[RISCV] Support INSERT_SUBVECTOR on vector masks

[frasercrmck]

Summary

Like with EXTRACT_SUBVECTOR, INSERT_SUBVECTOR poses a problem
for vector masks as RVV isn't able to slide mask types around. We choose
instead to bitcast to equivalently-sized i8 types where we can, else we
zero-extend, perform the operation, and truncate back down.

One test was left disabled due to a crash in the legalizer.

[frasercrmck]

• add missing bitcast

[craig.topper]

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert-subvector.ll

396

I'm working on a fix for this.

[craig.topper]

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert-subvector.ll

396

https://reviews.llvm.org/D97582 should fix this. Before the introduction of experimental.vector.insert there was almost no way to create an INSERT_SUBVECTOR until after type legalization so the code wasn't well exercised.

[craig.topper]

LGTM

llvm/lib/Target/RISCV/RISCVISelLowering.cpp

2583

You can call getBitcast unconditionally.

[frasercrmck]

llvm/lib/Target/RISCV/RISCVISelLowering.cpp

2583

Yep, did so prior to merging. Cheers.

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert-subvector.ll

396

Thanks. Now that this was merged you can hopefully enable this as your test case.

llvm/lib/Target/RISCV/RISCVISelLowering.cpp

if (Subtarget.is64Bit()) {
       setOperationAction(ISD::VECREDUCE_UMIN, MVT::i64, Custom);
     }
 
     for (MVT VT : BoolVecVTs) {
       setOperationAction(ISD::SPLAT_VECTOR, VT, Legal);
 
       // Mask VTs are custom-expanded into a series of standard nodes
       setOperationAction(ISD::TRUNCATE, VT, Custom);
+      setOperationAction(ISD::INSERT_SUBVECTOR, VT, Custom);
       setOperationAction(ISD::EXTRACT_SUBVECTOR, VT, Custom);
     }
 
     for (MVT VT : IntVecVTs) {
       setOperationAction(ISD::SPLAT_VECTOR, VT, Legal);
 
       setOperationAction(ISD::SMIN, VT, Legal);
       setOperationAction(ISD::SMAX, VT, Legal);
SDValue RISCVTargetLowering::lowerINSERT_SUBVECTOR(SDValue Op,
   MVT VecVT = Vec.getSimpleValueType();
   MVT SubVecVT = SubVec.getSimpleValueType();
 
   SDLoc DL(Op);
   MVT XLenVT = Subtarget.getXLenVT();
   unsigned OrigIdx = Op.getConstantOperandVal(2);
   const RISCVRegisterInfo *TRI = Subtarget.getRegisterInfo();
 
