policy operand.
+// (HasTU = false, IsCombined = false)
+//    No TA (unsuffixed) or _TU variants.  Masked version is only pseudo
+// (HasTU = false, IsCombined = true)
+//    The unsuffixed version has a merge operand; no explicit _TU variant
+//    exists.  If the unsuffixed version has a policy operand, and can thus
+//    represent all policy states.
+// (HasTU = true, IsCombined = true)
+//    Invalid and unused state.
+class RISCVMaskedPseudo<bits<4> MaskIdx, bit HasTU = true, bit IsCombined = false> {
   Pseudo MaskedPseudo = !cast<Pseudo>(NAME);
   Pseudo UnmaskedPseudo = !cast<Pseudo>(!subst("_MASK", "", NAME));
   Pseudo UnmaskedTUPseudo = !cond(HasTU : !cast<Pseudo>(!subst("_MASK", "", NAME # "_TU")),
-                                  IsTernary : UnmaskedPseudo,
+                                  IsCombined : UnmaskedPseudo,
                                   true  :  MaskedPseudo);
   bits<4> MaskOpIdx = MaskIdx;
 }
 
 def RISCVMaskedPseudosTable : GenericTable {
   let FilterClass = "RISCVMaskedPseudo";
   let CppTypeName = "RISCVMaskedPseudoInfo";
   let Fields = ["MaskedPseudo", "UnmaskedPseudo", "UnmaskedTUPseudo", "MaskOpIdx"];
class VPseudoSStoreMask<VReg StClass, int EEW>:
   let mayStore = 1;
   let hasSideEffects = 0;
   let HasVLOp = 1;
   let HasSEWOp = 1;
 }
 
 class VPseudoNullaryNoMask<VReg RegClass>:
       Pseudo<(outs RegClass:$rd),
-             (ins AVL:$vl, ixlenimm:$sew),
-             []>, RISCVVPseudo {
-  let mayLoad = 0;
-  let mayStore = 0;
-  let hasSideEffects = 0;
-  let HasVLOp = 1;
-  let HasSEWOp = 1;
-}
-
-class VPseudoNullaryNoMaskTU<VReg RegClass>:
-      Pseudo<(outs RegClass:$rd),
-             (ins RegClass:$merge, AVL:$vl, ixlenimm:$sew),
-             []>, RISCVVPseudo {
+             (ins RegClass:$merge, AVL:$vl, ixlenimm:$sew,
+              ixlenimm:$policy), []>, RISCVVPseudo {
   let mayLoad = 0;
   let mayStore = 0;
   let hasSideEffects = 0;
   let Constraints = "$rd = $merge";
   let HasVLOp = 1;
   let HasSEWOp = 1;
+  let HasVecPolicyOp = 1;
 }
 
 class VPseudoNullaryMask<VReg RegClass>:
       Pseudo<(outs GetVRegNoV0<RegClass>.R:$rd),
              (ins GetVRegNoV0<RegClass>.R:$merge, VMaskOp:$vm, AVL:$vl,
               ixlenimm:$sew, ixlenimm:$policy), []>, RISCVVPseudo {
   let mayLoad = 0;
   let mayStore = 0;
multiclass VPseudoVID_V {
   foreach m = MxList in {
     defvar mx = m.MX;
     defvar WriteVMIdxV_MX = !cast<SchedWrite>("WriteVMIdxV_" # mx);
     defvar ReadVMIdxV_MX = !cast<SchedRead>("ReadVMIdxV_" # mx);
 
     let VLMul = m.value in {
       def "_V_" # m.MX : VPseudoNullaryNoMask<m.vrclass>,
                          Sched<[WriteVMIdxV_MX, ReadVMask]>;
-      def "_V_" # m.MX # "_TU": VPseudoNullaryNoMaskTU<m.vrclass>,
-                                Sched<[WriteVMIdxV_MX, ReadVMask]>;
       def "_V_" # m.MX # "_MASK" : VPseudoNullaryMask<m.vrclass>,
-                                   RISCVMaskedPseudo</*MaskOpIdx*/ 1>,
+                                   RISCVMaskedPseudo</*MaskOpIdx*/ 1, /*HasTU=*/false, /*IsCombined=*/true>,
                                    Sched<[WriteVMIdxV_MX, ReadVMask]>;
     }
   }
 }
 
 multiclass VPseudoNullaryPseudoM <string BaseInst> {
   foreach mti = AllMasks in {
     defvar mx = mti.LMul.MX;
multiclass VPseudoTernaryWithPolicy<VReg RetClass,
                                     DAGOperand Op2Class,
                                     LMULInfo MInfo,
                                     string Constraint = "",
                                     bit Commutable = 0> {
   let VLMul = MInfo.value in {
     let isCommutable = Commutable in
     def "_" # MInfo.MX : VPseudoTernaryNoMaskWithPolicy<RetClass, Op1Class, Op2Class, Constraint>;
     def "_" # MInfo.MX # "_MASK" : VPseudoBinaryMaskPolicy<RetClass, Op1Class, Op2Class, Constraint>,
-                                   RISCVMaskedPseudo</*MaskOpIdx*/ 3, /*HasTU*/ false, /*IsTernary*/true>;
+                                   RISCVMaskedPseudo</*MaskOpIdx*/ 3, /*HasTU*/ false, /*IsCombined*/true>;
   }
 }
 
 multiclass VPseudoTernaryV_VV_AAXA<LMULInfo m, string Constraint = ""> {
   defm _VV : VPseudoTernaryWithPolicy<m.vrclass, m.vrclass, m.vrclass, m,
                                       Constraint, /*Commutable*/1>;
 }
 
foreach vti = vtilist in {
     }
   }
 }
 
 multiclass VPatNullaryV<string intrinsic, string instruction>
 {
   foreach vti = AllIntegerVectors in {
     let Predicates = GetVTypePredicates<vti>.Predicates in {
       def : Pat<(vti.Vector (!cast<Intrinsic>(intrinsic)
                             (vti.Vector undef),
                             VLOpFrag)),
-                            (!cast<Instruction>(instruction#"_V_" # vti.LMul.MX # "_TU")
-                            (vti.Vector (IMPLICIT_DEF)), GPR:$vl, vti.Log2SEW)>;
+                            (!cast<Instruction>(instruction#"_V_" # vti.LMul.MX)
+                            (vti.Vector (IMPLICIT_DEF)), GPR:$vl, vti.Log2SEW, TA_MA)>;
       def : Pat<(vti.Vector (!cast<Intrinsic>(intrinsic)
                             (vti.Vector vti.RegClass:$merge),
                             VLOpFrag)),
-                            (!cast<Instruction>(instruction#"_V_" # vti.LMul.MX # "_TU")
-                            vti.RegClass:$merge, GPR:$vl, vti.Log2SEW)>;
+                            (!cast<Instruction>(instruction#"_V_" # vti.LMul.MX)
+                            vti.RegClass:$merge, GPR:$vl, vti.Log2SEW,
+                            TAIL_UNDISTURBED_MASK_UNDISTURBED)>;
       def : Pat<(vti.Vector (!cast<Intrinsic>(intrinsic # "_mask")
                             (vti.Vector vti.RegClass:$merge),
                             (vti.Mask V0), VLOpFrag, (XLenVT timm:$policy))),
                             (!cast<Instruction>(instruction#"_V_" # vti.LMul.MX # "_MASK")
                             vti.RegClass:$merge, (vti.Mask V0),
                             GPR:$vl, vti.Log2SEW, (XLenVT timm:$policy))>;
   }
   }

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int-buildvec.ll

 ; CHECK-NEXT:    ret
   ret <4 x i8> <i8 3, i8 1, i8 -1, i8 -3>
 }
 
 define <4 x i8> @buildvec_vid_stepn3_add3_v4i8() {
 ; CHECK-LABEL: buildvec_vid_stepn3_add3_v4i8:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    vsetivli zero, 4, e8, mf4, ta, ma
-; CHECK-NEXT:    vmv.v.i v9, 3
-; CHECK-NEXT:    vid.v v8
+; CHECK-NEXT:    vid.v v9
+; CHECK-NEXT:    vmv.v.i v8, 3
 ; CHECK-NEXT:    li a0, -3
-; CHECK-NEXT:    vmadd.vx v8, a0, v9
+; CHECK-NEXT:    vmacc.vx v8, a0, v9
 ; CHECK-NEXT:    ret
   ret <4 x i8> <i8 3, i8 0, i8 -3, i8 -6>
 }
 
 define void @buildvec_vid_stepn3_addn3_v4i32(ptr %z0, ptr %z1, ptr %z2, ptr %z3) {
 ; CHECK-LABEL: buildvec_vid_stepn3_addn3_v4i32:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    vsetivli zero, 4, e32, m1, ta, ma
-; CHECK-NEXT:    vmv.v.i v8, -3
-; CHECK-NEXT:    vid.v v9
+; CHECK-NEXT:    vid.v v8
+; CHECK-NEXT:    vmv.v.i v9, -3
 ; CHECK-NEXT:    li a4, -3
-; CHECK-NEXT:    vmadd.vx v9, a4, v8
+; CHECK-NEXT:    vmacc.vx v9, a4, v8
 ; CHECK-NEXT:    vse32.v v9, (a0)
 ; CHECK-NEXT:    vse32.v v9, (a1)
 ; CHECK-NEXT:    vse32.v v9, (a2)
 ; CHECK-NEXT:    vse32.v v9, (a3)
 ; CHECK-NEXT:    ret
   store <4 x i32> <i32 -3, i32 -6, i32 -9, i32 -12>, ptr %z0
   store <4 x i32> <i32 undef, i32 -6, i32 -9, i32 -12>, ptr %z1
   store <4 x i32> <i32 undef, i32 undef, i32 -9, i32 -12>, ptr %z2

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-int.ll

 }
 
 define void @mulhs_v6i16(ptr %x) {
 ; CHECK-LABEL: mulhs_v6i16:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    vsetivli zero, 6, e16, m1, ta, ma
 ; CHECK-NEXT:    vle16.v v8, (a0)
 ; CHECK-NEXT:    vsetivli zero, 2, e16, mf4, ta, ma
-; CHECK-NEXT:    vmv.v.i v9, 7
-; CHECK-NEXT:    vid.v v10
+; CHECK-NEXT:    vid.v v9
+; CHECK-NEXT:    vmv.v.i v10, 7
 ; CHECK-NEXT:    li a1, -14
-; CHECK-NEXT:    vmadd.vx v10, a1, v9
+; CHECK-NEXT:    vmacc.vx v10, a1, v9
 ; CHECK-NEXT:    vsetivli zero, 2, e16, m1, ta, ma
 ; CHECK-NEXT:    vslidedown.vi v9, v8, 4
 ; CHECK-NEXT:    vsetivli zero, 2, e16, mf4, ta, ma
 ; CHECK-NEXT:    vdiv.vv v9, v9, v10
 ; CHECK-NEXT:    li a1, 6
 ; CHECK-NEXT:    vmv.s.x v0, a1
 ; CHECK-NEXT:    vsetivli zero, 4, e16, mf2, ta, ma
 ; CHECK-NEXT:    vmv.v.i v10, -7

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-interleaved-access.ll

 ; RV64-NEXT:    slli a2, a1, 6
 ; RV64-NEXT:    add a1, a2, a1
 ; RV64-NEXT:    add a1, sp, a1
 ; RV64-NEXT:    addi a1, a1, 16
 ; RV64-NEXT:    vl8r.v v24, (a1) # Unknown-size Folded Reload
 ; RV64-NEXT:    vrgather.vv v8, v24, v16, v0.t
 ; RV64-NEXT:    vsetivli zero, 5, e64, m4, tu, ma
 ; RV64-NEXT:    vmv.v.v v4, v8
-; RV64-NEXT:    addi a1, a0, 320
 ; RV64-NEXT:    vsetivli zero, 8, e64, m4, ta, ma
+; RV64-NEXT:    csrr a1, vlenb
+; RV64-NEXT:    li a2, 29
+; RV64-NEXT:    mul a1, a1, a2
+; RV64-NEXT:    add a1, sp, a1
+; RV64-NEXT:    addi a1, a1, 16
+; RV64-NEXT:    vl4r.v v8, (a1) # Unknown-size Folded Reload
+; RV64-NEXT:    vse64.v v8, (a0)
+; RV64-NEXT:    addi a1, a0, 320
 ; RV64-NEXT:    vse64.v v4, (a1)
 ; RV64-NEXT:    addi a1, a0, 256
 ; RV64-NEXT:    csrr a2, vlenb
 ; RV64-NEXT:    slli a3, a2, 5
 ; RV64-NEXT:    add a2, a3, a2
 ; RV64-NEXT:    add a2, sp, a2
 ; RV64-NEXT:    addi a2, a2, 16
 ; RV64-NEXT:    vl4r.v v8, (a2) # Unknown-size Folded Reload

