[llvm] [LV][EVL] Support interleaved access with tail folding by EVL (PR #152070)

[Mel Chen via llvm-commits]

https://github.com/Mel-Chen updated https://github.com/llvm/llvm-project/pull/152070

>From 1f680a55ae507072c212b1663bcafaeeaad59690 Mon Sep 17 00:00:00 2001
From: Mel Chen <mel.chen at sifive.com>
Date: Mon, 21 Jul 2025 08:06:02 -0700
Subject: [PATCH 1/3] Support EVL interleave access

---
 .../Transforms/Vectorize/LoopVectorize.cpp    |   4 +-
 llvm/lib/Transforms/Vectorize/VPlan.h         | 172 ++++++++++++---
 .../Transforms/Vectorize/VPlanAnalysis.cpp    |   2 +-
 .../lib/Transforms/Vectorize/VPlanRecipes.cpp | 162 +++++++++++++-
 .../Transforms/Vectorize/VPlanTransforms.cpp  |  17 +-
 llvm/lib/Transforms/Vectorize/VPlanValue.h    |   5 +-
 .../Transforms/Vectorize/VPlanVerifier.cpp    |   3 +-
 .../LoopVectorize/RISCV/dead-ops-cost.ll      |   7 +-
 .../RISCV/interleaved-accesses.ll             | 208 ++++--------------
 .../RISCV/interleaved-masked-access.ll        |  90 ++++----
 .../RISCV/tail-folding-interleave.ll          |  22 +-
 11 files changed, 402 insertions(+), 290 deletions(-)

diff --git a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
index d25add7c4be1f..9875783f670af 100644
--- a/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
+++ b/llvm/lib/Transforms/Vectorize/LoopVectorize.cpp
@@ -4170,6 +4170,7 @@ static bool willGenerateVectors(VPlan &Plan, ElementCount VF,
       case VPDef::VPWidenIntOrFpInductionSC:
       case VPDef::VPWidenPointerInductionSC:
       case VPDef::VPReductionPHISC:
+      case VPDef::VPInterleaveEVLSC:
       case VPDef::VPInterleaveSC:
       case VPDef::VPWidenLoadEVLSC:
       case VPDef::VPWidenLoadSC:
@@ -4198,8 +4199,7 @@ static bool willGenerateVectors(VPlan &Plan, ElementCount VF,
 
       // If no def nor is a store, e.g., branches, continue - no value to check.
       if (R.getNumDefinedValues() == 0 &&
-          !isa<VPWidenStoreRecipe, VPWidenStoreEVLRecipe, VPInterleaveRecipe>(
-              &R))
+          !isa<VPWidenStoreRecipe, VPWidenStoreEVLRecipe, VPInterleaveBase>(&R))
         continue;
       // For multi-def recipes, currently only interleaved loads, suffice to
       // check first def only.
diff --git a/llvm/lib/Transforms/Vectorize/VPlan.h b/llvm/lib/Transforms/Vectorize/VPlan.h
index 5c3edee13020f..e297a38ceddda 100644
--- a/llvm/lib/Transforms/Vectorize/VPlan.h
+++ b/llvm/lib/Transforms/Vectorize/VPlan.h
@@ -557,6 +557,7 @@ class VPSingleDefRecipe : public VPRecipeBase, public VPValue {
     case VPRecipeBase::VPPartialReductionSC:
       return true;
     case VPRecipeBase::VPBranchOnMaskSC:
+    case VPRecipeBase::VPInterleaveEVLSC:
     case VPRecipeBase::VPInterleaveSC:
     case VPRecipeBase::VPIRInstructionSC:
     case VPRecipeBase::VPWidenLoadEVLSC:
@@ -2385,11 +2386,14 @@ class LLVM_ABI_FOR_TEST VPBlendRecipe : public VPSingleDefRecipe {
   }
 };
 
-/// VPInterleaveRecipe is a recipe for transforming an interleave group of load
-/// or stores into one wide load/store and shuffles. The first operand of a
-/// VPInterleave recipe is the address, followed by the stored values, followed
-/// by an optional mask.
-class LLVM_ABI_FOR_TEST VPInterleaveRecipe : public VPRecipeBase {
+/// A common base class for interleaved memory operations.
+/// Interleaved memory operation is a memory access method that combines
+/// multiple strided loads/stores into a single wide load/store with shuffles.
+/// The first operand must be the address. The optional operands are, in order,
+/// the stored values and the mask.
+/// TODO: Inherit from VPIRMetadata
+class LLVM_ABI_FOR_TEST VPInterleaveBase : public VPRecipeBase {
+protected:
   const InterleaveGroup<Instruction> *IG;
 
   /// Indicates if the interleave group is in a conditional block and requires a
@@ -2400,14 +2404,13 @@ class LLVM_ABI_FOR_TEST VPInterleaveRecipe : public VPRecipeBase {
   /// unusued gaps can be loaded speculatively.
   bool NeedsMaskForGaps = false;
 