+  // We don't have the ability to slide mask vectors up indexed by their i1
+  // elements; the smallest we can do is i8. Often we are able to bitcast to
+  // equivalent i8 vectors. Note that when inserting a fixed-length vector
+  // into a scalable one, we might not necessarily have enough scalable
+  // elements to safely divide by 8: nxv1i1 = insert nxv1i1, v4i1 is valid.
+  if (SubVecVT.getVectorElementType() == MVT::i1 &&
+      (OrigIdx != 0 || !Vec.isUndef())) {
+    if (VecVT.getVectorMinNumElements() >= 8 &&
+        SubVecVT.getVectorMinNumElements() >= 8) {
+      assert(OrigIdx % 8 == 0 && "Invalid index");
+      assert(VecVT.getVectorMinNumElements() % 8 == 0 &&
+             SubVecVT.getVectorMinNumElements() % 8 == 0 &&
+             "Unexpected mask vector lowering");
+      OrigIdx /= 8;
+      SubVecVT =
+          MVT::getVectorVT(MVT::i8, SubVecVT.getVectorMinNumElements() / 8,
+                           SubVecVT.isScalableVector());
+      VecVT = MVT::getVectorVT(MVT::i8, VecVT.getVectorMinNumElements() / 8,
+                               VecVT.isScalableVector());
+      Vec = DAG.getBitcast(VecVT, Vec);
+      SubVec = DAG.getBitcast(SubVecVT, SubVec);
+    } else {
+      // We can't slide this mask vector up indexed by its i1 elements.
+      // This poses a problem when we wish to insert a scalable vector which
+      // can't be re-expressed as a larger type. Just choose the slow path and
+      // extend to a larger type, then truncate back down.
+      MVT ExtVecVT = VecVT.changeVectorElementType(MVT::i8);
+      MVT ExtSubVecVT = SubVecVT.changeVectorElementType(MVT::i8);
+      Vec = DAG.getNode(ISD::ZERO_EXTEND, DL, ExtVecVT, Vec);
+      SubVec = DAG.getNode(ISD::ZERO_EXTEND, DL, ExtSubVecVT, SubVec);
+      Vec = DAG.getNode(ISD::INSERT_SUBVECTOR, DL, ExtVecVT, Vec, SubVec,
+                        Op.getOperand(2));
+      SDValue SplatZero = DAG.getConstant(0, DL, ExtVecVT);
+      return DAG.getSetCC(DL, VecVT, Vec, SplatZero, ISD::SETNE);
+    }
+  }
+
   // If the subvector vector is a fixed-length type, we cannot use subregister
   // manipulation to simplify the codegen; we don't know which register of a
   // LMUL group contains the specific subvector as we only know the minimum
   // register size. Therefore we must slide the vector group up the full
   // amount.
   if (SubVecVT.isFixedLengthVector()) {
     if (OrigIdx == 0 && Vec.isUndef())
       return Op;
SDValue RISCVTargetLowering::lowerINSERT_SUBVECTOR(SDValue Op,
   SDValue Slideup = DAG.getNode(RISCVISD::VSLIDEUP_VL, DL, InterSubVT,
                                 AlignedExtract, SubVec, SlideupAmt, Mask, VL);
 
   // If required, insert this subvector back into the correct vector register.
   // This should resolve to an INSERT_SUBREG instruction.
   if (VecVT.bitsGT(InterSubVT))
     Slideup = DAG.getNode(ISD::INSERT_SUBVECTOR, DL, VecVT, Vec, Slideup,
                           DAG.getConstant(AlignedIdx, DL, XLenVT));
-  return Slideup;
+
+  // We might have bitcast from a mask type: cast back to the original type if
+  // required.
+  return DAG.getBitcast(Op.getSimpleValueType(), Slideup);
 }
 
 SDValue RISCVTargetLowering::lowerEXTRACT_SUBVECTOR(SDValue Op,
                                                     SelectionDAG &DAG) const {
   SDValue Vec = Op.getOperand(0);
   MVT SubVecVT = Op.getSimpleValueType();
   MVT VecVT = Vec.getSimpleValueType();
 

llvm/test/CodeGen/RISCV/rvv/fixed-vectors-insert-subvector.ll

 ; CHECK-NEXT:    ret
   %v = load <4 x i16>, <4 x i16>* %vp
   %sv = load <2 x i16>, <2 x i16>* %svp
   %c = call <4 x i16> @llvm.experimental.vector.insert.v2i16.v4i16(<4 x i16> %v, <2 x i16> %sv, i64 2)
   store <4 x i16> %c, <4 x i16>* %vp
   ret void
 }
 