 ; RV64-NEXT:    addi a1, a0, 128
 ; RV64-NEXT:    csrr a2, vlenb
 ; RV64-NEXT:    li a3, 45
 ; RV64-NEXT:    mul a2, a2, a3
 ; RV64-NEXT:    add a2, sp, a2
 ; RV64-NEXT:    addi a2, a2, 16
 ; RV64-NEXT:    vl4r.v v8, (a2) # Unknown-size Folded Reload
 ; RV64-NEXT:    vse64.v v8, (a1)
-; RV64-NEXT:    addi a1, a0, 64
-; RV64-NEXT:    csrr a2, vlenb
-; RV64-NEXT:    li a3, 25
-; RV64-NEXT:    mul a2, a2, a3
-; RV64-NEXT:    add a2, sp, a2
-; RV64-NEXT:    addi a2, a2, 16
-; RV64-NEXT:    vl4r.v v8, (a2) # Unknown-size Folded Reload
-; RV64-NEXT:    vse64.v v8, (a1)
+; RV64-NEXT:    addi a0, a0, 64
 ; RV64-NEXT:    csrr a1, vlenb
-; RV64-NEXT:    li a2, 29
+; RV64-NEXT:    li a2, 25
 ; RV64-NEXT:    mul a1, a1, a2
 ; RV64-NEXT:    add a1, sp, a1
 ; RV64-NEXT:    addi a1, a1, 16
 ; RV64-NEXT:    vl4r.v v8, (a1) # Unknown-size Folded Reload
 ; RV64-NEXT:    vse64.v v8, (a0)
 ; RV64-NEXT:    csrr a0, vlenb
 ; RV64-NEXT:    li a1, 90
 ; RV64-NEXT:    mul a0, a0, a1

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-reduction-int-vp.ll

 
 define signext i8 @vpreduce_mul_v64i8(i8 signext %s, <64 x i8> %v, <64 x i1> %m, i32 zeroext %evl) {
 ; RV32-LABEL: vpreduce_mul_v64i8:
 ; RV32:       # %bb.0:
 ; RV32-NEXT:    addi sp, sp, -16
 ; RV32-NEXT:    .cfi_def_cfa_offset 16
 ; RV32-NEXT:    sw ra, 12(sp) # 4-byte Folded Spill
 ; RV32-NEXT:    .cfi_offset ra, -4
-; RV32-NEXT:    li a3, 32
-; RV32-NEXT:    vsetvli zero, a3, e32, m8, ta, ma
 ; RV32-NEXT:    lui a2, %hi(.LCPI72_0)
 ; RV32-NEXT:    addi a2, a2, %lo(.LCPI72_0)
+; RV32-NEXT:    li a3, 32
+; RV32-NEXT:    vsetvli zero, a3, e32, m8, ta, ma
 ; RV32-NEXT:    vle32.v v16, (a2)
 ; RV32-NEXT:    mv a2, a0
-; RV32-NEXT:    vid.v v24
-; RV32-NEXT:    vmsltu.vx v12, v24, a1
+; RV32-NEXT:    vmsltu.vx v12, v16, a1
+; RV32-NEXT:    vid.v v16
 ; RV32-NEXT:    vmsltu.vx v13, v16, a1
 ; RV32-NEXT:    vsetivli zero, 8, e8, mf2, ta, ma
-; RV32-NEXT:    vslideup.vi v12, v13, 4
+; RV32-NEXT:    vslideup.vi v13, v12, 4
 ; RV32-NEXT:    li a0, 64
 ; RV32-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
-; RV32-NEXT:    vmand.mm v0, v12, v0
+; RV32-NEXT:    vmand.mm v0, v13, v0
 ; RV32-NEXT:    vmv.v.i v12, 1
 ; RV32-NEXT:    vmerge.vvm v8, v12, v8, v0
 ; RV32-NEXT:    vslidedown.vx v12, v8, a3
 ; RV32-NEXT:    vmul.vv v8, v8, v12
 ; RV32-NEXT:    vslidedown.vi v12, v8, 16
 ; RV32-NEXT:    vmul.vv v8, v8, v12
 ; RV32-NEXT:    vslidedown.vi v12, v8, 8
 ; RV32-NEXT:    vmul.vv v8, v8, v12

 ; RV32-NEXT:    ret
 ;
 ; RV64-LABEL: vpreduce_mul_v64i8:
 ; RV64:       # %bb.0:
 ; RV64-NEXT:    addi sp, sp, -16
 ; RV64-NEXT:    .cfi_def_cfa_offset 16
 ; RV64-NEXT:    sd ra, 8(sp) # 8-byte Folded Spill
 ; RV64-NEXT:    .cfi_offset ra, -8
-; RV64-NEXT:    li a3, 32
-; RV64-NEXT:    vsetvli zero, a3, e32, m8, ta, ma
 ; RV64-NEXT:    lui a2, %hi(.LCPI72_0)
 ; RV64-NEXT:    addi a2, a2, %lo(.LCPI72_0)
+; RV64-NEXT:    li a3, 32
+; RV64-NEXT:    vsetvli zero, a3, e32, m8, ta, ma
 ; RV64-NEXT:    vle32.v v16, (a2)
 ; RV64-NEXT:    mv a2, a0
-; RV64-NEXT:    vid.v v24
-; RV64-NEXT:    vmsltu.vx v12, v24, a1
+; RV64-NEXT:    vmsltu.vx v12, v16, a1
+; RV64-NEXT:    vid.v v16
 ; RV64-NEXT:    vmsltu.vx v13, v16, a1
 ; RV64-NEXT:    vsetivli zero, 8, e8, mf2, ta, ma
-; RV64-NEXT:    vslideup.vi v12, v13, 4
+; RV64-NEXT:    vslideup.vi v13, v12, 4
 ; RV64-NEXT:    li a0, 64
 ; RV64-NEXT:    vsetvli zero, a0, e8, m4, ta, ma
-; RV64-NEXT:    vmand.mm v0, v12, v0
+; RV64-NEXT:    vmand.mm v0, v13, v0
 ; RV64-NEXT:    vmv.v.i v12, 1
 ; RV64-NEXT:    vmerge.vvm v8, v12, v8, v0
 ; RV64-NEXT:    vslidedown.vx v12, v8, a3
 ; RV64-NEXT:    vmul.vv v8, v8, v12
 ; RV64-NEXT:    vslidedown.vi v12, v8, 16
 ; RV64-NEXT:    vmul.vv v8, v8, v12
 ; RV64-NEXT:    vslidedown.vi v12, v8, 8
 ; RV64-NEXT:    vmul.vv v8, v8, v12

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-store.ll

; RV64-NEXT:    ret
   store <2 x i16> <i16 3, i16 6>, ptr %p
   ret void
 }
 
 define void @store_constant_v2i32(ptr %p) {
 ; CHECK-LABEL: store_constant_v2i32:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    vsetivli zero, 2, e32, mf2, ta, ma
-; CHECK-NEXT:    vmv.v.i v8, 3
-; CHECK-NEXT:    vid.v v9
+; CHECK-NEXT:    vid.v v8
+; CHECK-NEXT:    vmv.v.i v9, 3
 ; CHECK-NEXT:    li a1, 3
-; CHECK-NEXT:    vmadd.vx v9, a1, v8
+; CHECK-NEXT:    vmacc.vx v9, a1, v8
 ; CHECK-NEXT:    vse32.v v9, (a0)
 ; CHECK-NEXT:    ret
   store <2 x i32> <i32 3, i32 6>, ptr %p
   ret void
 }
 
 define void @store_constant_v4i8(ptr %p) {
 ; RV32-LABEL: store_constant_v4i8:
; RV64-NEXT:    ret
   store <4 x i8> <i8 0, i8 1, i8 2, i8 3>, ptr %p
   ret void
 }
 
 define void @store_constant_v2i8_align1(ptr %p) {
 ; CHECK-LABEL: store_constant_v2i8_align1:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    vsetivli zero, 2, e8, mf8, ta, ma
-; CHECK-NEXT:    vmv.v.i v8, 3
-; CHECK-NEXT:    vid.v v9
+; CHECK-NEXT:    vid.v v8
+; CHECK-NEXT:    vmv.v.i v9, 3
 ; CHECK-NEXT:    li a1, 3
-; CHECK-NEXT:    vmadd.vx v9, a1, v8
+; CHECK-NEXT:    vmacc.vx v9, a1, v8
 ; CHECK-NEXT:    vse8.v v9, (a0)
 ; CHECK-NEXT:    ret
   store <2 x i8> <i8 3, i8 6>, ptr %p, align 1
   ret void
 }
 
 define void @store_constant_splat_v2i8(ptr %p) {
 ; CHECK-LABEL: store_constant_splat_v2i8:

llvm/test/CodeGen/RISCV/rvv/named-vector-shuffle-reverse.ll

 ;
 
 define <vscale x 2 x i1> @reverse_nxv2i1(<vscale x 2 x i1> %a) {
 ; RV32-BITS-UNKNOWN-LABEL: reverse_nxv2i1:
 ; RV32-BITS-UNKNOWN:       # %bb.0:
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-UNKNOWN-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV32-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV32-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV32-BITS-UNKNOWN-NEXT:    srli a0, a0, 2
 ; RV32-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV32-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV32-BITS-UNKNOWN-NEXT:    vrsub.vx v9, v9, a0
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, mf4, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v10, v8, v9
 ; RV32-BITS-UNKNOWN-NEXT:    vand.vi v8, v10, 1
 ; RV32-BITS-UNKNOWN-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV32-BITS-256-LABEL: reverse_nxv2i1:
 ; RV32-BITS-256:       # %bb.0:
 ; RV32-BITS-256-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; RV32-BITS-256-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-256-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-256-NEXT:    vid.v v9
 ; RV32-BITS-256-NEXT:    csrr a0, vlenb
 ; RV32-BITS-256-NEXT:    srli a0, a0, 2
 ; RV32-BITS-256-NEXT:    addi a0, a0, -1
-; RV32-BITS-256-NEXT:    vid.v v9
 ; RV32-BITS-256-NEXT:    vrsub.vx v9, v9, a0
 ; RV32-BITS-256-NEXT:    vrgather.vv v10, v8, v9
 ; RV32-BITS-256-NEXT:    vand.vi v8, v10, 1
 ; RV32-BITS-256-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-256-NEXT:    ret
 ;
 ; RV32-BITS-512-LABEL: reverse_nxv2i1:
 ; RV32-BITS-512:       # %bb.0:
 ; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; RV32-BITS-512-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-512-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-512-NEXT:    vid.v v9
 ; RV32-BITS-512-NEXT:    csrr a0, vlenb
 ; RV32-BITS-512-NEXT:    srli a0, a0, 2
 ; RV32-BITS-512-NEXT:    addi a0, a0, -1
-; RV32-BITS-512-NEXT:    vid.v v9
 ; RV32-BITS-512-NEXT:    vrsub.vx v9, v9, a0
 ; RV32-BITS-512-NEXT:    vrgather.vv v10, v8, v9
 ; RV32-BITS-512-NEXT:    vand.vi v8, v10, 1
 ; RV32-BITS-512-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-512-NEXT:    ret
 ;
 ; RV64-BITS-UNKNOWN-LABEL: reverse_nxv2i1:
 ; RV64-BITS-UNKNOWN:       # %bb.0:
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-UNKNOWN-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
+; RV64-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV64-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV64-BITS-UNKNOWN-NEXT:    srli a0, a0, 2
 ; RV64-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, mf2, ta, ma
-; RV64-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV64-BITS-UNKNOWN-NEXT:    vrsub.vx v9, v9, a0
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, mf4, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v10, v8, v9
 ; RV64-BITS-UNKNOWN-NEXT:    vand.vi v8, v10, 1
 ; RV64-BITS-UNKNOWN-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV64-BITS-256-LABEL: reverse_nxv2i1:
 ; RV64-BITS-256:       # %bb.0:
 ; RV64-BITS-256-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; RV64-BITS-256-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-256-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-256-NEXT:    vid.v v9
 ; RV64-BITS-256-NEXT:    csrr a0, vlenb
 ; RV64-BITS-256-NEXT:    srli a0, a0, 2
 ; RV64-BITS-256-NEXT:    addi a0, a0, -1
-; RV64-BITS-256-NEXT:    vid.v v9
 ; RV64-BITS-256-NEXT:    vrsub.vx v9, v9, a0
 ; RV64-BITS-256-NEXT:    vrgather.vv v10, v8, v9
 ; RV64-BITS-256-NEXT:    vand.vi v8, v10, 1
 ; RV64-BITS-256-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-256-NEXT:    ret
 ;
 ; RV64-BITS-512-LABEL: reverse_nxv2i1:
 ; RV64-BITS-512:       # %bb.0:
 ; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
 ; RV64-BITS-512-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-512-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-512-NEXT:    vid.v v9
 ; RV64-BITS-512-NEXT:    csrr a0, vlenb
 ; RV64-BITS-512-NEXT:    srli a0, a0, 2
 ; RV64-BITS-512-NEXT:    addi a0, a0, -1
-; RV64-BITS-512-NEXT:    vid.v v9
 ; RV64-BITS-512-NEXT:    vrsub.vx v9, v9, a0
 ; RV64-BITS-512-NEXT:    vrgather.vv v10, v8, v9
 ; RV64-BITS-512-NEXT:    vand.vi v8, v10, 1
 ; RV64-BITS-512-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-512-NEXT:    ret
   %res = call <vscale x 2 x i1> @llvm.experimental.vector.reverse.nxv2i1(<vscale x 2 x i1> %a)
   ret <vscale x 2 x i1> %res
 }
 