-public:
-  VPInterleaveRecipe(const InterleaveGroup<Instruction> *IG, VPValue *Addr,
-                     ArrayRef<VPValue *> StoredValues, VPValue *Mask,
-                     bool NeedsMaskForGaps, DebugLoc DL)
-      : VPRecipeBase(VPDef::VPInterleaveSC, {Addr},
-                     DL),
-
-        IG(IG), NeedsMaskForGaps(NeedsMaskForGaps) {
+  VPInterleaveBase(const unsigned char SC,
+                   const InterleaveGroup<Instruction> *IG,
+                   ArrayRef<VPValue *> Operands,
+                   ArrayRef<VPValue *> StoredValues, VPValue *Mask,
+                   bool NeedsMaskForGaps, DebugLoc DL)
+      : VPRecipeBase(SC, Operands, DL), IG(IG),
+        NeedsMaskForGaps(NeedsMaskForGaps) {
     // TODO: extend the masked interleaved-group support to reversed access.
     assert((!Mask || !IG->isReverse()) &&
            "Reversed masked interleave-group not supported.");
@@ -2420,70 +2423,169 @@ class LLVM_ABI_FOR_TEST VPInterleaveRecipe : public VPRecipeBase {
 
     for (auto *SV : StoredValues)
       addOperand(SV);
+
     if (Mask) {
       HasMask = true;
       addOperand(Mask);
     }
   }
-  ~VPInterleaveRecipe() override = default;
 
-  VPInterleaveRecipe *clone() override {
-    return new VPInterleaveRecipe(IG, getAddr(), getStoredValues(), getMask(),
-                                  NeedsMaskForGaps, getDebugLoc());
+public:
+  VPInterleaveBase *clone() override {
+    llvm_unreachable("cloning not supported");
   }
 
-  VP_CLASSOF_IMPL(VPDef::VPInterleaveSC)
+  static inline bool classof(const VPRecipeBase *R) {
+    return R->getVPDefID() == VPRecipeBase::VPInterleaveSC ||
+           R->getVPDefID() == VPRecipeBase::VPInterleaveEVLSC;
+  }
+
+  static inline bool classof(const VPUser *U) {
+    auto *R = dyn_cast<VPRecipeBase>(U);
+    return R && classof(R);
+  }
 
   /// Return the address accessed by this recipe.
   VPValue *getAddr() const {
     return getOperand(0); // Address is the 1st, mandatory operand.
   }
 
+  /// Return true if the access needs a mask because of the gaps.
+  bool needsMaskForGaps() const { return NeedsMaskForGaps; }
+
   /// Return the mask used by this recipe. Note that a full mask is represented
   /// by a nullptr.
   VPValue *getMask() const {
-    // Mask is optional and therefore the last, currently 2nd operand.
+    // Mask is optional and the last operand.
     return HasMask ? getOperand(getNumOperands() - 1) : nullptr;
   }
 
+  const InterleaveGroup<Instruction> *getInterleaveGroup() { return IG; }
+
+  Instruction *getInsertPos() const { return IG->getInsertPos(); }
+
+  void execute(VPTransformState &State) override {
+    llvm_unreachable("VPInterleaveBase should not be instantiated.");
+  }
+
+  /// Return the cost of this VPInterleaveRecipe.
+  InstructionCost computeCost(ElementCount VF,
+                              VPCostContext &Ctx) const override;
+
+  /// Returns true if the recipe only uses the first lane of operand \p Op.
+  virtual bool onlyFirstLaneUsed(const VPValue *Op) const = 0;
+
+  /// Returns the number of stored operands of this interleave group. Returns 0
+  /// for load interleave groups.
+  virtual unsigned getNumStoreOperands() const = 0;
+
   /// Return the VPValues stored by this interleave group. If it is a load
   /// interleave group, return an empty ArrayRef.
-  ArrayRef<VPValue *> getStoredValues() const {
-    // The first operand is the address, followed by the stored values, followed
-    // by an optional mask.
-    return ArrayRef<VPValue *>(op_begin(), getNumOperands())
-        .slice(1, getNumStoreOperands());
+  virtual ArrayRef<VPValue *> getStoredValues() const = 0;
+};
+
+/// VPInterleaveRecipe is a recipe for transforming an interleave group of load
+/// or stores into one wide load/store and shuffles. The first operand of a
+/// VPInterleave recipe is the address, followed by the stored values, followed
+/// by an optional mask.
+class LLVM_ABI_FOR_TEST VPInterleaveRecipe final : public VPInterleaveBase {
+public:
+  VPInterleaveRecipe(const InterleaveGroup<Instruction> *IG, VPValue *Addr,
+                     ArrayRef<VPValue *> StoredValues, VPValue *Mask,
+                     bool NeedsMaskForGaps, DebugLoc DL)
+      : VPInterleaveBase(VPDef::VPInterleaveSC, IG, ArrayRef<VPValue *>({Addr}),
+                         StoredValues, Mask, NeedsMaskForGaps, DL) {}
+
+  ~VPInterleaveRecipe() override = default;
+
+  VPInterleaveRecipe *clone() override {
+    return new VPInterleaveRecipe(IG, getAddr(), getStoredValues(), getMask(),
+                                  NeedsMaskForGaps, getDebugLoc());
   }
 
+  VP_CLASSOF_IMPL(VPDef::VPInterleaveSC)
+
   /// Generate the wide load or store, and shuffles.
   void execute(VPTransformState &State) override;
 
-  /// Return the cost of this VPInterleaveRecipe.
-  InstructionCost computeCost(ElementCount VF,
-                              VPCostContext &Ctx) const override;
-
 #if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
   /// Print the recipe.
   void print(raw_ostream &O, const Twine &Indent,
              VPSlotTracker &SlotTracker) const override;
 #endif
 
-  const InterleaveGroup<Instruction> *getInterleaveGroup() { return IG; }
+  bool onlyFirstLaneUsed(const VPValue *Op) const override {
+    assert(is_contained(operands(), Op) &&
+           "Op must be an operand of the recipe");
+    return Op == getAddr() && !llvm::is_contained(getStoredValues(), Op);
+  }
 