+define void @insert_v32i1_v8i1_0(<32 x i1>* %vp, <8 x i1>* %svp) {
+; LMULMAX2-LABEL: insert_v32i1_v8i1_0:
+; LMULMAX2:       # %bb.0:
+; LMULMAX2-NEXT:    addi a2, zero, 32
+; LMULMAX2-NEXT:    vsetvli a3, a2, e8,m2,ta,mu
+; LMULMAX2-NEXT:    vle1.v v25, (a0)
+; LMULMAX2-NEXT:    vsetivli a3, 8, e8,m1,ta,mu
+; LMULMAX2-NEXT:    vle1.v v26, (a1)
+; LMULMAX2-NEXT:    vsetivli a1, 1, e8,m1,tu,mu
+; LMULMAX2-NEXT:    vslideup.vi v25, v26, 0
+; LMULMAX2-NEXT:    vsetvli a1, a2, e8,m2,ta,mu
+; LMULMAX2-NEXT:    vse1.v v25, (a0)
+; LMULMAX2-NEXT:    ret
+;
+; LMULMAX1-LABEL: insert_v32i1_v8i1_0:
+; LMULMAX1:       # %bb.0:
+; LMULMAX1-NEXT:    vsetivli a2, 16, e8,m1,ta,mu
+; LMULMAX1-NEXT:    vle1.v v25, (a0)
+; LMULMAX1-NEXT:    vsetivli a2, 8, e8,m1,ta,mu
+; LMULMAX1-NEXT:    vle1.v v26, (a1)
+; LMULMAX1-NEXT:    vsetivli a1, 1, e8,m1,tu,mu
+; LMULMAX1-NEXT:    vslideup.vi v25, v26, 0
+; LMULMAX1-NEXT:    vsetivli a1, 16, e8,m1,ta,mu
+; LMULMAX1-NEXT:    vse1.v v25, (a0)
+; LMULMAX1-NEXT:    ret
+  %v = load <32 x i1>, <32 x i1>* %vp
+  %sv = load <8 x i1>, <8 x i1>* %svp
+  %c = call <32 x i1> @llvm.experimental.vector.insert.v8i1.v32i1(<32 x i1> %v, <8 x i1> %sv, i64 0)
+  store <32 x i1> %c, <32 x i1>* %vp
+  ret void
+}
+
+; FIXME: SplitVecRes_INSERT_SUBVECTOR crashes on this one when trying to spill
+; to the stack.
+;define void @insert_v32i1_v8i1_16(<32 x i1>* %vp, <8 x i1>* %svp) {
+;  %v = load <32 x i1>, <32 x i1>* %vp
+;  %sv = load <8 x i1>, <8 x i1>* %svp
+;  %c = call <32 x i1> @llvm.experimental.vector.insert.v8i1.v32i1(<32 x i1> %v, <8 x i1> %sv, i64 16)
+;  store <32 x i1> %c, <32 x i1>* %vp
+;  ret void
+;}
+
+define void @insert_v8i1_v4i1_0(<8 x i1>* %vp, <4 x i1>* %svp) {
+; CHECK-LABEL: insert_v8i1_v4i1_0:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli a2, 8, e8,m1,ta,mu
+; CHECK-NEXT:    vle1.v v27, (a0)
+; CHECK-NEXT:    vsetivli a2, 4, e8,m1,ta,mu
+; CHECK-NEXT:    vle1.v v0, (a1)
+; CHECK-NEXT:    vmv.v.i v25, 0
+; CHECK-NEXT:    vmerge.vim v25, v25, 1, v0
+; CHECK-NEXT:    vsetivli a1, 8, e8,m1,ta,mu
+; CHECK-NEXT:    vmv.v.i v26, 0
+; CHECK-NEXT:    vmv1r.v v0, v27
+; CHECK-NEXT:    vmerge.vim v26, v26, 1, v0
+; CHECK-NEXT:    vsetivli a1, 4, e8,m1,tu,mu
+; CHECK-NEXT:    vslideup.vi v26, v25, 0
+; CHECK-NEXT:    vsetivli a1, 8, e8,m1,ta,mu
+; CHECK-NEXT:    vmsne.vi v25, v26, 0
+; CHECK-NEXT:    vse1.v v25, (a0)
+; CHECK-NEXT:    ret
+  %v = load <8 x i1>, <8 x i1>* %vp
+  %sv = load <4 x i1>, <4 x i1>* %svp
+  %c = call <8 x i1> @llvm.experimental.vector.insert.v4i1.v8i1(<8 x i1> %v, <4 x i1> %sv, i64 0)
+  store <8 x i1> %c, <8 x i1>* %vp
+  ret void
+}
+
+define void @insert_v8i1_v4i1_4(<8 x i1>* %vp, <4 x i1>* %svp) {
+; CHECK-LABEL: insert_v8i1_v4i1_4:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli a2, 8, e8,m1,ta,mu
+; CHECK-NEXT:    vle1.v v27, (a0)
+; CHECK-NEXT:    vsetivli a2, 4, e8,m1,ta,mu
+; CHECK-NEXT:    vle1.v v0, (a1)
+; CHECK-NEXT:    vmv.v.i v25, 0
+; CHECK-NEXT:    vmerge.vim v25, v25, 1, v0
+; CHECK-NEXT:    vsetivli a1, 8, e8,m1,ta,mu
+; CHECK-NEXT:    vmv.v.i v26, 0
+; CHECK-NEXT:    vmv1r.v v0, v27
+; CHECK-NEXT:    vmerge.vim v26, v26, 1, v0
+; CHECK-NEXT:    vsetivli a1, 8, e8,m1,tu,mu
+; CHECK-NEXT:    vslideup.vi v26, v25, 4
+; CHECK-NEXT:    vsetivli a1, 8, e8,m1,ta,mu
+; CHECK-NEXT:    vmsne.vi v25, v26, 0
+; CHECK-NEXT:    vse1.v v25, (a0)
+; CHECK-NEXT:    ret
+  %v = load <8 x i1>, <8 x i1>* %vp
+  %sv = load <4 x i1>, <4 x i1>* %svp
+  %c = call <8 x i1> @llvm.experimental.vector.insert.v4i1.v8i1(<8 x i1> %v, <4 x i1> %sv, i64 4)
+  store <8 x i1> %c, <8 x i1>* %vp
+  ret void
+}
+
 define <vscale x 2 x i16> @insert_nxv2i16_v2i16_0(<vscale x 2 x i16> %v, <2 x i16>* %svp) {
 ; CHECK-LABEL: insert_nxv2i16_v2i16_0:
 ; CHECK:       # %bb.0:
 ; CHECK-NEXT:    vsetivli a1, 2, e16,m1,ta,mu
 ; CHECK-NEXT:    vle16.v v25, (a0)
 ; CHECK-NEXT:    vsetivli a0, 2, e16,mf2,tu,mu
 ; CHECK-NEXT:    vslideup.vi v8, v25, 0
 ; CHECK-NEXT:    ret