 define <vscale x 4 x i1> @reverse_nxv4i1(<vscale x 4 x i1> %a) {
 ; RV32-BITS-UNKNOWN-LABEL: reverse_nxv4i1:
 ; RV32-BITS-UNKNOWN:       # %bb.0:
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-UNKNOWN-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, m1, ta, ma
+; RV32-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV32-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV32-BITS-UNKNOWN-NEXT:    srli a0, a0, 1
 ; RV32-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, m1, ta, ma
-; RV32-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV32-BITS-UNKNOWN-NEXT:    vrsub.vx v9, v9, a0
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v10, v8, v9
 ; RV32-BITS-UNKNOWN-NEXT:    vand.vi v8, v10, 1
 ; RV32-BITS-UNKNOWN-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV32-BITS-256-LABEL: reverse_nxv4i1:
 ; RV32-BITS-256:       # %bb.0:
 ; RV32-BITS-256-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; RV32-BITS-256-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-256-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-256-NEXT:    vid.v v9
 ; RV32-BITS-256-NEXT:    csrr a0, vlenb
 ; RV32-BITS-256-NEXT:    srli a0, a0, 1
 ; RV32-BITS-256-NEXT:    addi a0, a0, -1
-; RV32-BITS-256-NEXT:    vid.v v9
 ; RV32-BITS-256-NEXT:    vrsub.vx v9, v9, a0
 ; RV32-BITS-256-NEXT:    vrgather.vv v10, v8, v9
 ; RV32-BITS-256-NEXT:    vand.vi v8, v10, 1
 ; RV32-BITS-256-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-256-NEXT:    ret
 ;
 ; RV32-BITS-512-LABEL: reverse_nxv4i1:
 ; RV32-BITS-512:       # %bb.0:
 ; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; RV32-BITS-512-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-512-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-512-NEXT:    vid.v v9
 ; RV32-BITS-512-NEXT:    csrr a0, vlenb
 ; RV32-BITS-512-NEXT:    srli a0, a0, 1
 ; RV32-BITS-512-NEXT:    addi a0, a0, -1
-; RV32-BITS-512-NEXT:    vid.v v9
 ; RV32-BITS-512-NEXT:    vrsub.vx v9, v9, a0
 ; RV32-BITS-512-NEXT:    vrgather.vv v10, v8, v9
 ; RV32-BITS-512-NEXT:    vand.vi v8, v10, 1
 ; RV32-BITS-512-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-512-NEXT:    ret
 ;
 ; RV64-BITS-UNKNOWN-LABEL: reverse_nxv4i1:
 ; RV64-BITS-UNKNOWN:       # %bb.0:
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-UNKNOWN-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, m1, ta, ma
+; RV64-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV64-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV64-BITS-UNKNOWN-NEXT:    srli a0, a0, 1
 ; RV64-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, m1, ta, ma
-; RV64-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV64-BITS-UNKNOWN-NEXT:    vrsub.vx v9, v9, a0
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v10, v8, v9
 ; RV64-BITS-UNKNOWN-NEXT:    vand.vi v8, v10, 1
 ; RV64-BITS-UNKNOWN-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV64-BITS-256-LABEL: reverse_nxv4i1:
 ; RV64-BITS-256:       # %bb.0:
 ; RV64-BITS-256-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; RV64-BITS-256-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-256-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-256-NEXT:    vid.v v9
 ; RV64-BITS-256-NEXT:    csrr a0, vlenb
 ; RV64-BITS-256-NEXT:    srli a0, a0, 1
 ; RV64-BITS-256-NEXT:    addi a0, a0, -1
-; RV64-BITS-256-NEXT:    vid.v v9
 ; RV64-BITS-256-NEXT:    vrsub.vx v9, v9, a0
 ; RV64-BITS-256-NEXT:    vrgather.vv v10, v8, v9
 ; RV64-BITS-256-NEXT:    vand.vi v8, v10, 1
 ; RV64-BITS-256-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-256-NEXT:    ret
 ;
 ; RV64-BITS-512-LABEL: reverse_nxv4i1:
 ; RV64-BITS-512:       # %bb.0:
 ; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
 ; RV64-BITS-512-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-512-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-512-NEXT:    vid.v v9
 ; RV64-BITS-512-NEXT:    csrr a0, vlenb
 ; RV64-BITS-512-NEXT:    srli a0, a0, 1
 ; RV64-BITS-512-NEXT:    addi a0, a0, -1
-; RV64-BITS-512-NEXT:    vid.v v9
 ; RV64-BITS-512-NEXT:    vrsub.vx v9, v9, a0
 ; RV64-BITS-512-NEXT:    vrgather.vv v10, v8, v9
 ; RV64-BITS-512-NEXT:    vand.vi v8, v10, 1
 ; RV64-BITS-512-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-512-NEXT:    ret
   %res = call <vscale x 4 x i1> @llvm.experimental.vector.reverse.nxv4i1(<vscale x 4 x i1> %a)
   ret <vscale x 4 x i1> %res
 }

 }
 
 define <vscale x 16 x i1> @reverse_nxv16i1(<vscale x 16 x i1> %a) {
 ; RV32-BITS-UNKNOWN-LABEL: reverse_nxv16i1:
 ; RV32-BITS-UNKNOWN:       # %bb.0:
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-UNKNOWN-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, m4, ta, ma
+; RV32-BITS-UNKNOWN-NEXT:    vid.v v12
 ; RV32-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV32-BITS-UNKNOWN-NEXT:    slli a0, a0, 1
 ; RV32-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, m4, ta, ma
-; RV32-BITS-UNKNOWN-NEXT:    vid.v v12
 ; RV32-BITS-UNKNOWN-NEXT:    vrsub.vx v12, v12, a0
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m2, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v10, v8, v12
 ; RV32-BITS-UNKNOWN-NEXT:    vand.vi v8, v10, 1
 ; RV32-BITS-UNKNOWN-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV32-BITS-256-LABEL: reverse_nxv16i1:
 ; RV32-BITS-256:       # %bb.0:
 ; RV32-BITS-256-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
 ; RV32-BITS-256-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-256-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-256-NEXT:    vid.v v10
 ; RV32-BITS-256-NEXT:    csrr a0, vlenb
 ; RV32-BITS-256-NEXT:    slli a0, a0, 1
 ; RV32-BITS-256-NEXT:    addi a0, a0, -1
-; RV32-BITS-256-NEXT:    vid.v v10
 ; RV32-BITS-256-NEXT:    vrsub.vx v10, v10, a0
 ; RV32-BITS-256-NEXT:    vrgather.vv v12, v8, v10
 ; RV32-BITS-256-NEXT:    vand.vi v8, v12, 1
 ; RV32-BITS-256-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-256-NEXT:    ret
 ;
 ; RV32-BITS-512-LABEL: reverse_nxv16i1:
 ; RV32-BITS-512:       # %bb.0:
 ; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
 ; RV32-BITS-512-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-512-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-512-NEXT:    vid.v v10
 ; RV32-BITS-512-NEXT:    csrr a0, vlenb
 ; RV32-BITS-512-NEXT:    slli a0, a0, 1
 ; RV32-BITS-512-NEXT:    addi a0, a0, -1
-; RV32-BITS-512-NEXT:    vid.v v10
 ; RV32-BITS-512-NEXT:    vrsub.vx v10, v10, a0
 ; RV32-BITS-512-NEXT:    vrgather.vv v12, v8, v10
 ; RV32-BITS-512-NEXT:    vand.vi v8, v12, 1
 ; RV32-BITS-512-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-512-NEXT:    ret
 ;
 ; RV64-BITS-UNKNOWN-LABEL: reverse_nxv16i1:
 ; RV64-BITS-UNKNOWN:       # %bb.0:
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-UNKNOWN-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, m4, ta, ma
+; RV64-BITS-UNKNOWN-NEXT:    vid.v v12
 ; RV64-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV64-BITS-UNKNOWN-NEXT:    slli a0, a0, 1
 ; RV64-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, m4, ta, ma
-; RV64-BITS-UNKNOWN-NEXT:    vid.v v12
 ; RV64-BITS-UNKNOWN-NEXT:    vrsub.vx v12, v12, a0
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m2, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v10, v8, v12
 ; RV64-BITS-UNKNOWN-NEXT:    vand.vi v8, v10, 1
 ; RV64-BITS-UNKNOWN-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV64-BITS-256-LABEL: reverse_nxv16i1:
 ; RV64-BITS-256:       # %bb.0:
 ; RV64-BITS-256-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
 ; RV64-BITS-256-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-256-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-256-NEXT:    vid.v v10
 ; RV64-BITS-256-NEXT:    csrr a0, vlenb
 ; RV64-BITS-256-NEXT:    slli a0, a0, 1
 ; RV64-BITS-256-NEXT:    addi a0, a0, -1
-; RV64-BITS-256-NEXT:    vid.v v10
 ; RV64-BITS-256-NEXT:    vrsub.vx v10, v10, a0
 ; RV64-BITS-256-NEXT:    vrgather.vv v12, v8, v10
 ; RV64-BITS-256-NEXT:    vand.vi v8, v12, 1
 ; RV64-BITS-256-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-256-NEXT:    ret
 ;
 ; RV64-BITS-512-LABEL: reverse_nxv16i1:
 ; RV64-BITS-512:       # %bb.0:
 ; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
 ; RV64-BITS-512-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-512-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-512-NEXT:    vid.v v10
 ; RV64-BITS-512-NEXT:    csrr a0, vlenb
 ; RV64-BITS-512-NEXT:    slli a0, a0, 1
 ; RV64-BITS-512-NEXT:    addi a0, a0, -1
-; RV64-BITS-512-NEXT:    vid.v v10
 ; RV64-BITS-512-NEXT:    vrsub.vx v10, v10, a0
 ; RV64-BITS-512-NEXT:    vrgather.vv v12, v8, v10
 ; RV64-BITS-512-NEXT:    vand.vi v8, v12, 1
 ; RV64-BITS-512-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-512-NEXT:    ret
   %res = call <vscale x 16 x i1> @llvm.experimental.vector.reverse.nxv16i1(<vscale x 16 x i1> %a)
   ret <vscale x 16 x i1> %res
 }
 