-  /// Returns the number of stored operands of this interleave group. Returns 0
-  /// for load interleave groups.
-  unsigned getNumStoreOperands() const {
+  unsigned getNumStoreOperands() const override {
     return getNumOperands() - (HasMask ? 2 : 1);
   }
 
-  /// The recipe only uses the first lane of the address.
+  ArrayRef<VPValue *> getStoredValues() const override {
+    // The first operand is the address, followed by the stored values, followed
+    // by an optional mask.
+    return ArrayRef<VPValue *>(op_begin(), getNumOperands())
+        .slice(1, getNumStoreOperands());
+  }
+};
+
+/// A recipe for interleaved access operations with vector-predication
+/// intrinsics. The first operand is the address, the second operand is the
+/// explicit vector length . Stored values and mask are optional operands.
+class LLVM_ABI_FOR_TEST VPInterleaveEVLRecipe final : public VPInterleaveBase {
+public:
+  VPInterleaveEVLRecipe(VPInterleaveRecipe &R, VPValue &EVL, VPValue *Mask)
+      : VPInterleaveBase(VPDef::VPInterleaveEVLSC, R.getInterleaveGroup(),
+                         ArrayRef<VPValue *>({R.getAddr(), &EVL}),
+                         R.getStoredValues(), Mask, R.needsMaskForGaps(),
+                         R.getDebugLoc()) {
+    assert(!IG->isReverse() &&
+           "Reversed interleave-group with tail folding is not supported.");
+    assert(!needsMaskForGaps() && "Interleaved access with gap mask is not "
+                                  "supported for scalable vector.");
+  }
+
+  ~VPInterleaveEVLRecipe() override = default;
+
+  VPInterleaveEVLRecipe *clone() override {
+    llvm_unreachable("cloning not implemented yet");
+  }
+
+  VP_CLASSOF_IMPL(VPDef::VPInterleaveEVLSC)
+
+  /// The VPValue of the explicit vector length.
+  VPValue *getEVL() const { return getOperand(1); }
+
+  /// Generate the wide load or store, and shuffles.
+  void execute(VPTransformState &State) override;
+
+#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
+  /// Print the recipe.
+  void print(raw_ostream &O, const Twine &Indent,
+             VPSlotTracker &SlotTracker) const override;
+#endif
+
+  /// The recipe only uses the first lane of the address, and EVL operand.
   bool onlyFirstLaneUsed(const VPValue *Op) const override {
     assert(is_contained(operands(), Op) &&
            "Op must be an operand of the recipe");
-    return Op == getAddr() && !llvm::is_contained(getStoredValues(), Op);
+    return (Op == getAddr() && !llvm::is_contained(getStoredValues(), Op)) ||
+           Op == getEVL();
   }
 
-  Instruction *getInsertPos() const { return IG->getInsertPos(); }
+  unsigned getNumStoreOperands() const override {
+    return getNumOperands() - (HasMask ? 3 : 2);
+  }
+
+  ArrayRef<VPValue *> getStoredValues() const override {
+    // The first operand is the address, and the second operand is EVL, followed
+    // by the stored values, followe by an optional mask.
+    return ArrayRef<VPValue *>(op_begin(), getNumOperands())
+        .slice(2, getNumStoreOperands());
+  }
 };
 
 /// A recipe to represent inloop reduction operations, performing a reduction on
diff --git a/llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp b/llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp
index 4c4adebc64ff4..b5efe9b560c7e 100644
--- a/llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp
+++ b/llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp
@@ -297,7 +297,7 @@ Type *VPTypeAnalysis::inferScalarType(const VPValue *V) {
           .Case<VPBlendRecipe, VPInstruction, VPWidenRecipe, VPReplicateRecipe,
                 VPWidenCallRecipe, VPWidenMemoryRecipe, VPWidenSelectRecipe>(
               [this](const auto *R) { return inferScalarTypeForRecipe(R); })
-          .Case<VPInterleaveRecipe>([V](const VPInterleaveRecipe *R) {
+          .Case<VPInterleaveRecipe, VPInterleaveEVLRecipe>([V](const auto *R) {
             // TODO: Use info from interleave group.
             return V->getUnderlyingValue()->getType();
           })
diff --git a/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp b/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
index ada25acc3636a..5d7e56b38d063 100644
--- a/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
+++ b/llvm/lib/Transforms/Vectorize/VPlanRecipes.cpp
@@ -53,8 +53,9 @@ bool VPRecipeBase::mayWriteToMemory() const {
     return cast<VPExpressionRecipe>(this)->mayReadOrWriteMemory();
   case VPInstructionSC:
     return cast<VPInstruction>(this)->opcodeMayReadOrWriteFromMemory();
+  case VPInterleaveEVLSC:
   case VPInterleaveSC:
-    return cast<VPInterleaveRecipe>(this)->getNumStoreOperands() > 0;
+    return cast<VPInterleaveBase>(this)->getNumStoreOperands() > 0;
   case VPWidenStoreEVLSC:
   case VPWidenStoreSC:
     return true;
@@ -108,6 +109,9 @@ bool VPRecipeBase::mayReadFromMemory() const {
   case VPWidenLoadEVLSC:
   case VPWidenLoadSC:
     return true;
+  case VPInterleaveEVLSC:
+  case VPInterleaveSC:
+    return cast<VPInterleaveBase>(this)->getNumStoreOperands() == 0;
   case VPReplicateSC:
     return cast<Instruction>(getVPSingleValue()->getUnderlyingValue())
         ->mayReadFromMemory();
@@ -184,6 +188,7 @@ bool VPRecipeBase::mayHaveSideEffects() const {
            "underlying instruction has side-effects");
     return false;
   }
+  case VPInterleaveEVLSC:
   case VPInterleaveSC:
     return mayWriteToMemory();
   case VPWidenLoadEVLSC:
@@ -256,7 +261,7 @@ InstructionCost VPRecipeBase::cost(ElementCount VF, VPCostContext &Ctx) {
   Instruction *UI = nullptr;
   if (auto *S = dyn_cast<VPSingleDefRecipe>(this))
     UI = dyn_cast_or_null<Instruction>(S->getUnderlyingValue());
-  else if (auto *IG = dyn_cast<VPInterleaveRecipe>(this))
+  else if (auto *IG = dyn_cast<VPInterleaveBase>(this))
     UI = IG->getInsertPos();
   else if (auto *WidenMem = dyn_cast<VPWidenMemoryRecipe>(this))
     UI = &WidenMem->getIngredient();
@@ -2110,7 +2115,7 @@ InstructionCost VPWidenCastRecipe::computeCost(ElementCount VF,
   auto ComputeCCH = [&](const VPRecipeBase *R) -> TTI::CastContextHint {
     if (VF.isScalar())
       return TTI::CastContextHint::Normal;
-    if (isa<VPInterleaveRecipe>(R))
+    if (isa<VPInterleaveBase>(R))
       return TTI::CastContextHint::Interleave;
     if (const auto *ReplicateRecipe = dyn_cast<VPReplicateRecipe>(R))
       return ReplicateRecipe->isPredicated() ? TTI::CastContextHint::Masked
@@ -3646,8 +3651,155 @@ void VPInterleaveRecipe::print(raw_ostream &O, const Twine &Indent,
 }
 #endif
 