 ; CHECK-NEXT:    vsetivli a0, 6, e16,mf2,tu,mu
 ; CHECK-NEXT:    vslideup.vi v8, v25, 4
 ; CHECK-NEXT:    ret
   %sv = load <2 x i16>, <2 x i16>* %svp
   %c = call <vscale x 2 x i16> @llvm.experimental.vector.insert.v2i16.nxv2i16(<vscale x 2 x i16> %v, <2 x i16> %sv, i64 4)
   ret <vscale x 2 x i16> %c
 }
 
+define <vscale x 2 x i1> @insert_nxv2i1_v4i1_0(<vscale x 2 x i1> %v, <4 x i1>* %svp) {
+; CHECK-LABEL: insert_nxv2i1_v4i1_0:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli a1, 4, e8,m1,ta,mu
+; CHECK-NEXT:    vle1.v v27, (a0)
+; CHECK-NEXT:    vsetvli a0, zero, e8,mf4,ta,mu
+; CHECK-NEXT:    vmv.v.i v25, 0
+; CHECK-NEXT:    vmerge.vim v25, v25, 1, v0
+; CHECK-NEXT:    vsetivli a0, 4, e8,m1,ta,mu
+; CHECK-NEXT:    vmv.v.i v26, 0
+; CHECK-NEXT:    vmv1r.v v0, v27
+; CHECK-NEXT:    vmerge.vim v26, v26, 1, v0
+; CHECK-NEXT:    vsetivli a0, 4, e8,mf4,tu,mu
+; CHECK-NEXT:    vslideup.vi v25, v26, 0
+; CHECK-NEXT:    vsetvli a0, zero, e8,mf4,ta,mu
+; CHECK-NEXT:    vmsne.vi v0, v25, 0
+; CHECK-NEXT:    ret
+  %sv = load <4 x i1>, <4 x i1>* %svp
+  %c = call <vscale x 2 x i1> @llvm.experimental.vector.insert.v4i1.nxv2i1(<vscale x 2 x i1> %v, <4 x i1> %sv, i64 0)
+  ret <vscale x 2 x i1> %c
+}
+
+define <vscale x 2 x i1> @insert_nxv2i1_v4i1_6(<vscale x 2 x i1> %v, <4 x i1>* %svp) {
+; CHECK-LABEL: insert_nxv2i1_v4i1_6:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli a1, 4, e8,m1,ta,mu
+; CHECK-NEXT:    vle1.v v27, (a0)
+; CHECK-NEXT:    vsetvli a0, zero, e8,mf4,ta,mu
+; CHECK-NEXT:    vmv.v.i v25, 0
+; CHECK-NEXT:    vmerge.vim v25, v25, 1, v0
+; CHECK-NEXT:    vsetivli a0, 4, e8,m1,ta,mu
+; CHECK-NEXT:    vmv.v.i v26, 0
+; CHECK-NEXT:    vmv1r.v v0, v27
+; CHECK-NEXT:    vmerge.vim v26, v26, 1, v0
+; CHECK-NEXT:    vsetivli a0, 10, e8,mf4,tu,mu
+; CHECK-NEXT:    vslideup.vi v25, v26, 6
+; CHECK-NEXT:    vsetvli a0, zero, e8,mf4,ta,mu
+; CHECK-NEXT:    vmsne.vi v0, v25, 0
+; CHECK-NEXT:    ret
+  %sv = load <4 x i1>, <4 x i1>* %svp
+  %c = call <vscale x 2 x i1> @llvm.experimental.vector.insert.v4i1.nxv2i1(<vscale x 2 x i1> %v, <4 x i1> %sv, i64 6)
+  ret <vscale x 2 x i1> %c
+}
+
+define <vscale x 8 x i1> @insert_nxv8i1_v4i1_0(<vscale x 8 x i1> %v, <8 x i1>* %svp) {
+; CHECK-LABEL: insert_nxv8i1_v4i1_0:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli a1, 8, e8,m1,ta,mu
+; CHECK-NEXT:    vle1.v v25, (a0)
+; CHECK-NEXT:    vsetivli a0, 1, e8,mf8,tu,mu
+; CHECK-NEXT:    vslideup.vi v0, v25, 0
+; CHECK-NEXT:    ret
+  %sv = load <8 x i1>, <8 x i1>* %svp
+  %c = call <vscale x 8 x i1> @llvm.experimental.vector.insert.v8i1.nxv8i1(<vscale x 8 x i1> %v, <8 x i1> %sv, i64 0)
+  ret <vscale x 8 x i1> %c
+}
+
+define <vscale x 8 x i1> @insert_nxv8i1_v8i1_16(<vscale x 8 x i1> %v, <8 x i1>* %svp) {
+; CHECK-LABEL: insert_nxv8i1_v8i1_16:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetivli a1, 8, e8,m1,ta,mu
+; CHECK-NEXT:    vle1.v v25, (a0)
+; CHECK-NEXT:    vsetivli a0, 3, e8,mf8,tu,mu
+; CHECK-NEXT:    vslideup.vi v0, v25, 2
+; CHECK-NEXT:    ret
+  %sv = load <8 x i1>, <8 x i1>* %svp
+  %c = call <vscale x 8 x i1> @llvm.experimental.vector.insert.v8i1.nxv8i1(<vscale x 8 x i1> %v, <8 x i1> %sv, i64 16)
+  ret <vscale x 8 x i1> %c
+}
+
+declare <8 x i1> @llvm.experimental.vector.insert.v4i1.v8i1(<8 x i1>, <4 x i1>, i64)
+declare <32 x i1> @llvm.experimental.vector.insert.v8i1.v32i1(<32 x i1>, <8 x i1>, i64)
+
 declare <4 x i16> @llvm.experimental.vector.insert.v2i16.v4i16(<4 x i16>, <2 x i16>, i64)
 
 declare <4 x i32> @llvm.experimental.vector.insert.v2i32.v4i32(<4 x i32>, <2 x i32>, i64)
 declare <8 x i32> @llvm.experimental.vector.insert.v2i32.v8i32(<8 x i32>, <2 x i32>, i64)
 