 define <vscale x 32 x i1> @reverse_nxv32i1(<vscale x 32 x i1> %a) {
 ; RV32-BITS-UNKNOWN-LABEL: reverse_nxv32i1:
 ; RV32-BITS-UNKNOWN:       # %bb.0:
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-UNKNOWN-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, m8, ta, ma
+; RV32-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV32-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV32-BITS-UNKNOWN-NEXT:    slli a0, a0, 2
 ; RV32-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, m8, ta, ma
-; RV32-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV32-BITS-UNKNOWN-NEXT:    vrsub.vx v16, v16, a0
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m4, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v12, v8, v16
 ; RV32-BITS-UNKNOWN-NEXT:    vand.vi v8, v12, 1
 ; RV32-BITS-UNKNOWN-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV32-BITS-256-LABEL: reverse_nxv32i1:
 ; RV32-BITS-256:       # %bb.0:
 ; RV32-BITS-256-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
 ; RV32-BITS-256-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-256-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-256-NEXT:    vid.v v12
 ; RV32-BITS-256-NEXT:    csrr a0, vlenb
 ; RV32-BITS-256-NEXT:    slli a0, a0, 2
 ; RV32-BITS-256-NEXT:    addi a0, a0, -1
-; RV32-BITS-256-NEXT:    vid.v v12
 ; RV32-BITS-256-NEXT:    vrsub.vx v12, v12, a0
 ; RV32-BITS-256-NEXT:    vrgather.vv v16, v8, v12
 ; RV32-BITS-256-NEXT:    vand.vi v8, v16, 1
 ; RV32-BITS-256-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-256-NEXT:    ret
 ;
 ; RV32-BITS-512-LABEL: reverse_nxv32i1:
 ; RV32-BITS-512:       # %bb.0:
 ; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
 ; RV32-BITS-512-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-512-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-512-NEXT:    vid.v v12
 ; RV32-BITS-512-NEXT:    csrr a0, vlenb
 ; RV32-BITS-512-NEXT:    slli a0, a0, 2
 ; RV32-BITS-512-NEXT:    addi a0, a0, -1
-; RV32-BITS-512-NEXT:    vid.v v12
 ; RV32-BITS-512-NEXT:    vrsub.vx v12, v12, a0
 ; RV32-BITS-512-NEXT:    vrgather.vv v16, v8, v12
 ; RV32-BITS-512-NEXT:    vand.vi v8, v16, 1
 ; RV32-BITS-512-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-512-NEXT:    ret
 ;
 ; RV64-BITS-UNKNOWN-LABEL: reverse_nxv32i1:
 ; RV64-BITS-UNKNOWN:       # %bb.0:
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-UNKNOWN-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, m8, ta, ma
+; RV64-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV64-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV64-BITS-UNKNOWN-NEXT:    slli a0, a0, 2
 ; RV64-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e16, m8, ta, ma
-; RV64-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV64-BITS-UNKNOWN-NEXT:    vrsub.vx v16, v16, a0
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m4, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v12, v8, v16
 ; RV64-BITS-UNKNOWN-NEXT:    vand.vi v8, v12, 1
 ; RV64-BITS-UNKNOWN-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV64-BITS-256-LABEL: reverse_nxv32i1:
 ; RV64-BITS-256:       # %bb.0:
 ; RV64-BITS-256-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
 ; RV64-BITS-256-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-256-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-256-NEXT:    vid.v v12
 ; RV64-BITS-256-NEXT:    csrr a0, vlenb
 ; RV64-BITS-256-NEXT:    slli a0, a0, 2
 ; RV64-BITS-256-NEXT:    addi a0, a0, -1
-; RV64-BITS-256-NEXT:    vid.v v12
 ; RV64-BITS-256-NEXT:    vrsub.vx v12, v12, a0
 ; RV64-BITS-256-NEXT:    vrgather.vv v16, v8, v12
 ; RV64-BITS-256-NEXT:    vand.vi v8, v16, 1
 ; RV64-BITS-256-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-256-NEXT:    ret
 ;
 ; RV64-BITS-512-LABEL: reverse_nxv32i1:
 ; RV64-BITS-512:       # %bb.0:
 ; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
 ; RV64-BITS-512-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-512-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-512-NEXT:    vid.v v12
 ; RV64-BITS-512-NEXT:    csrr a0, vlenb
 ; RV64-BITS-512-NEXT:    slli a0, a0, 2
 ; RV64-BITS-512-NEXT:    addi a0, a0, -1
-; RV64-BITS-512-NEXT:    vid.v v12
 ; RV64-BITS-512-NEXT:    vrsub.vx v12, v12, a0
 ; RV64-BITS-512-NEXT:    vrgather.vv v16, v8, v12
 ; RV64-BITS-512-NEXT:    vand.vi v8, v16, 1
 ; RV64-BITS-512-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-512-NEXT:    ret
   %res = call <vscale x 32 x i1> @llvm.experimental.vector.reverse.nxv32i1(<vscale x 32 x i1> %a)
   ret <vscale x 32 x i1> %res
 }
 
 define <vscale x 64 x i1> @reverse_nxv64i1(<vscale x 64 x i1> %a) {
 ; RV32-BITS-UNKNOWN-LABEL: reverse_nxv64i1:
 ; RV32-BITS-UNKNOWN:       # %bb.0:
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-UNKNOWN-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, m8, ta, ma
+; RV32-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV32-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV32-BITS-UNKNOWN-NEXT:    slli a0, a0, 2
 ; RV32-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV32-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
-; RV32-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV32-BITS-UNKNOWN-NEXT:    vrsub.vx v16, v16, a0
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m4, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v28, v8, v16
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v24, v12, v16
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vand.vi v8, v24, 1
 ; RV32-BITS-UNKNOWN-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV32-BITS-256-LABEL: reverse_nxv64i1:
 ; RV32-BITS-256:       # %bb.0:
 ; RV32-BITS-256-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; RV32-BITS-256-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-256-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-256-NEXT:    vid.v v16
 ; RV32-BITS-256-NEXT:    csrr a0, vlenb
 ; RV32-BITS-256-NEXT:    slli a0, a0, 3
 ; RV32-BITS-256-NEXT:    addi a0, a0, -1
-; RV32-BITS-256-NEXT:    vid.v v16
 ; RV32-BITS-256-NEXT:    vrsub.vx v16, v16, a0
 ; RV32-BITS-256-NEXT:    vrgather.vv v24, v8, v16
 ; RV32-BITS-256-NEXT:    vand.vi v8, v24, 1
 ; RV32-BITS-256-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-256-NEXT:    ret
 ;
 ; RV32-BITS-512-LABEL: reverse_nxv64i1:
 ; RV32-BITS-512:       # %bb.0:
 ; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; RV32-BITS-512-NEXT:    vmv.v.i v8, 0
 ; RV32-BITS-512-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
+; RV32-BITS-512-NEXT:    vid.v v16
 ; RV32-BITS-512-NEXT:    csrr a0, vlenb
 ; RV32-BITS-512-NEXT:    slli a0, a0, 2
 ; RV32-BITS-512-NEXT:    addi a0, a0, -1
-; RV32-BITS-512-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
-; RV32-BITS-512-NEXT:    vid.v v16
 ; RV32-BITS-512-NEXT:    vrsub.vx v16, v16, a0
 ; RV32-BITS-512-NEXT:    vrgather.vv v28, v8, v16
 ; RV32-BITS-512-NEXT:    vrgather.vv v24, v12, v16
 ; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; RV32-BITS-512-NEXT:    vand.vi v8, v24, 1
 ; RV32-BITS-512-NEXT:    vmsne.vi v0, v8, 0
 ; RV32-BITS-512-NEXT:    ret
 ;
 ; RV64-BITS-UNKNOWN-LABEL: reverse_nxv64i1:
 ; RV64-BITS-UNKNOWN:       # %bb.0:
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-UNKNOWN-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, m8, ta, ma
+; RV64-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV64-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV64-BITS-UNKNOWN-NEXT:    slli a0, a0, 2
 ; RV64-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV64-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
-; RV64-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV64-BITS-UNKNOWN-NEXT:    vrsub.vx v16, v16, a0
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m4, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v28, v8, v16
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v24, v12, v16
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vand.vi v8, v24, 1
 ; RV64-BITS-UNKNOWN-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV64-BITS-256-LABEL: reverse_nxv64i1:
 ; RV64-BITS-256:       # %bb.0:
 ; RV64-BITS-256-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; RV64-BITS-256-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-256-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-256-NEXT:    vid.v v16
 ; RV64-BITS-256-NEXT:    csrr a0, vlenb
 ; RV64-BITS-256-NEXT:    slli a0, a0, 3
 ; RV64-BITS-256-NEXT:    addi a0, a0, -1
-; RV64-BITS-256-NEXT:    vid.v v16
 ; RV64-BITS-256-NEXT:    vrsub.vx v16, v16, a0
 ; RV64-BITS-256-NEXT:    vrgather.vv v24, v8, v16
 ; RV64-BITS-256-NEXT:    vand.vi v8, v24, 1
 ; RV64-BITS-256-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-256-NEXT:    ret
 ;
 ; RV64-BITS-512-LABEL: reverse_nxv64i1:
 ; RV64-BITS-512:       # %bb.0:
 ; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; RV64-BITS-512-NEXT:    vmv.v.i v8, 0
 ; RV64-BITS-512-NEXT:    vmerge.vim v8, v8, 1, v0
+; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
+; RV64-BITS-512-NEXT:    vid.v v16
 ; RV64-BITS-512-NEXT:    csrr a0, vlenb
 ; RV64-BITS-512-NEXT:    slli a0, a0, 2
 ; RV64-BITS-512-NEXT:    addi a0, a0, -1
-; RV64-BITS-512-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
-; RV64-BITS-512-NEXT:    vid.v v16
 ; RV64-BITS-512-NEXT:    vrsub.vx v16, v16, a0
 ; RV64-BITS-512-NEXT:    vrgather.vv v28, v8, v16
 ; RV64-BITS-512-NEXT:    vrgather.vv v24, v12, v16
 ; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
 ; RV64-BITS-512-NEXT:    vand.vi v8, v24, 1
 ; RV64-BITS-512-NEXT:    vmsne.vi v0, v8, 0
 ; RV64-BITS-512-NEXT:    ret
   %res = call <vscale x 64 x i1> @llvm.experimental.vector.reverse.nxv64i1(<vscale x 64 x i1> %a)
   ret <vscale x 64 x i1> %res
 }
 
 ;
 ; VECTOR_REVERSE - integer
 ;
 
 define <vscale x 1 x i8> @reverse_nxv1i8(<vscale x 1 x i8> %a) {
 ; RV32-BITS-UNKNOWN-LABEL: reverse_nxv1i8:
 ; RV32-BITS-UNKNOWN:       # %bb.0:
+; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, mf4, ta, ma
+; RV32-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV32-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV32-BITS-UNKNOWN-NEXT:    srli a0, a0, 3
 ; RV32-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV32-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, mf4, ta, ma
-; RV32-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV32-BITS-UNKNOWN-NEXT:    vrsub.vx v10, v9, a0
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, mf8, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v9, v8, v10
 ; RV32-BITS-UNKNOWN-NEXT:    vmv1r.v v8, v9
 ; RV32-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV32-BITS-256-LABEL: reverse_nxv1i8:
 ; RV32-BITS-256:       # %bb.0:
+; RV32-BITS-256-NEXT:    vsetvli a0, zero, e8, mf8, ta, ma
+; RV32-BITS-256-NEXT:    vid.v v9
 ; RV32-BITS-256-NEXT:    csrr a0, vlenb
 ; RV32-BITS-256-NEXT:    srli a0, a0, 3
 ; RV32-BITS-256-NEXT:    addi a0, a0, -1
-; RV32-BITS-256-NEXT:    vsetvli a1, zero, e8, mf8, ta, ma
-; RV32-BITS-256-NEXT:    vid.v v9
 ; RV32-BITS-256-NEXT:    vrsub.vx v10, v9, a0
 ; RV32-BITS-256-NEXT:    vrgather.vv v9, v8, v10
 ; RV32-BITS-256-NEXT:    vmv1r.v v8, v9
 ; RV32-BITS-256-NEXT:    ret
 ;
 ; RV32-BITS-512-LABEL: reverse_nxv1i8:
 ; RV32-BITS-512:       # %bb.0:
+; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, mf8, ta, ma
+; RV32-BITS-512-NEXT:    vid.v v9
 ; RV32-BITS-512-NEXT:    csrr a0, vlenb
 ; RV32-BITS-512-NEXT:    srli a0, a0, 3
 ; RV32-BITS-512-NEXT:    addi a0, a0, -1
-; RV32-BITS-512-NEXT:    vsetvli a1, zero, e8, mf8, ta, ma
-; RV32-BITS-512-NEXT:    vid.v v9
 ; RV32-BITS-512-NEXT:    vrsub.vx v10, v9, a0
 ; RV32-BITS-512-NEXT:    vrgather.vv v9, v8, v10
 ; RV32-BITS-512-NEXT:    vmv1r.v v8, v9
 ; RV32-BITS-512-NEXT:    ret
 ;
 ; RV64-BITS-UNKNOWN-LABEL: reverse_nxv1i8:
 ; RV64-BITS-UNKNOWN:       # %bb.0:
+; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, mf4, ta, ma
+; RV64-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV64-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV64-BITS-UNKNOWN-NEXT:    srli a0, a0, 3
 ; RV64-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV64-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, mf4, ta, ma
-; RV64-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV64-BITS-UNKNOWN-NEXT:    vrsub.vx v10, v9, a0
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, mf8, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v9, v8, v10
 ; RV64-BITS-UNKNOWN-NEXT:    vmv1r.v v8, v9
 ; RV64-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV64-BITS-256-LABEL: reverse_nxv1i8:
 ; RV64-BITS-256:       # %bb.0:
+; RV64-BITS-256-NEXT:    vsetvli a0, zero, e8, mf8, ta, ma
+; RV64-BITS-256-NEXT:    vid.v v9
 ; RV64-BITS-256-NEXT:    csrr a0, vlenb
 ; RV64-BITS-256-NEXT:    srli a0, a0, 3
 ; RV64-BITS-256-NEXT:    addi a0, a0, -1
-; RV64-BITS-256-NEXT:    vsetvli a1, zero, e8, mf8, ta, ma
-; RV64-BITS-256-NEXT:    vid.v v9
 ; RV64-BITS-256-NEXT:    vrsub.vx v10, v9, a0
 ; RV64-BITS-256-NEXT:    vrgather.vv v9, v8, v10
 ; RV64-BITS-256-NEXT:    vmv1r.v v8, v9
 ; RV64-BITS-256-NEXT:    ret
 ;
 ; RV64-BITS-512-LABEL: reverse_nxv1i8:
 ; RV64-BITS-512:       # %bb.0:
+; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, mf8, ta, ma
+; RV64-BITS-512-NEXT:    vid.v v9
 ; RV64-BITS-512-NEXT:    csrr a0, vlenb
 ; RV64-BITS-512-NEXT:    srli a0, a0, 3
 ; RV64-BITS-512-NEXT:    addi a0, a0, -1
-; RV64-BITS-512-NEXT:    vsetvli a1, zero, e8, mf8, ta, ma
-; RV64-BITS-512-NEXT:    vid.v v9
 ; RV64-BITS-512-NEXT:    vrsub.vx v10, v9, a0
 ; RV64-BITS-512-NEXT:    vrgather.vv v9, v8, v10
 ; RV64-BITS-512-NEXT:    vmv1r.v v8, v9
 ; RV64-BITS-512-NEXT:    ret
   %res = call <vscale x 1 x i8> @llvm.experimental.vector.reverse.nxv1i8(<vscale x 1 x i8> %a)
   ret <vscale x 1 x i8> %res
 }
 