-InstructionCost VPInterleaveRecipe::computeCost(ElementCount VF,
-                                                VPCostContext &Ctx) const {
+void VPInterleaveEVLRecipe::execute(VPTransformState &State) {
+  assert(!State.Lane && "Interleave group being replicated.");
+  assert(State.VF.isScalable() &&
+         "Only support scalable VF for EVL tail-folding.");
+  assert(!NeedsMaskForGaps &&
+         "Masking gaps for scalable vectors is not yet supported.");
+  const InterleaveGroup<Instruction> *Group = IG;
+  Instruction *Instr = Group->getInsertPos();
+
+  // Prepare for the vector type of the interleaved load/store.
+  Type *ScalarTy = getLoadStoreType(Instr);
+  unsigned InterleaveFactor = Group->getFactor();
+  assert(InterleaveFactor <= 8 &&
+         "Unsupported deinterleave/interleave factor for scalable vectors");
+  ElementCount WideVF = State.VF * InterleaveFactor;
+  auto *VecTy = VectorType::get(ScalarTy, WideVF);
+
+  VPValue *BlockInMask = getMask();
+  VPValue *Addr = getAddr();
+  Value *ResAddr = State.get(Addr, VPLane(0));
+  Value *EVL = State.get(getEVL(), VPLane(0));
+
+  auto CreateGroupMask = [&BlockInMask, &State,
+                          &InterleaveFactor]() -> Value * {
+    auto *ResBlockInMask = State.get(BlockInMask);
+    SmallVector<Value *> Ops(InterleaveFactor, ResBlockInMask);
+    return interleaveVectors(State.Builder, Ops, "interleaved.mask");
+  };
+
+  Value *GroupMask = nullptr;
+  if (BlockInMask)
+    GroupMask = CreateGroupMask();
+  else
+    GroupMask =
+        State.Builder.CreateVectorSplat(WideVF, State.Builder.getTrue());
+
+  const DataLayout &DL = Instr->getDataLayout();
+  // Vectorize the interleaved load group.
+  if (isa<LoadInst>(Instr)) {
+    CallInst *NewLoad = State.Builder.CreateIntrinsic(VecTy, Intrinsic::vp_load,
+                                                      {ResAddr, GroupMask, EVL},
+                                                      nullptr, "wide.vp.load");
+    NewLoad->addParamAttr(0, Attribute::getWithAlignment(NewLoad->getContext(),
+                                                         Group->getAlign()));
+
+    Group->addMetadata(NewLoad);
+
+    ArrayRef<VPValue *> VPDefs = definedValues();
+    // Scalable vectors cannot use arbitrary shufflevectors (only splats),
+    // so must use intrinsics to deinterleave.
+    NewLoad = State.Builder.CreateIntrinsic(
+        Intrinsic::getDeinterleaveIntrinsicID(InterleaveFactor),
+        NewLoad->getType(), NewLoad,
+        /*FMFSource=*/nullptr, "strided.vec");
+
+    for (unsigned I = 0, J = 0; I < InterleaveFactor; ++I) {
+      Instruction *Member = Group->getMember(I);
+
+      // Skip the gaps in the group.
+      if (!Member)
+        continue;
+
+      Value *StridedVec = State.Builder.CreateExtractValue(NewLoad, I);
+
+      // If this member has different type, cast the result type.
+      if (Member->getType() != ScalarTy) {
+        VectorType *OtherVTy = VectorType::get(Member->getType(), State.VF);
+        StridedVec =
+            createBitOrPointerCast(State.Builder, StridedVec, OtherVTy, DL);
+      }
+
+      State.set(VPDefs[J], StridedVec);
+      ++J;
+    }
+    return;
+  }
+
+  // The sub vector type for current instruction.
+  auto *SubVT = VectorType::get(ScalarTy, State.VF);
+
+  // Vectorize the interleaved store group.
+  ArrayRef<VPValue *> StoredValues = getStoredValues();
+  // Collect the stored vector from each member.
+  SmallVector<Value *, 4> StoredVecs;
+  unsigned StoredIdx = 0;
+  for (unsigned I = 0; I < InterleaveFactor; I++) {
+    Instruction *Member = Group->getMember(I);
+
+    // Skip the gaps in the group.
+    if (!Member) {
+      Value *Undef = PoisonValue::get(SubVT);
+      StoredVecs.push_back(Undef);
+      continue;
+    }
+
+    Value *StoredVec = State.get(StoredValues[StoredIdx]);
+    ++StoredIdx;
+
+    // If this member has different type, cast it to a unified type.
+    if (StoredVec->getType() != SubVT)
+      StoredVec = createBitOrPointerCast(State.Builder, StoredVec, SubVT, DL);
+
+    StoredVecs.push_back(StoredVec);
+  }
+
+  // Interleave all the smaller vectors into one wider vector.
+  Value *IVec = interleaveVectors(State.Builder, StoredVecs, "interleaved.vec");
+  CallInst *NewStore = State.Builder.CreateIntrinsic(
+      Type::getVoidTy(EVL->getContext()), Intrinsic::vp_store,
+      {IVec, ResAddr, GroupMask, EVL});
+  NewStore->addParamAttr(1, Attribute::getWithAlignment(NewStore->getContext(),
+                                                        Group->getAlign()));
+
+  Group->addMetadata(NewStore);
+}
+#if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
+void VPInterleaveEVLRecipe::print(raw_ostream &O, const Twine &Indent,
+                                  VPSlotTracker &SlotTracker) const {
+  O << Indent << "INTERLEAVE-GROUP with factor " << IG->getFactor() << " at ";
+  IG->getInsertPos()->printAsOperand(O, false);
+  O << ", ";
+  getAddr()->printAsOperand(O, SlotTracker);
+  O << ", ";
+  getEVL()->printAsOperand(O, SlotTracker);
+  if (VPValue *Mask = getMask()) {
+    O << ", ";
+    Mask->printAsOperand(O, SlotTracker);
+  }
+
+  unsigned OpIdx = 0;
+  for (unsigned i = 0; i < IG->getFactor(); ++i) {
+    if (!IG->getMember(i))
+      continue;
+    if (getNumStoreOperands() > 0) {
+      O << "\n" << Indent << "  vp.store ";
+      getOperand(2 + OpIdx)->printAsOperand(O, SlotTracker);
+      O << " to index " << i;
+    } else {
+      O << "\n" << Indent << "  ";
+      getVPValue(OpIdx)->printAsOperand(O, SlotTracker);
+      O << " = vp.load from index " << i;
+    }
+    ++OpIdx;
+  }
+}
+#endif
+
+InstructionCost VPInterleaveBase::computeCost(ElementCount VF,
+                                              VPCostContext &Ctx) const {
   Instruction *InsertPos = getInsertPos();
   // Find the VPValue index of the interleave group. We need to skip gaps.
   unsigned InsertPosIdx = 0;
diff --git a/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp b/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp
index c4a6994276274..19644338997db 100644
--- a/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp
+++ b/llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp
@@ -2148,6 +2148,10 @@ static VPRecipeBase *optimizeMaskToEVL(VPValue *HeaderMask,
         VPValue *NewMask = GetNewMask(S->getMask());
         return new VPWidenStoreEVLRecipe(*S, EVL, NewMask);
       })
+      .Case<VPInterleaveRecipe>([&](VPInterleaveRecipe *IR) {
+        VPValue *NewMask = GetNewMask(IR->getMask());
+        return new VPInterleaveEVLRecipe(*IR, EVL, NewMask);
+      })
       .Case<VPReductionRecipe>([&](VPReductionRecipe *Red) {
         VPValue *NewMask = GetNewMask(Red->getCondOp());
         return new VPReductionEVLRecipe(*Red, EVL, NewMask);
@@ -2269,16 +2273,17 @@ static void transformRecipestoEVLRecipes(VPlan &Plan, VPValue &EVL) {
     if (!EVLRecipe)
       continue;
 