+declare <vscale x 2 x i1> @llvm.experimental.vector.insert.v4i1.nxv2i1(<vscale x 2 x i1>, <4 x i1>, i64)
+declare <vscale x 8 x i1> @llvm.experimental.vector.insert.v8i1.nxv8i1(<vscale x 8 x i1>, <8 x i1>, i64)
+
 declare <vscale x 2 x i16> @llvm.experimental.vector.insert.v2i16.nxv2i16(<vscale x 2 x i16>, <2 x i16>, i64)
 
 declare <vscale x 8 x i32> @llvm.experimental.vector.insert.v2i32.nxv8i32(<vscale x 8 x i32>, <2 x i32>, i64)
 declare <vscale x 8 x i32> @llvm.experimental.vector.insert.v4i32.nxv8i32(<vscale x 8 x i32>, <4 x i32>, i64)
 declare <vscale x 8 x i32> @llvm.experimental.vector.insert.v8i32.nxv8i32(<vscale x 8 x i32>, <8 x i32>, i64)

llvm/test/CodeGen/RISCV/rvv/insert-subvector.ll

 ; CHECK-NEXT:    vsetvli a0, a0, e16,m1,ta,mu
 ; CHECK-NEXT:    vslideup.vx v22, v8, a1
 ; CHECK-NEXT:    vmv8r.v v8, v16
 ; CHECK-NEXT:    ret
   %v = call <vscale x 32 x half> @llvm.experimental.vector.insert.nxv1f16.nxv32f16(<vscale x 32 x half> undef, <vscale x 1 x half> %subvec, i64 26)
   ret <vscale x 32 x half> %v
 }
 
+define <vscale x 32 x i1> @insert_nxv32i1_nxv8i1_0(<vscale x 32 x i1> %v, <vscale x 8 x i1> %sv) {
+; CHECK-LABEL: insert_nxv32i1_nxv8i1_0:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    csrr a0, vlenb
+; CHECK-NEXT:    srli a0, a0, 3
+; CHECK-NEXT:    vsetvli a0, a0, e8,mf2,tu,mu
+; CHECK-NEXT:    vslideup.vi v0, v8, 0
+; CHECK-NEXT:    ret
+  %vec = call <vscale x 32 x i1> @llvm.experimental.vector.insert.nxv8i1.nxv32i1(<vscale x 32 x i1> %v, <vscale x 8 x i1> %sv, i64 0)
+  ret <vscale x 32 x i1> %vec
+}
+
+define <vscale x 32 x i1> @insert_nxv32i1_nxv8i1_8(<vscale x 32 x i1> %v, <vscale x 8 x i1> %sv) {
+; CHECK-LABEL: insert_nxv32i1_nxv8i1_8:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    csrr a0, vlenb
+; CHECK-NEXT:    srli a0, a0, 3
+; CHECK-NEXT:    add a1, a0, a0
+; CHECK-NEXT:    vsetvli a1, a1, e8,mf2,tu,mu
+; CHECK-NEXT:    vslideup.vx v0, v8, a0
+; CHECK-NEXT:    ret
+  %vec = call <vscale x 32 x i1> @llvm.experimental.vector.insert.nxv8i1.nxv32i1(<vscale x 32 x i1> %v, <vscale x 8 x i1> %sv, i64 8)
+  ret <vscale x 32 x i1> %vec
+}
+