 define <vscale x 2 x i8> @reverse_nxv2i8(<vscale x 2 x i8> %a) {
 ; RV32-BITS-UNKNOWN-LABEL: reverse_nxv2i8:
 ; RV32-BITS-UNKNOWN:       # %bb.0:
+; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, mf2, ta, ma
+; RV32-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV32-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV32-BITS-UNKNOWN-NEXT:    srli a0, a0, 2
 ; RV32-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV32-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, mf2, ta, ma
-; RV32-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV32-BITS-UNKNOWN-NEXT:    vrsub.vx v10, v9, a0
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, mf4, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v9, v8, v10
 ; RV32-BITS-UNKNOWN-NEXT:    vmv1r.v v8, v9
 ; RV32-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV32-BITS-256-LABEL: reverse_nxv2i8:
 ; RV32-BITS-256:       # %bb.0:
+; RV32-BITS-256-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
+; RV32-BITS-256-NEXT:    vid.v v9
 ; RV32-BITS-256-NEXT:    csrr a0, vlenb
 ; RV32-BITS-256-NEXT:    srli a0, a0, 2
 ; RV32-BITS-256-NEXT:    addi a0, a0, -1
-; RV32-BITS-256-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
-; RV32-BITS-256-NEXT:    vid.v v9
 ; RV32-BITS-256-NEXT:    vrsub.vx v10, v9, a0
 ; RV32-BITS-256-NEXT:    vrgather.vv v9, v8, v10
 ; RV32-BITS-256-NEXT:    vmv1r.v v8, v9
 ; RV32-BITS-256-NEXT:    ret
 ;
 ; RV32-BITS-512-LABEL: reverse_nxv2i8:
 ; RV32-BITS-512:       # %bb.0:
+; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
+; RV32-BITS-512-NEXT:    vid.v v9
 ; RV32-BITS-512-NEXT:    csrr a0, vlenb
 ; RV32-BITS-512-NEXT:    srli a0, a0, 2
 ; RV32-BITS-512-NEXT:    addi a0, a0, -1
-; RV32-BITS-512-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
-; RV32-BITS-512-NEXT:    vid.v v9
 ; RV32-BITS-512-NEXT:    vrsub.vx v10, v9, a0
 ; RV32-BITS-512-NEXT:    vrgather.vv v9, v8, v10
 ; RV32-BITS-512-NEXT:    vmv1r.v v8, v9
 ; RV32-BITS-512-NEXT:    ret
 ;
 ; RV64-BITS-UNKNOWN-LABEL: reverse_nxv2i8:
 ; RV64-BITS-UNKNOWN:       # %bb.0:
+; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, mf2, ta, ma
+; RV64-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV64-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV64-BITS-UNKNOWN-NEXT:    srli a0, a0, 2
 ; RV64-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV64-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, mf2, ta, ma
-; RV64-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV64-BITS-UNKNOWN-NEXT:    vrsub.vx v10, v9, a0
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, mf4, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v9, v8, v10
 ; RV64-BITS-UNKNOWN-NEXT:    vmv1r.v v8, v9
 ; RV64-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV64-BITS-256-LABEL: reverse_nxv2i8:
 ; RV64-BITS-256:       # %bb.0:
+; RV64-BITS-256-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
+; RV64-BITS-256-NEXT:    vid.v v9
 ; RV64-BITS-256-NEXT:    csrr a0, vlenb
 ; RV64-BITS-256-NEXT:    srli a0, a0, 2
 ; RV64-BITS-256-NEXT:    addi a0, a0, -1
-; RV64-BITS-256-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
-; RV64-BITS-256-NEXT:    vid.v v9
 ; RV64-BITS-256-NEXT:    vrsub.vx v10, v9, a0
 ; RV64-BITS-256-NEXT:    vrgather.vv v9, v8, v10
 ; RV64-BITS-256-NEXT:    vmv1r.v v8, v9
 ; RV64-BITS-256-NEXT:    ret
 ;
 ; RV64-BITS-512-LABEL: reverse_nxv2i8:
 ; RV64-BITS-512:       # %bb.0:
+; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, mf4, ta, ma
+; RV64-BITS-512-NEXT:    vid.v v9
 ; RV64-BITS-512-NEXT:    csrr a0, vlenb
 ; RV64-BITS-512-NEXT:    srli a0, a0, 2
 ; RV64-BITS-512-NEXT:    addi a0, a0, -1
-; RV64-BITS-512-NEXT:    vsetvli a1, zero, e8, mf4, ta, ma
-; RV64-BITS-512-NEXT:    vid.v v9
 ; RV64-BITS-512-NEXT:    vrsub.vx v10, v9, a0
 ; RV64-BITS-512-NEXT:    vrgather.vv v9, v8, v10
 ; RV64-BITS-512-NEXT:    vmv1r.v v8, v9
 ; RV64-BITS-512-NEXT:    ret
   %res = call <vscale x 2 x i8> @llvm.experimental.vector.reverse.nxv2i8(<vscale x 2 x i8> %a)
   ret <vscale x 2 x i8> %res
 }
 
 define <vscale x 4 x i8> @reverse_nxv4i8(<vscale x 4 x i8> %a) {
 ; RV32-BITS-UNKNOWN-LABEL: reverse_nxv4i8:
 ; RV32-BITS-UNKNOWN:       # %bb.0:
+; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, m1, ta, ma
+; RV32-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV32-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV32-BITS-UNKNOWN-NEXT:    srli a0, a0, 1
 ; RV32-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV32-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
-; RV32-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV32-BITS-UNKNOWN-NEXT:    vrsub.vx v10, v9, a0
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v9, v8, v10
 ; RV32-BITS-UNKNOWN-NEXT:    vmv1r.v v8, v9
 ; RV32-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV32-BITS-256-LABEL: reverse_nxv4i8:
 ; RV32-BITS-256:       # %bb.0:
+; RV32-BITS-256-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
+; RV32-BITS-256-NEXT:    vid.v v9
 ; RV32-BITS-256-NEXT:    csrr a0, vlenb
 ; RV32-BITS-256-NEXT:    srli a0, a0, 1
 ; RV32-BITS-256-NEXT:    addi a0, a0, -1
-; RV32-BITS-256-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
-; RV32-BITS-256-NEXT:    vid.v v9
 ; RV32-BITS-256-NEXT:    vrsub.vx v10, v9, a0
 ; RV32-BITS-256-NEXT:    vrgather.vv v9, v8, v10
 ; RV32-BITS-256-NEXT:    vmv1r.v v8, v9
 ; RV32-BITS-256-NEXT:    ret
 ;
 ; RV32-BITS-512-LABEL: reverse_nxv4i8:
 ; RV32-BITS-512:       # %bb.0:
+; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
+; RV32-BITS-512-NEXT:    vid.v v9
 ; RV32-BITS-512-NEXT:    csrr a0, vlenb
 ; RV32-BITS-512-NEXT:    srli a0, a0, 1
 ; RV32-BITS-512-NEXT:    addi a0, a0, -1
-; RV32-BITS-512-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
-; RV32-BITS-512-NEXT:    vid.v v9
 ; RV32-BITS-512-NEXT:    vrsub.vx v10, v9, a0
 ; RV32-BITS-512-NEXT:    vrgather.vv v9, v8, v10
 ; RV32-BITS-512-NEXT:    vmv1r.v v8, v9
 ; RV32-BITS-512-NEXT:    ret
 ;
 ; RV64-BITS-UNKNOWN-LABEL: reverse_nxv4i8:
 ; RV64-BITS-UNKNOWN:       # %bb.0:
+; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, m1, ta, ma
+; RV64-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV64-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV64-BITS-UNKNOWN-NEXT:    srli a0, a0, 1
 ; RV64-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV64-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
-; RV64-BITS-UNKNOWN-NEXT:    vid.v v9
 ; RV64-BITS-UNKNOWN-NEXT:    vrsub.vx v10, v9, a0
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, mf2, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v9, v8, v10
 ; RV64-BITS-UNKNOWN-NEXT:    vmv1r.v v8, v9
 ; RV64-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV64-BITS-256-LABEL: reverse_nxv4i8:
 ; RV64-BITS-256:       # %bb.0:
+; RV64-BITS-256-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
+; RV64-BITS-256-NEXT:    vid.v v9
 ; RV64-BITS-256-NEXT:    csrr a0, vlenb
 ; RV64-BITS-256-NEXT:    srli a0, a0, 1
 ; RV64-BITS-256-NEXT:    addi a0, a0, -1
-; RV64-BITS-256-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
-; RV64-BITS-256-NEXT:    vid.v v9
 ; RV64-BITS-256-NEXT:    vrsub.vx v10, v9, a0
 ; RV64-BITS-256-NEXT:    vrgather.vv v9, v8, v10
 ; RV64-BITS-256-NEXT:    vmv1r.v v8, v9
 ; RV64-BITS-256-NEXT:    ret
 ;
 ; RV64-BITS-512-LABEL: reverse_nxv4i8:
 ; RV64-BITS-512:       # %bb.0:
+; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, mf2, ta, ma
+; RV64-BITS-512-NEXT:    vid.v v9
 ; RV64-BITS-512-NEXT:    csrr a0, vlenb
 ; RV64-BITS-512-NEXT:    srli a0, a0, 1
 ; RV64-BITS-512-NEXT:    addi a0, a0, -1
-; RV64-BITS-512-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
-; RV64-BITS-512-NEXT:    vid.v v9
 ; RV64-BITS-512-NEXT:    vrsub.vx v10, v9, a0
 ; RV64-BITS-512-NEXT:    vrgather.vv v9, v8, v10
 ; RV64-BITS-512-NEXT:    vmv1r.v v8, v9
 ; RV64-BITS-512-NEXT:    ret
   %res = call <vscale x 4 x i8> @llvm.experimental.vector.reverse.nxv4i8(<vscale x 4 x i8> %a)
   ret <vscale x 4 x i8> %res
 }
 

 ; RV64-BITS-512-NEXT:    ret
   %res = call <vscale x 8 x i8> @llvm.experimental.vector.reverse.nxv8i8(<vscale x 8 x i8> %a)
   ret <vscale x 8 x i8> %res
 }
 