-    [[maybe_unused]] unsigned NumDefVal = EVLRecipe->getNumDefinedValues();
+    unsigned NumDefVal = EVLRecipe->getNumDefinedValues();
     assert(NumDefVal == CurRecipe->getNumDefinedValues() &&
            "New recipe must define the same number of values as the "
            "original.");
-    assert(NumDefVal <= 1 &&
-           "Only supports recipes with a single definition or without users.");
     EVLRecipe->insertBefore(CurRecipe);
-    if (isa<VPSingleDefRecipe, VPWidenLoadEVLRecipe>(EVLRecipe)) {
-      VPValue *CurVPV = CurRecipe->getVPSingleValue();
-      CurVPV->replaceAllUsesWith(EVLRecipe->getVPSingleValue());
+    if (isa<VPSingleDefRecipe, VPWidenLoadEVLRecipe, VPInterleaveEVLRecipe>(
+            EVLRecipe)) {
+      for (unsigned I = 0; I < NumDefVal; ++I) {
+        VPValue *CurVPV = CurRecipe->getVPValue(I);
+        CurVPV->replaceAllUsesWith(EVLRecipe->getVPValue(I));
+      }
     }
     ToErase.push_back(CurRecipe);
   }
diff --git a/llvm/lib/Transforms/Vectorize/VPlanValue.h b/llvm/lib/Transforms/Vectorize/VPlanValue.h
index 24f6d61512ef6..85c6c2c8d7965 100644
--- a/llvm/lib/Transforms/Vectorize/VPlanValue.h
+++ b/llvm/lib/Transforms/Vectorize/VPlanValue.h
@@ -38,7 +38,7 @@ struct VPDoubleValueDef;
 class VPSlotTracker;
 class VPUser;
 class VPRecipeBase;
-class VPInterleaveRecipe;
+class VPInterleaveBase;
 class VPPhiAccessors;
 