+define <vscale x 4 x i1> @insert_nxv4i1_nxv1i1_0(<vscale x 4 x i1> %v, <vscale x 1 x i1> %sv) {
+; CHECK-LABEL: insert_nxv4i1_nxv1i1_0:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8,mf2,ta,mu
+; CHECK-NEXT:    vmv.v.i v25, 0
+; CHECK-NEXT:    vmerge.vim v25, v25, 1, v0
+; CHECK-NEXT:    vsetvli a0, zero, e8,mf8,ta,mu
+; CHECK-NEXT:    vmv.v.i v26, 0
+; CHECK-NEXT:    vmv1r.v v0, v8
+; CHECK-NEXT:    vmerge.vim v26, v26, 1, v0
+; CHECK-NEXT:    csrr a0, vlenb
+; CHECK-NEXT:    srli a0, a0, 3
+; CHECK-NEXT:    vsetvli a0, a0, e8,mf2,tu,mu
+; CHECK-NEXT:    vslideup.vi v25, v26, 0
+; CHECK-NEXT:    vsetvli a0, zero, e8,mf2,ta,mu
+; CHECK-NEXT:    vmsne.vi v0, v25, 0
+; CHECK-NEXT:    ret
+  %vec = call <vscale x 4 x i1> @llvm.experimental.vector.insert.nxv1i1.nxv4i1(<vscale x 4 x i1> %v, <vscale x 1 x i1> %sv, i64 0)
+  ret <vscale x 4 x i1> %vec
+}
+
+define <vscale x 4 x i1> @insert_nxv4i1_nxv1i1_2(<vscale x 4 x i1> %v, <vscale x 1 x i1> %sv) {
+; CHECK-LABEL: insert_nxv4i1_nxv1i1_2:
+; CHECK:       # %bb.0:
+; CHECK-NEXT:    vsetvli a0, zero, e8,mf2,ta,mu
+; CHECK-NEXT:    vmv.v.i v25, 0
+; CHECK-NEXT:    vmerge.vim v25, v25, 1, v0
+; CHECK-NEXT:    csrr a0, vlenb
+; CHECK-NEXT:    srli a0, a0, 3
+; CHECK-NEXT:    slli a1, a0, 1
+; CHECK-NEXT:    add a0, a1, a0
+; CHECK-NEXT:    vsetvli a2, zero, e8,mf8,ta,mu
+; CHECK-NEXT:    vmv.v.i v26, 0
+; CHECK-NEXT:    vmv1r.v v0, v8
+; CHECK-NEXT:    vmerge.vim v26, v26, 1, v0
+; CHECK-NEXT:    vsetvli a0, a0, e8,mf2,tu,mu
+; CHECK-NEXT:    vslideup.vx v25, v26, a1
+; CHECK-NEXT:    vsetvli a0, zero, e8,mf2,ta,mu
+; CHECK-NEXT:    vmsne.vi v0, v25, 0
+; CHECK-NEXT:    ret
+  %vec = call <vscale x 4 x i1> @llvm.experimental.vector.insert.nxv1i1.nxv4i1(<vscale x 4 x i1> %v, <vscale x 1 x i1> %sv, i64 2)
+  ret <vscale x 4 x i1> %vec
+}
+
+declare <vscale x 4 x i1> @llvm.experimental.vector.insert.nxv1i1.nxv4i1(<vscale x 4 x i1>, <vscale x 1 x i1>, i64)
+declare <vscale x 32 x i1> @llvm.experimental.vector.insert.nxv8i1.nxv32i1(<vscale x 32 x i1>, <vscale x 8 x i1>, i64)
+
 declare <vscale x 16 x i8> @llvm.experimental.vector.insert.nxv1i8.nxv16i8(<vscale x 16 x i8>, <vscale x 1 x i8>, i64)
 