 define <vscale x 16 x i8> @reverse_nxv16i8(<vscale x 16 x i8> %a) {
 ; RV32-BITS-UNKNOWN-LABEL: reverse_nxv16i8:
 ; RV32-BITS-UNKNOWN:       # %bb.0:
+; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; RV32-BITS-UNKNOWN-NEXT:    vid.v v12
 ; RV32-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV32-BITS-UNKNOWN-NEXT:    slli a0, a0, 1
 ; RV32-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV32-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
-; RV32-BITS-UNKNOWN-NEXT:    vid.v v12
 ; RV32-BITS-UNKNOWN-NEXT:    vrsub.vx v12, v12, a0
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m2, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v10, v8, v12
 ; RV32-BITS-UNKNOWN-NEXT:    vmv.v.v v8, v10
 ; RV32-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV32-BITS-256-LABEL: reverse_nxv16i8:
 ; RV32-BITS-256:       # %bb.0:
+; RV32-BITS-256-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
+; RV32-BITS-256-NEXT:    vid.v v10
 ; RV32-BITS-256-NEXT:    csrr a0, vlenb
 ; RV32-BITS-256-NEXT:    slli a0, a0, 1
 ; RV32-BITS-256-NEXT:    addi a0, a0, -1
-; RV32-BITS-256-NEXT:    vsetvli a1, zero, e8, m2, ta, ma
-; RV32-BITS-256-NEXT:    vid.v v10
 ; RV32-BITS-256-NEXT:    vrsub.vx v12, v10, a0
 ; RV32-BITS-256-NEXT:    vrgather.vv v10, v8, v12
 ; RV32-BITS-256-NEXT:    vmv.v.v v8, v10
 ; RV32-BITS-256-NEXT:    ret
 ;
 ; RV32-BITS-512-LABEL: reverse_nxv16i8:
 ; RV32-BITS-512:       # %bb.0:
+; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
+; RV32-BITS-512-NEXT:    vid.v v10
 ; RV32-BITS-512-NEXT:    csrr a0, vlenb
 ; RV32-BITS-512-NEXT:    slli a0, a0, 1
 ; RV32-BITS-512-NEXT:    addi a0, a0, -1
-; RV32-BITS-512-NEXT:    vsetvli a1, zero, e8, m2, ta, ma
-; RV32-BITS-512-NEXT:    vid.v v10
 ; RV32-BITS-512-NEXT:    vrsub.vx v12, v10, a0
 ; RV32-BITS-512-NEXT:    vrgather.vv v10, v8, v12
 ; RV32-BITS-512-NEXT:    vmv.v.v v8, v10
 ; RV32-BITS-512-NEXT:    ret
 ;
 ; RV64-BITS-UNKNOWN-LABEL: reverse_nxv16i8:
 ; RV64-BITS-UNKNOWN:       # %bb.0:
+; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; RV64-BITS-UNKNOWN-NEXT:    vid.v v12
 ; RV64-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV64-BITS-UNKNOWN-NEXT:    slli a0, a0, 1
 ; RV64-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV64-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
-; RV64-BITS-UNKNOWN-NEXT:    vid.v v12
 ; RV64-BITS-UNKNOWN-NEXT:    vrsub.vx v12, v12, a0
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m2, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v10, v8, v12
 ; RV64-BITS-UNKNOWN-NEXT:    vmv.v.v v8, v10
 ; RV64-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV64-BITS-256-LABEL: reverse_nxv16i8:
 ; RV64-BITS-256:       # %bb.0:
+; RV64-BITS-256-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
+; RV64-BITS-256-NEXT:    vid.v v10
 ; RV64-BITS-256-NEXT:    csrr a0, vlenb
 ; RV64-BITS-256-NEXT:    slli a0, a0, 1
 ; RV64-BITS-256-NEXT:    addi a0, a0, -1
-; RV64-BITS-256-NEXT:    vsetvli a1, zero, e8, m2, ta, ma
-; RV64-BITS-256-NEXT:    vid.v v10
 ; RV64-BITS-256-NEXT:    vrsub.vx v12, v10, a0
 ; RV64-BITS-256-NEXT:    vrgather.vv v10, v8, v12
 ; RV64-BITS-256-NEXT:    vmv.v.v v8, v10
 ; RV64-BITS-256-NEXT:    ret
 ;
 ; RV64-BITS-512-LABEL: reverse_nxv16i8:
 ; RV64-BITS-512:       # %bb.0:
+; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, m2, ta, ma
+; RV64-BITS-512-NEXT:    vid.v v10
 ; RV64-BITS-512-NEXT:    csrr a0, vlenb
 ; RV64-BITS-512-NEXT:    slli a0, a0, 1
 ; RV64-BITS-512-NEXT:    addi a0, a0, -1
-; RV64-BITS-512-NEXT:    vsetvli a1, zero, e8, m2, ta, ma
-; RV64-BITS-512-NEXT:    vid.v v10
 ; RV64-BITS-512-NEXT:    vrsub.vx v12, v10, a0
 ; RV64-BITS-512-NEXT:    vrgather.vv v10, v8, v12
 ; RV64-BITS-512-NEXT:    vmv.v.v v8, v10
 ; RV64-BITS-512-NEXT:    ret
   %res = call <vscale x 16 x i8> @llvm.experimental.vector.reverse.nxv16i8(<vscale x 16 x i8> %a)
   ret <vscale x 16 x i8> %res
 }
 
 define <vscale x 32 x i8> @reverse_nxv32i8(<vscale x 32 x i8> %a) {
 ; RV32-BITS-UNKNOWN-LABEL: reverse_nxv32i8:
 ; RV32-BITS-UNKNOWN:       # %bb.0:
+; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, m8, ta, ma
+; RV32-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV32-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV32-BITS-UNKNOWN-NEXT:    slli a0, a0, 2
 ; RV32-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV32-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
-; RV32-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV32-BITS-UNKNOWN-NEXT:    vrsub.vx v16, v16, a0
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m4, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v12, v8, v16
 ; RV32-BITS-UNKNOWN-NEXT:    vmv.v.v v8, v12
 ; RV32-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV32-BITS-256-LABEL: reverse_nxv32i8:
 ; RV32-BITS-256:       # %bb.0:
+; RV32-BITS-256-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
+; RV32-BITS-256-NEXT:    vid.v v12
 ; RV32-BITS-256-NEXT:    csrr a0, vlenb
 ; RV32-BITS-256-NEXT:    slli a0, a0, 2
 ; RV32-BITS-256-NEXT:    addi a0, a0, -1
-; RV32-BITS-256-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
-; RV32-BITS-256-NEXT:    vid.v v12
 ; RV32-BITS-256-NEXT:    vrsub.vx v16, v12, a0
 ; RV32-BITS-256-NEXT:    vrgather.vv v12, v8, v16
 ; RV32-BITS-256-NEXT:    vmv.v.v v8, v12
 ; RV32-BITS-256-NEXT:    ret
 ;
 ; RV32-BITS-512-LABEL: reverse_nxv32i8:
 ; RV32-BITS-512:       # %bb.0:
+; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
+; RV32-BITS-512-NEXT:    vid.v v12
 ; RV32-BITS-512-NEXT:    csrr a0, vlenb
 ; RV32-BITS-512-NEXT:    slli a0, a0, 2
 ; RV32-BITS-512-NEXT:    addi a0, a0, -1
-; RV32-BITS-512-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
-; RV32-BITS-512-NEXT:    vid.v v12
 ; RV32-BITS-512-NEXT:    vrsub.vx v16, v12, a0
 ; RV32-BITS-512-NEXT:    vrgather.vv v12, v8, v16
 ; RV32-BITS-512-NEXT:    vmv.v.v v8, v12
 ; RV32-BITS-512-NEXT:    ret
 ;
 ; RV64-BITS-UNKNOWN-LABEL: reverse_nxv32i8:
 ; RV64-BITS-UNKNOWN:       # %bb.0:
+; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, m8, ta, ma
+; RV64-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV64-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV64-BITS-UNKNOWN-NEXT:    slli a0, a0, 2
 ; RV64-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV64-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
-; RV64-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV64-BITS-UNKNOWN-NEXT:    vrsub.vx v16, v16, a0
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m4, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v12, v8, v16
 ; RV64-BITS-UNKNOWN-NEXT:    vmv.v.v v8, v12
 ; RV64-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV64-BITS-256-LABEL: reverse_nxv32i8:
 ; RV64-BITS-256:       # %bb.0:
+; RV64-BITS-256-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
+; RV64-BITS-256-NEXT:    vid.v v12
 ; RV64-BITS-256-NEXT:    csrr a0, vlenb
 ; RV64-BITS-256-NEXT:    slli a0, a0, 2
 ; RV64-BITS-256-NEXT:    addi a0, a0, -1
-; RV64-BITS-256-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
-; RV64-BITS-256-NEXT:    vid.v v12
 ; RV64-BITS-256-NEXT:    vrsub.vx v16, v12, a0
 ; RV64-BITS-256-NEXT:    vrgather.vv v12, v8, v16
 ; RV64-BITS-256-NEXT:    vmv.v.v v8, v12
 ; RV64-BITS-256-NEXT:    ret
 ;
 ; RV64-BITS-512-LABEL: reverse_nxv32i8:
 ; RV64-BITS-512:       # %bb.0:
+; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
+; RV64-BITS-512-NEXT:    vid.v v12
 ; RV64-BITS-512-NEXT:    csrr a0, vlenb
 ; RV64-BITS-512-NEXT:    slli a0, a0, 2
 ; RV64-BITS-512-NEXT:    addi a0, a0, -1
-; RV64-BITS-512-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
-; RV64-BITS-512-NEXT:    vid.v v12
 ; RV64-BITS-512-NEXT:    vrsub.vx v16, v12, a0
 ; RV64-BITS-512-NEXT:    vrgather.vv v12, v8, v16
 ; RV64-BITS-512-NEXT:    vmv.v.v v8, v12
 ; RV64-BITS-512-NEXT:    ret
   %res = call <vscale x 32 x i8> @llvm.experimental.vector.reverse.nxv32i8(<vscale x 32 x i8> %a)
   ret <vscale x 32 x i8> %res
 }
 
 define <vscale x 64 x i8> @reverse_nxv64i8(<vscale x 64 x i8> %a) {
 ; RV32-BITS-UNKNOWN-LABEL: reverse_nxv64i8:
 ; RV32-BITS-UNKNOWN:       # %bb.0:
+; RV32-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, m8, ta, ma
+; RV32-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV32-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV32-BITS-UNKNOWN-NEXT:    slli a0, a0, 2
 ; RV32-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV32-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
-; RV32-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV32-BITS-UNKNOWN-NEXT:    vrsub.vx v24, v16, a0
 ; RV32-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m4, ta, ma
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v20, v8, v24
 ; RV32-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v16, v12, v24
 ; RV32-BITS-UNKNOWN-NEXT:    vmv8r.v v8, v16
 ; RV32-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV32-BITS-256-LABEL: reverse_nxv64i8:
 ; RV32-BITS-256:       # %bb.0:
+; RV32-BITS-256-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
+; RV32-BITS-256-NEXT:    vid.v v16
 ; RV32-BITS-256-NEXT:    csrr a0, vlenb
 ; RV32-BITS-256-NEXT:    slli a0, a0, 3
 ; RV32-BITS-256-NEXT:    addi a0, a0, -1
-; RV32-BITS-256-NEXT:    vsetvli a1, zero, e8, m8, ta, ma
-; RV32-BITS-256-NEXT:    vid.v v16
 ; RV32-BITS-256-NEXT:    vrsub.vx v24, v16, a0
 ; RV32-BITS-256-NEXT:    vrgather.vv v16, v8, v24
 ; RV32-BITS-256-NEXT:    vmv.v.v v8, v16
 ; RV32-BITS-256-NEXT:    ret
 ;
 ; RV32-BITS-512-LABEL: reverse_nxv64i8:
 ; RV32-BITS-512:       # %bb.0:
+; RV32-BITS-512-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
+; RV32-BITS-512-NEXT:    vid.v v16
 ; RV32-BITS-512-NEXT:    csrr a0, vlenb
 ; RV32-BITS-512-NEXT:    slli a0, a0, 2
 ; RV32-BITS-512-NEXT:    addi a0, a0, -1
-; RV32-BITS-512-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
-; RV32-BITS-512-NEXT:    vid.v v16
 ; RV32-BITS-512-NEXT:    vrsub.vx v24, v16, a0
 ; RV32-BITS-512-NEXT:    vrgather.vv v20, v8, v24
 ; RV32-BITS-512-NEXT:    vrgather.vv v16, v12, v24
 ; RV32-BITS-512-NEXT:    vmv8r.v v8, v16
 ; RV32-BITS-512-NEXT:    ret
 ;
 ; RV64-BITS-UNKNOWN-LABEL: reverse_nxv64i8:
 ; RV64-BITS-UNKNOWN:       # %bb.0:
+; RV64-BITS-UNKNOWN-NEXT:    vsetvli a0, zero, e16, m8, ta, ma
+; RV64-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV64-BITS-UNKNOWN-NEXT:    csrr a0, vlenb
 ; RV64-BITS-UNKNOWN-NEXT:    slli a0, a0, 2
 ; RV64-BITS-UNKNOWN-NEXT:    addi a0, a0, -1
-; RV64-BITS-UNKNOWN-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
-; RV64-BITS-UNKNOWN-NEXT:    vid.v v16
 ; RV64-BITS-UNKNOWN-NEXT:    vrsub.vx v24, v16, a0
 ; RV64-BITS-UNKNOWN-NEXT:    vsetvli zero, zero, e8, m4, ta, ma
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v20, v8, v24
 ; RV64-BITS-UNKNOWN-NEXT:    vrgatherei16.vv v16, v12, v24
 ; RV64-BITS-UNKNOWN-NEXT:    vmv8r.v v8, v16
 ; RV64-BITS-UNKNOWN-NEXT:    ret
 ;
 ; RV64-BITS-256-LABEL: reverse_nxv64i8:
 ; RV64-BITS-256:       # %bb.0:
+; RV64-BITS-256-NEXT:    vsetvli a0, zero, e8, m8, ta, ma
+; RV64-BITS-256-NEXT:    vid.v v16
 ; RV64-BITS-256-NEXT:    csrr a0, vlenb
 ; RV64-BITS-256-NEXT:    slli a0, a0, 3
 ; RV64-BITS-256-NEXT:    addi a0, a0, -1
-; RV64-BITS-256-NEXT:    vsetvli a1, zero, e8, m8, ta, ma
-; RV64-BITS-256-NEXT:    vid.v v16
 ; RV64-BITS-256-NEXT:    vrsub.vx v24, v16, a0
 ; RV64-BITS-256-NEXT:    vrgather.vv v16, v8, v24
 ; RV64-BITS-256-NEXT:    vmv.v.v v8, v16
 ; RV64-BITS-256-NEXT:    ret
 ;
 ; RV64-BITS-512-LABEL: reverse_nxv64i8:
 ; RV64-BITS-512:       # %bb.0:
+; RV64-BITS-512-NEXT:    vsetvli a0, zero, e8, m4, ta, ma
+; RV64-BITS-512-NEXT:    vid.v v16
 ; RV64-BITS-512-NEXT:    csrr a0, vlenb
 ; RV64-BITS-512-NEXT:    slli a0, a0, 2
 ; RV64-BITS-512-NEXT:    addi a0, a0, -1
-; RV64-BITS-512-NEXT:    vsetvli a1, zero, e8, m4, ta, ma
-; RV64-BITS-512-NEXT:    vid.v v16
 ; RV64-BITS-512-NEXT:    vrsub.vx v24, v16, a0
 ; RV64-BITS-512-NEXT:    vrgather.vv v20, v8, v24
 ; RV64-BITS-512-NEXT:    vrgather.vv v16, v12, v24
 ; RV64-BITS-512-NEXT:    vmv8r.v v8, v16
 ; RV64-BITS-512-NEXT:    ret
   %res = call <vscale x 64 x i8> @llvm.experimental.vector.reverse.nxv64i8(<vscale x 64 x i8> %a)
   ret <vscale x 64 x i8> %res
 }
 