 // This is the base class of the VPlan Def/Use graph, used for modeling the data
@@ -48,7 +48,7 @@ class VPPhiAccessors;
 class LLVM_ABI_FOR_TEST VPValue {
   friend class VPDef;
   friend struct VPDoubleValueDef;
-  friend class VPInterleaveRecipe;
+  friend class VPInterleaveBase;
   friend class VPlan;
   friend class VPExpressionRecipe;
 
@@ -335,6 +335,7 @@ class VPDef {
     VPExpressionSC,
     VPIRInstructionSC,
     VPInstructionSC,
+    VPInterleaveEVLSC,
     VPInterleaveSC,
     VPReductionEVLSC,
     VPReductionSC,
diff --git a/llvm/lib/Transforms/Vectorize/VPlanVerifier.cpp b/llvm/lib/Transforms/Vectorize/VPlanVerifier.cpp
index 3417e1c8dc1ea..4c0dc9bb08046 100644
--- a/llvm/lib/Transforms/Vectorize/VPlanVerifier.cpp
+++ b/llvm/lib/Transforms/Vectorize/VPlanVerifier.cpp
@@ -166,7 +166,8 @@ bool VPlanVerifier::verifyEVLRecipe(const VPInstruction &EVL) const {
           }
           return VerifyEVLUse(*R, 2);
         })
-        .Case<VPWidenLoadEVLRecipe, VPVectorEndPointerRecipe>(
+        .Case<VPWidenLoadEVLRecipe, VPVectorEndPointerRecipe,
+              VPInterleaveEVLRecipe>(
             [&](const VPRecipeBase *R) { return VerifyEVLUse(*R, 1); })
         .Case<VPInstructionWithType>(
             [&](const VPInstructionWithType *S) { return VerifyEVLUse(*S, 0); })
diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/dead-ops-cost.ll b/llvm/test/Transforms/LoopVectorize/RISCV/dead-ops-cost.ll
index 045e816e6ce1e..a65cafaf0fbf7 100644
--- a/llvm/test/Transforms/LoopVectorize/RISCV/dead-ops-cost.ll
+++ b/llvm/test/Transforms/LoopVectorize/RISCV/dead-ops-cost.ll
@@ -407,18 +407,13 @@ define void @gather_interleave_group_with_dead_insert_pos(i64 %N, ptr noalias %s
 ; CHECK-NEXT:    [[VEC_IND:%.*]] = phi <vscale x 4 x i64> [ [[INDUCTION]], %[[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], %[[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[AVL:%.*]] = phi i64 [ [[TMP2]], %[[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], %[[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[TMP10:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 4, i1 true)
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[TMP10]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT2:%.*]] = shufflevector <vscale x 4 x i32> [[BROADCAST_SPLATINSERT1]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
 ; CHECK-NEXT:    [[TMP16:%.*]] = zext i32 [[TMP10]] to i64
 ; CHECK-NEXT:    [[TMP12:%.*]] = mul i64 2, [[TMP16]]
 ; CHECK-NEXT:    [[DOTSPLATINSERT:%.*]] = insertelement <vscale x 4 x i64> poison, i64 [[TMP12]], i64 0
 ; CHECK-NEXT:    [[DOTSPLAT:%.*]] = shufflevector <vscale x 4 x i64> [[DOTSPLATINSERT]], <vscale x 4 x i64> poison, <vscale x 4 x i32> zeroinitializer
-; CHECK-NEXT:    [[TMP13:%.*]] = call <vscale x 4 x i32> @llvm.stepvector.nxv4i32()
-; CHECK-NEXT:    [[TMP14:%.*]] = icmp ult <vscale x 4 x i32> [[TMP13]], [[BROADCAST_SPLAT2]]
 ; CHECK-NEXT:    [[OFFSET_IDX:%.*]] = mul i64 [[EVL_BASED_IV]], 2
 ; CHECK-NEXT:    [[TMP22:%.*]] = getelementptr i8, ptr [[SRC]], i64 [[OFFSET_IDX]]
-; CHECK-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave2.nxv8i1(<vscale x 4 x i1> [[TMP14]], <vscale x 4 x i1> [[TMP14]])
-; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[TMP22]], i32 1, <vscale x 8 x i1> [[INTERLEAVED_MASK]], <vscale x 8 x i8> poison)
+; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i8> @llvm.vp.load.nxv8i8.p0(ptr align 1 [[TMP22]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP10]])
 ; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 4 x i8>, <vscale x 4 x i8> } @llvm.vector.deinterleave2.nxv8i8(<vscale x 8 x i8> [[WIDE_MASKED_VEC]])
 ; CHECK-NEXT:    [[TMP23:%.*]] = extractvalue { <vscale x 4 x i8>, <vscale x 4 x i8> } [[STRIDED_VEC]], 0
 ; CHECK-NEXT:    [[TMP17:%.*]] = extractvalue { <vscale x 4 x i8>, <vscale x 4 x i8> } [[STRIDED_VEC]], 1
diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/interleaved-accesses.ll b/llvm/test/Transforms/LoopVectorize/RISCV/interleaved-accesses.ll
index a2997b3c6652f..9b38973fc6b2a 100644
--- a/llvm/test/Transforms/LoopVectorize/RISCV/interleaved-accesses.ll
+++ b/llvm/test/Transforms/LoopVectorize/RISCV/interleaved-accesses.ll
@@ -15,22 +15,16 @@ define void @load_store_factor2_i32(ptr %p) {
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 4, i1 true)
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[TMP7]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-; CHECK-NEXT:    [[TMP6:%.*]] = call <vscale x 4 x i32> @llvm.stepvector.nxv4i32()