 declare <vscale x 32 x half> @llvm.experimental.vector.insert.nxv1f16.nxv32f16(<vscale x 32 x half>, <vscale x 1 x half>, i64)
 declare <vscale x 32 x half> @llvm.experimental.vector.insert.nxv2f16.nxv32f16(<vscale x 32 x half>, <vscale x 2 x half>, i64)
 
 declare <vscale x 4 x i8> @llvm.experimental.vector.insert.nxv1i8.nxv4i8(<vscale x 4 x i8>, <vscale x 1 x i8>, i64 %idx)
 
 declare <vscale x 8 x i32> @llvm.experimental.vector.insert.nxv2i32.nxv8i32(<vscale x 8 x i32>, <vscale x 2 x i32>, i64 %idx)
 declare <vscale x 8 x i32> @llvm.experimental.vector.insert.nxv4i32.nxv8i32(<vscale x 8 x i32>, <vscale x 4 x i32>, i64 %idx)
 
 declare <vscale x 16 x i32> @llvm.experimental.vector.insert.nxv1i32.nxv16i32(<vscale x 16 x i32>, <vscale x 1 x i32>, i64 %idx)
 declare <vscale x 16 x i32> @llvm.experimental.vector.insert.nxv2i32.nxv16i32(<vscale x 16 x i32>, <vscale x 2 x i32>, i64 %idx)
 declare <vscale x 16 x i32> @llvm.experimental.vector.insert.nxv4i32.nxv16i32(<vscale x 16 x i32>, <vscale x 4 x i32>, i64 %idx)
 declare <vscale x 16 x i32> @llvm.experimental.vector.insert.nxv8i32.nxv16i32(<vscale x 16 x i32>, <vscale x 8 x i32>, i64 %idx)