 define <vscale x 1 x i16> @reverse_nxv1i16(<vscale x 1 x i16> %a) {
 ; CHECK-LABEL: reverse_nxv1i16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, mf4, ta, ma
+; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 3
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, mf4, ta, ma
-; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    vrsub.vx v10, v9, a0
 ; CHECK-NEXT:    vrgather.vv v9, v8, v10
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %res = call <vscale x 1 x i16> @llvm.experimental.vector.reverse.nxv1i16(<vscale x 1 x i16> %a)
   ret <vscale x 1 x i16> %res
 }
 
 define <vscale x 2 x i16> @reverse_nxv2i16(<vscale x 2 x i16> %a) {
 ; CHECK-LABEL: reverse_nxv2i16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, mf2, ta, ma
+; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 2
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, mf2, ta, ma
-; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    vrsub.vx v10, v9, a0
 ; CHECK-NEXT:    vrgather.vv v9, v8, v10
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %res = call <vscale x 2 x i16> @llvm.experimental.vector.reverse.nxv2i16(<vscale x 2 x i16> %a)
   ret <vscale x 2 x i16> %res
 }
 
 define <vscale x 4 x i16> @reverse_nxv4i16(<vscale x 4 x i16> %a) {
 ; CHECK-LABEL: reverse_nxv4i16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m1, ta, ma
+; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
-; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    vrsub.vx v10, v9, a0
 ; CHECK-NEXT:    vrgather.vv v9, v8, v10
 ; CHECK-NEXT:    vmv.v.v v8, v9
 ; CHECK-NEXT:    ret
   %res = call <vscale x 4 x i16> @llvm.experimental.vector.reverse.nxv4i16(<vscale x 4 x i16> %a)
   ret <vscale x 4 x i16> %res
 }
 

 ; CHECK-NEXT:    ret
   %res = call <vscale x 8 x i16> @llvm.experimental.vector.reverse.nxv8i16(<vscale x 8 x i16> %a)
   ret <vscale x 8 x i16> %res
 }
 
 define <vscale x 16 x i16> @reverse_nxv16i16(<vscale x 16 x i16> %a) {
 ; CHECK-LABEL: reverse_nxv16i16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vid.v v12
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
-; CHECK-NEXT:    vid.v v12
 ; CHECK-NEXT:    vrsub.vx v16, v12, a0
 ; CHECK-NEXT:    vrgather.vv v12, v8, v16
 ; CHECK-NEXT:    vmv.v.v v8, v12
 ; CHECK-NEXT:    ret
   %res = call <vscale x 16 x i16> @llvm.experimental.vector.reverse.nxv16i16(<vscale x 16 x i16> %a)
   ret <vscale x 16 x i16> %res
 }
 
 define <vscale x 32 x i16> @reverse_nxv32i16(<vscale x 32 x i16> %a) {
 ; CHECK-LABEL: reverse_nxv32i16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m8, ta, ma
+; CHECK-NEXT:    vid.v v16
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 2
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
-; CHECK-NEXT:    vid.v v16
 ; CHECK-NEXT:    vrsub.vx v24, v16, a0
 ; CHECK-NEXT:    vrgather.vv v16, v8, v24
 ; CHECK-NEXT:    vmv.v.v v8, v16
 ; CHECK-NEXT:    ret
   %res = call <vscale x 32 x i16> @llvm.experimental.vector.reverse.nxv32i16(<vscale x 32 x i16> %a)
   ret <vscale x 32 x i16> %res
 }
 
 define <vscale x 1 x i32> @reverse_nxv1i32(<vscale x 1 x i32> %a) {
 ; CHECK-LABEL: reverse_nxv1i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, mf2, ta, ma
+; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 3
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e32, mf2, ta, ma
-; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    vrsub.vx v10, v9, a0
 ; CHECK-NEXT:    vrgather.vv v9, v8, v10
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %res = call <vscale x 1 x i32> @llvm.experimental.vector.reverse.nxv1i32(<vscale x 1 x i32> %a)
   ret <vscale x 1 x i32> %res
 }
 
 define <vscale x 2 x i32> @reverse_nxv2i32(<vscale x 2 x i32> %a) {
 ; CHECK-LABEL: reverse_nxv2i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m1, ta, ma
+; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 2
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e32, m1, ta, ma
-; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    vrsub.vx v10, v9, a0
 ; CHECK-NEXT:    vrgather.vv v9, v8, v10
 ; CHECK-NEXT:    vmv.v.v v8, v9
 ; CHECK-NEXT:    ret
   %res = call <vscale x 2 x i32> @llvm.experimental.vector.reverse.nxv2i32(<vscale x 2 x i32> %a)
   ret <vscale x 2 x i32> %res
 }
 
 define <vscale x 4 x i32> @reverse_nxv4i32(<vscale x 4 x i32> %a) {
 ; CHECK-LABEL: reverse_nxv4i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m2, ta, ma
+; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e32, m2, ta, ma
-; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    vrsub.vx v12, v10, a0
 ; CHECK-NEXT:    vrgather.vv v10, v8, v12
 ; CHECK-NEXT:    vmv.v.v v8, v10
 ; CHECK-NEXT:    ret
   %res = call <vscale x 4 x i32> @llvm.experimental.vector.reverse.nxv4i32(<vscale x 4 x i32> %a)
   ret <vscale x 4 x i32> %res
 }
 

 ; CHECK-NEXT:    ret
   %res = call <vscale x 8 x i32> @llvm.experimental.vector.reverse.nxv8i32(<vscale x 8 x i32> %a)
   ret <vscale x 8 x i32> %res
 }
 
 define <vscale x 16 x i32> @reverse_nxv16i32(<vscale x 16 x i32> %a) {
 ; CHECK-LABEL: reverse_nxv16i32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m8, ta, ma
+; CHECK-NEXT:    vid.v v16
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e32, m8, ta, ma
-; CHECK-NEXT:    vid.v v16
 ; CHECK-NEXT:    vrsub.vx v24, v16, a0
 ; CHECK-NEXT:    vrgather.vv v16, v8, v24
 ; CHECK-NEXT:    vmv.v.v v8, v16
 ; CHECK-NEXT:    ret
   %res = call <vscale x 16 x i32> @llvm.experimental.vector.reverse.nxv16i32(<vscale x 16 x i32> %a)
   ret <vscale x 16 x i32> %res
 }
 
 define <vscale x 1 x i64> @reverse_nxv1i64(<vscale x 1 x i64> %a) {
 ; CHECK-LABEL: reverse_nxv1i64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e64, m1, ta, ma
+; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 3
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e64, m1, ta, ma
-; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    vrsub.vx v10, v9, a0
 ; CHECK-NEXT:    vrgather.vv v9, v8, v10
 ; CHECK-NEXT:    vmv.v.v v8, v9
 ; CHECK-NEXT:    ret
   %res = call <vscale x 1 x i64> @llvm.experimental.vector.reverse.nxv1i64(<vscale x 1 x i64> %a)
   ret <vscale x 1 x i64> %res
 }
 
 define <vscale x 2 x i64> @reverse_nxv2i64(<vscale x 2 x i64> %a) {
 ; CHECK-LABEL: reverse_nxv2i64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e64, m2, ta, ma
+; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 2
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e64, m2, ta, ma
-; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    vrsub.vx v12, v10, a0
 ; CHECK-NEXT:    vrgather.vv v10, v8, v12
 ; CHECK-NEXT:    vmv.v.v v8, v10
 ; CHECK-NEXT:    ret
   %res = call <vscale x 2 x i64> @llvm.experimental.vector.reverse.nxv2i64(<vscale x 2 x i64> %a)
   ret <vscale x 2 x i64> %res
 }
 
 define <vscale x 4 x i64> @reverse_nxv4i64(<vscale x 4 x i64> %a) {
 ; CHECK-LABEL: reverse_nxv4i64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e64, m4, ta, ma
+; CHECK-NEXT:    vid.v v12
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e64, m4, ta, ma
-; CHECK-NEXT:    vid.v v12
 ; CHECK-NEXT:    vrsub.vx v16, v12, a0
 ; CHECK-NEXT:    vrgather.vv v12, v8, v16
 ; CHECK-NEXT:    vmv.v.v v8, v12
 ; CHECK-NEXT:    ret
   %res = call <vscale x 4 x i64> @llvm.experimental.vector.reverse.nxv4i64(<vscale x 4 x i64> %a)
   ret <vscale x 4 x i64> %res
 }
 

 
 ;
 ; VECTOR_REVERSE - floating point
 ;
 
 define <vscale x 1 x half> @reverse_nxv1f16(<vscale x 1 x half> %a) {
 ; CHECK-LABEL: reverse_nxv1f16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, mf4, ta, ma
+; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 3
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, mf4, ta, ma
-; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    vrsub.vx v10, v9, a0
 ; CHECK-NEXT:    vrgather.vv v9, v8, v10
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %res = call <vscale x 1 x half> @llvm.experimental.vector.reverse.nxv1f16(<vscale x 1 x half> %a)
   ret <vscale x 1 x half> %res
 }
 
 define <vscale x 2 x half> @reverse_nxv2f16(<vscale x 2 x half> %a) {
 ; CHECK-LABEL: reverse_nxv2f16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, mf2, ta, ma
+; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 2
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, mf2, ta, ma
-; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    vrsub.vx v10, v9, a0
 ; CHECK-NEXT:    vrgather.vv v9, v8, v10
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %res = call <vscale x 2 x half> @llvm.experimental.vector.reverse.nxv2f16(<vscale x 2 x half> %a)
   ret <vscale x 2 x half> %res
 }
 
 define <vscale x 4 x half> @reverse_nxv4f16(<vscale x 4 x half> %a) {
 ; CHECK-LABEL: reverse_nxv4f16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m1, ta, ma
+; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, m1, ta, ma
-; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    vrsub.vx v10, v9, a0
 ; CHECK-NEXT:    vrgather.vv v9, v8, v10
 ; CHECK-NEXT:    vmv.v.v v8, v9
 ; CHECK-NEXT:    ret
   %res = call <vscale x 4 x half> @llvm.experimental.vector.reverse.nxv4f16(<vscale x 4 x half> %a)
   ret <vscale x 4 x half> %res
 }
 