-; CHECK-NEXT:    [[TMP13:%.*]] = icmp ult <vscale x 4 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; CHECK-NEXT:    [[TMP14:%.*]] = shl i64 [[INDEX]], 1
 ; CHECK-NEXT:    [[TMP15:%.*]] = getelementptr i32, ptr [[P:%.*]], i64 [[TMP14]]
-; CHECK-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave2.nxv8i1(<vscale x 4 x i1> [[TMP13]], <vscale x 4 x i1> [[TMP13]])
-; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i32> @llvm.masked.load.nxv8i32.p0(ptr [[TMP15]], i32 4, <vscale x 8 x i1> [[INTERLEAVED_MASK]], <vscale x 8 x i32> poison)
+; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i32> @llvm.vp.load.nxv8i32.p0(ptr align 4 [[TMP15]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.vector.deinterleave2.nxv8i32(<vscale x 8 x i32> [[WIDE_MASKED_VEC]])
 ; CHECK-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 0
 ; CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 1
 ; CHECK-NEXT:    [[TMP10:%.*]] = add <vscale x 4 x i32> [[TMP8]], splat (i32 1)
 ; CHECK-NEXT:    [[TMP11:%.*]] = add <vscale x 4 x i32> [[TMP9]], splat (i32 2)
 ; CHECK-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 8 x i32> @llvm.vector.interleave2.nxv8i32(<vscale x 4 x i32> [[TMP10]], <vscale x 4 x i32> [[TMP11]])
-; CHECK-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave2.nxv8i1(<vscale x 4 x i1> [[TMP13]], <vscale x 4 x i1> [[TMP13]])
-; CHECK-NEXT:    call void @llvm.masked.store.nxv8i32.p0(<vscale x 8 x i32> [[INTERLEAVED_VEC]], ptr [[TMP15]], i32 4, <vscale x 8 x i1> [[INTERLEAVED_MASK1]])
+; CHECK-NEXT:    call void @llvm.vp.store.nxv8i32.p0(<vscale x 8 x i32> [[INTERLEAVED_VEC]], ptr align 4 [[TMP15]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[TMP16:%.*]] = zext i32 [[TMP7]] to i64
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP16]], [[INDEX]]
 ; CHECK-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP16]]
@@ -112,22 +106,16 @@ define void @load_store_factor2_i32(ptr %p) {
 ; SCALABLE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 4, i1 true)
-; SCALABLE-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[TMP7]], i64 0
-; SCALABLE-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-; SCALABLE-NEXT:    [[TMP6:%.*]] = call <vscale x 4 x i32> @llvm.stepvector.nxv4i32()
-; SCALABLE-NEXT:    [[TMP13:%.*]] = icmp ult <vscale x 4 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; SCALABLE-NEXT:    [[TMP14:%.*]] = shl i64 [[INDEX]], 1
 ; SCALABLE-NEXT:    [[TMP15:%.*]] = getelementptr i32, ptr [[P:%.*]], i64 [[TMP14]]
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave2.nxv8i1(<vscale x 4 x i1> [[TMP13]], <vscale x 4 x i1> [[TMP13]])
-; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i32> @llvm.masked.load.nxv8i32.p0(ptr [[TMP15]], i32 4, <vscale x 8 x i1> [[INTERLEAVED_MASK]], <vscale x 8 x i32> poison)
+; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i32> @llvm.vp.load.nxv8i32.p0(ptr align 4 [[TMP15]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.vector.deinterleave2.nxv8i32(<vscale x 8 x i32> [[WIDE_MASKED_VEC]])
 ; SCALABLE-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 0
 ; SCALABLE-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 1
 ; SCALABLE-NEXT:    [[TMP10:%.*]] = add <vscale x 4 x i32> [[TMP8]], splat (i32 1)
 ; SCALABLE-NEXT:    [[TMP11:%.*]] = add <vscale x 4 x i32> [[TMP9]], splat (i32 2)
 ; SCALABLE-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 8 x i32> @llvm.vector.interleave2.nxv8i32(<vscale x 4 x i32> [[TMP10]], <vscale x 4 x i32> [[TMP11]])
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave2.nxv8i1(<vscale x 4 x i1> [[TMP13]], <vscale x 4 x i1> [[TMP13]])
-; SCALABLE-NEXT:    call void @llvm.masked.store.nxv8i32.p0(<vscale x 8 x i32> [[INTERLEAVED_VEC]], ptr [[TMP15]], i32 4, <vscale x 8 x i1> [[INTERLEAVED_MASK1]])
+; SCALABLE-NEXT:    call void @llvm.vp.store.nxv8i32.p0(<vscale x 8 x i32> [[INTERLEAVED_VEC]], ptr align 4 [[TMP15]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[TMP16:%.*]] = zext i32 [[TMP7]] to i64
 ; SCALABLE-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP16]], [[INDEX]]
 ; SCALABLE-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP16]]
@@ -191,22 +179,16 @@ define void @load_store_factor2_i64(ptr %p) {
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 2, i1 true)
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 2 x i32> poison, i32 [[TMP7]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 2 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 2 x i32> poison, <vscale x 2 x i32> zeroinitializer
-; CHECK-NEXT:    [[TMP6:%.*]] = call <vscale x 2 x i32> @llvm.stepvector.nxv2i32()