 ; CHECK-NEXT:    ret
   %res = call <vscale x 8 x half> @llvm.experimental.vector.reverse.nxv8f16(<vscale x 8 x half> %a)
   ret <vscale x 8 x half> %res
 }
 
 define <vscale x 16 x half> @reverse_nxv16f16(<vscale x 16 x half> %a) {
 ; CHECK-LABEL: reverse_nxv16f16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m4, ta, ma
+; CHECK-NEXT:    vid.v v12
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, m4, ta, ma
-; CHECK-NEXT:    vid.v v12
 ; CHECK-NEXT:    vrsub.vx v16, v12, a0
 ; CHECK-NEXT:    vrgather.vv v12, v8, v16
 ; CHECK-NEXT:    vmv.v.v v8, v12
 ; CHECK-NEXT:    ret
   %res = call <vscale x 16 x half> @llvm.experimental.vector.reverse.nxv16f16(<vscale x 16 x half> %a)
   ret <vscale x 16 x half> %res
 }
 
 define <vscale x 32 x half> @reverse_nxv32f16(<vscale x 32 x half> %a) {
 ; CHECK-LABEL: reverse_nxv32f16:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e16, m8, ta, ma
+; CHECK-NEXT:    vid.v v16
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 2
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e16, m8, ta, ma
-; CHECK-NEXT:    vid.v v16
 ; CHECK-NEXT:    vrsub.vx v24, v16, a0
 ; CHECK-NEXT:    vrgather.vv v16, v8, v24
 ; CHECK-NEXT:    vmv.v.v v8, v16
 ; CHECK-NEXT:    ret
   %res = call <vscale x 32 x half> @llvm.experimental.vector.reverse.nxv32f16(<vscale x 32 x half> %a)
   ret <vscale x 32 x half> %res
 }
 
 define <vscale x 1 x float> @reverse_nxv1f32(<vscale x 1 x float> %a) {
 ; CHECK-LABEL: reverse_nxv1f32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, mf2, ta, ma
+; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 3
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e32, mf2, ta, ma
-; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    vrsub.vx v10, v9, a0
 ; CHECK-NEXT:    vrgather.vv v9, v8, v10
 ; CHECK-NEXT:    vmv1r.v v8, v9
 ; CHECK-NEXT:    ret
   %res = call <vscale x 1 x float> @llvm.experimental.vector.reverse.nxv1f32(<vscale x 1 x float> %a)
   ret <vscale x 1 x float> %res
 }
 
 define <vscale x 2 x float> @reverse_nxv2f32(<vscale x 2 x float> %a) {
 ; CHECK-LABEL: reverse_nxv2f32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m1, ta, ma
+; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 2
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e32, m1, ta, ma
-; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    vrsub.vx v10, v9, a0
 ; CHECK-NEXT:    vrgather.vv v9, v8, v10
 ; CHECK-NEXT:    vmv.v.v v8, v9
 ; CHECK-NEXT:    ret
   %res = call <vscale x 2 x float> @llvm.experimental.vector.reverse.nxv2f32(<vscale x 2 x float> %a)
   ret <vscale x 2 x float> %res
 }
 
 define <vscale x 4 x float> @reverse_nxv4f32(<vscale x 4 x float> %a) {
 ; CHECK-LABEL: reverse_nxv4f32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m2, ta, ma
+; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e32, m2, ta, ma
-; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    vrsub.vx v12, v10, a0
 ; CHECK-NEXT:    vrgather.vv v10, v8, v12
 ; CHECK-NEXT:    vmv.v.v v8, v10
 ; CHECK-NEXT:    ret
   %res = call <vscale x 4 x float> @llvm.experimental.vector.reverse.nxv4f32(<vscale x 4 x float> %a)
   ret <vscale x 4 x float> %res
 }
 

 ; CHECK-NEXT:    ret
   %res = call <vscale x 8 x float> @llvm.experimental.vector.reverse.nxv8f32(<vscale x 8 x float> %a)
   ret <vscale x 8 x float> %res
 }
 
 define <vscale x 16 x float> @reverse_nxv16f32(<vscale x 16 x float> %a) {
 ; CHECK-LABEL: reverse_nxv16f32:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e32, m8, ta, ma
+; CHECK-NEXT:    vid.v v16
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    slli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e32, m8, ta, ma
-; CHECK-NEXT:    vid.v v16
 ; CHECK-NEXT:    vrsub.vx v24, v16, a0
 ; CHECK-NEXT:    vrgather.vv v16, v8, v24
 ; CHECK-NEXT:    vmv.v.v v8, v16
 ; CHECK-NEXT:    ret
   %res = call <vscale x 16 x float> @llvm.experimental.vector.reverse.nxv16f32(<vscale x 16 x float> %a)
   ret <vscale x 16 x float> %res
 }
 
 define <vscale x 1 x double> @reverse_nxv1f64(<vscale x 1 x double> %a) {
 ; CHECK-LABEL: reverse_nxv1f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e64, m1, ta, ma
+; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 3
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e64, m1, ta, ma
-; CHECK-NEXT:    vid.v v9
 ; CHECK-NEXT:    vrsub.vx v10, v9, a0
 ; CHECK-NEXT:    vrgather.vv v9, v8, v10
 ; CHECK-NEXT:    vmv.v.v v8, v9
 ; CHECK-NEXT:    ret
   %res = call <vscale x 1 x double> @llvm.experimental.vector.reverse.nxv1f64(<vscale x 1 x double> %a)
   ret <vscale x 1 x double> %res
 }
 
 define <vscale x 2 x double> @reverse_nxv2f64(<vscale x 2 x double> %a) {
 ; CHECK-LABEL: reverse_nxv2f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e64, m2, ta, ma
+; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 2
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e64, m2, ta, ma
-; CHECK-NEXT:    vid.v v10
 ; CHECK-NEXT:    vrsub.vx v12, v10, a0
 ; CHECK-NEXT:    vrgather.vv v10, v8, v12
 ; CHECK-NEXT:    vmv.v.v v8, v10
 ; CHECK-NEXT:    ret
   %res = call <vscale x 2 x double> @llvm.experimental.vector.reverse.nxv2f64(<vscale x 2 x double> %a)
   ret <vscale x 2 x double> %res
 }
 
 define <vscale x 4 x double> @reverse_nxv4f64(<vscale x 4 x double> %a) {
 ; CHECK-LABEL: reverse_nxv4f64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e64, m4, ta, ma
+; CHECK-NEXT:    vid.v v12
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e64, m4, ta, ma
-; CHECK-NEXT:    vid.v v12
 ; CHECK-NEXT:    vrsub.vx v16, v12, a0
 ; CHECK-NEXT:    vrgather.vv v12, v8, v16
 ; CHECK-NEXT:    vmv.v.v v8, v12
 ; CHECK-NEXT:    ret
   %res = call <vscale x 4 x double> @llvm.experimental.vector.reverse.nxv4f64(<vscale x 4 x double> %a)
   ret <vscale x 4 x double> %res
 }
 
; CHECK-NEXT:    ret
   ret <vscale x 8 x double> %res
 }
 
 ; Test widen reverse vector
 
 define <vscale x 3 x i64> @reverse_nxv3i64(<vscale x 3 x i64> %a) {
 ; CHECK-LABEL: reverse_nxv3i64:
 ; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e64, m4, ta, ma
+; CHECK-NEXT:    vid.v v12
 ; CHECK-NEXT:    csrr a0, vlenb
 ; CHECK-NEXT:    srli a0, a0, 1
 ; CHECK-NEXT:    addi a0, a0, -1
-; CHECK-NEXT:    vsetvli a1, zero, e64, m4, ta, ma
-; CHECK-NEXT:    vid.v v12
 ; CHECK-NEXT:    vrsub.vx v12, v12, a0
 ; CHECK-NEXT:    vrgather.vv v16, v8, v12
 ; CHECK-NEXT:    vmv1r.v v8, v17
 ; CHECK-NEXT:    vmv1r.v v9, v18
 ; CHECK-NEXT:    vmv1r.v v10, v19
 ; CHECK-NEXT:    ret
   %res = call <vscale x 3 x i64> @llvm.experimental.vector.reverse.nxv3i64(<vscale x 3 x i64> %a)
   ret <vscale x 3 x i64> %res

llvm/test/CodeGen/RISCV/rvv/stepvector.ll

 ; RV32:       # %bb.0: # %entry
 ; RV32-NEXT:    addi sp, sp, -16
 ; RV32-NEXT:    .cfi_def_cfa_offset 16
 ; RV32-NEXT:    li a0, 7
 ; RV32-NEXT:    sw a0, 12(sp)
 ; RV32-NEXT:    lui a0, 797989
 ; RV32-NEXT:    addi a0, a0, -683
 ; RV32-NEXT:    sw a0, 8(sp)
+; RV32-NEXT:    vsetvli a0, zero, e64, m8, ta, ma
 ; RV32-NEXT:    addi a0, sp, 8
-; RV32-NEXT:    vsetvli a1, zero, e64, m8, ta, ma
 ; RV32-NEXT:    vlse64.v v8, (a0), zero
 ; RV32-NEXT:    vid.v v16
 ; RV32-NEXT:    vmul.vv v8, v16, v8
 ; RV32-NEXT:    addi sp, sp, 16
 ; RV32-NEXT:    ret
 ;
 ; RV64-LABEL: mul_bigimm_stepvector_nxv8i64:
 ; RV64:       # %bb.0: # %entry

 define <vscale x 16 x i64> @stepvector_nxv16i64() {
 ; RV32-LABEL: stepvector_nxv16i64:
 ; RV32:       # %bb.0:
 ; RV32-NEXT:    addi sp, sp, -16
 ; RV32-NEXT:    .cfi_def_cfa_offset 16
 ; RV32-NEXT:    sw zero, 12(sp)
 ; RV32-NEXT:    csrr a0, vlenb
 ; RV32-NEXT:    sw a0, 8(sp)
+; RV32-NEXT:    vsetvli a0, zero, e64, m8, ta, ma
 ; RV32-NEXT:    addi a0, sp, 8
-; RV32-NEXT:    vsetvli a1, zero, e64, m8, ta, ma
 ; RV32-NEXT:    vlse64.v v16, (a0), zero
 ; RV32-NEXT:    vid.v v8
 ; RV32-NEXT:    vadd.vv v16, v8, v16
 ; RV32-NEXT:    addi sp, sp, 16
 ; RV32-NEXT:    ret
 ;
 ; RV64-LABEL: stepvector_nxv16i64:
 ; RV64:       # %bb.0:
-; RV64-NEXT:    csrr a0, vlenb
-; RV64-NEXT:    vsetvli a1, zero, e64, m8, ta, ma
+; RV64-NEXT:    vsetvli a0, zero, e64, m8, ta, ma
 ; RV64-NEXT:    vid.v v8
+; RV64-NEXT:    csrr a0, vlenb
 ; RV64-NEXT:    vadd.vx v16, v8, a0
 ; RV64-NEXT:    ret
   %v = call <vscale x 16 x i64> @llvm.experimental.stepvector.nxv16i64()
   ret <vscale x 16 x i64> %v
 }
 
 define <vscale x 16 x i64> @add_stepvector_nxv16i64() {
 ; RV32-LABEL: add_stepvector_nxv16i64:

 ; RV32-NEXT:    li a0, 3
 ; RV32-NEXT:    vmul.vx v8, v8, a0
 ; RV32-NEXT:    vadd.vv v16, v8, v16
 ; RV32-NEXT:    addi sp, sp, 16
 ; RV32-NEXT:    ret
 ;
 ; RV64-LABEL: mul_stepvector_nxv16i64:
 ; RV64:       # %bb.0: # %entry
-; RV64-NEXT:    vsetvli a0, zero, e64, m8, ta, ma
-; RV64-NEXT:    vid.v v8
-; RV64-NEXT:    li a0, 3
-; RV64-NEXT:    vmul.vx v8, v8, a0
 ; RV64-NEXT:    csrr a0, vlenb
 ; RV64-NEXT:    slli a1, a0, 1
 ; RV64-NEXT:    add a0, a1, a0
+; RV64-NEXT:    vsetvli a1, zero, e64, m8, ta, ma
+; RV64-NEXT:    vid.v v8
+; RV64-NEXT:    li a1, 3
+; RV64-NEXT:    vmul.vx v8, v8, a1
 ; RV64-NEXT:    vadd.vx v16, v8, a0
 ; RV64-NEXT:    ret
 entry:
   %0 = insertelement <vscale x 16 x i64> poison, i64 3, i32 0
   %1 = shufflevector <vscale x 16 x i64> %0, <vscale x 16 x i64> poison, <vscale x 16 x i32> zeroinitializer
   %2 = call <vscale x 16 x i64> @llvm.experimental.stepvector.nxv16i64()
   %3 = mul <vscale x 16 x i64> %2, %1
   ret <vscale x 16 x i64> %3

 ; RV32-NEXT:    srli a0, a0, 3
 ; RV32-NEXT:    li a1, 62
 ; RV32-NEXT:    mul a1, a0, a1
 ; RV32-NEXT:    lui a2, 92455
 ; RV32-NEXT:    addi a2, a2, -1368
 ; RV32-NEXT:    mulhu a0, a0, a2
 ; RV32-NEXT:    add a0, a0, a1
 ; RV32-NEXT:    sw a0, 12(sp)
+; RV32-NEXT:    vsetvli a0, zero, e64, m8, ta, ma
 ; RV32-NEXT:    addi a0, sp, 8
-; RV32-NEXT:    vsetvli a1, zero, e64, m8, ta, ma
 ; RV32-NEXT:    vlse64.v v8, (a0), zero
 ; RV32-NEXT:    vlse64.v v16, (a0), zero
 ; RV32-NEXT:    vid.v v24
 ; RV32-NEXT:    vmul.vv v8, v24, v8
 ; RV32-NEXT:    vadd.vv v16, v8, v16
 ; RV32-NEXT:    addi sp, sp, 16
 ; RV32-NEXT:    ret
 ;