-; CHECK-NEXT:    [[TMP10:%.*]] = icmp ult <vscale x 2 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; CHECK-NEXT:    [[TMP8:%.*]] = shl i64 [[INDEX]], 1
 ; CHECK-NEXT:    [[TMP14:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP8]]
-; CHECK-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 4 x i1> @llvm.vector.interleave2.nxv4i1(<vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]])
-; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 4 x i64> @llvm.masked.load.nxv4i64.p0(ptr [[TMP14]], i32 8, <vscale x 4 x i1> [[INTERLEAVED_MASK]], <vscale x 4 x i64> poison)
+; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 4 x i64> @llvm.vp.load.nxv4i64.p0(ptr align 8 [[TMP14]], <vscale x 4 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.vector.deinterleave2.nxv4i64(<vscale x 4 x i64> [[WIDE_MASKED_VEC]])
 ; CHECK-NEXT:    [[TMP20:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 0
 ; CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 1
 ; CHECK-NEXT:    [[TMP13:%.*]] = add <vscale x 2 x i64> [[TMP20]], splat (i64 1)
 ; CHECK-NEXT:    [[TMP11:%.*]] = add <vscale x 2 x i64> [[TMP9]], splat (i64 2)
 ; CHECK-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 4 x i64> @llvm.vector.interleave2.nxv4i64(<vscale x 2 x i64> [[TMP13]], <vscale x 2 x i64> [[TMP11]])
-; CHECK-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 4 x i1> @llvm.vector.interleave2.nxv4i1(<vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]])
-; CHECK-NEXT:    call void @llvm.masked.store.nxv4i64.p0(<vscale x 4 x i64> [[INTERLEAVED_VEC]], ptr [[TMP14]], i32 8, <vscale x 4 x i1> [[INTERLEAVED_MASK1]])
+; CHECK-NEXT:    call void @llvm.vp.store.nxv4i64.p0(<vscale x 4 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP14]], <vscale x 4 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[TMP16:%.*]] = zext i32 [[TMP7]] to i64
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP16]], [[INDEX]]
 ; CHECK-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP16]]
@@ -288,22 +270,16 @@ define void @load_store_factor2_i64(ptr %p) {
 ; SCALABLE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 2, i1 true)
-; SCALABLE-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 2 x i32> poison, i32 [[TMP7]], i64 0
-; SCALABLE-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 2 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 2 x i32> poison, <vscale x 2 x i32> zeroinitializer
-; SCALABLE-NEXT:    [[TMP6:%.*]] = call <vscale x 2 x i32> @llvm.stepvector.nxv2i32()
-; SCALABLE-NEXT:    [[TMP10:%.*]] = icmp ult <vscale x 2 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; SCALABLE-NEXT:    [[TMP8:%.*]] = shl i64 [[INDEX]], 1
 ; SCALABLE-NEXT:    [[TMP14:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP8]]
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 4 x i1> @llvm.vector.interleave2.nxv4i1(<vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]])
-; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 4 x i64> @llvm.masked.load.nxv4i64.p0(ptr [[TMP14]], i32 8, <vscale x 4 x i1> [[INTERLEAVED_MASK]], <vscale x 4 x i64> poison)
+; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 4 x i64> @llvm.vp.load.nxv4i64.p0(ptr align 8 [[TMP14]], <vscale x 4 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.vector.deinterleave2.nxv4i64(<vscale x 4 x i64> [[WIDE_MASKED_VEC]])
 ; SCALABLE-NEXT:    [[TMP20:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 0
 ; SCALABLE-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 1
 ; SCALABLE-NEXT:    [[TMP13:%.*]] = add <vscale x 2 x i64> [[TMP20]], splat (i64 1)
 ; SCALABLE-NEXT:    [[TMP11:%.*]] = add <vscale x 2 x i64> [[TMP9]], splat (i64 2)
 ; SCALABLE-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 4 x i64> @llvm.vector.interleave2.nxv4i64(<vscale x 2 x i64> [[TMP13]], <vscale x 2 x i64> [[TMP11]])
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 4 x i1> @llvm.vector.interleave2.nxv4i1(<vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]])
-; SCALABLE-NEXT:    call void @llvm.masked.store.nxv4i64.p0(<vscale x 4 x i64> [[INTERLEAVED_VEC]], ptr [[TMP14]], i32 8, <vscale x 4 x i1> [[INTERLEAVED_MASK1]])
+; SCALABLE-NEXT:    call void @llvm.vp.store.nxv4i64.p0(<vscale x 4 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP14]], <vscale x 4 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[TMP16:%.*]] = zext i32 [[TMP7]] to i64
 ; SCALABLE-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP16]], [[INDEX]]
 ; SCALABLE-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP16]]
@@ -367,14 +343,9 @@ define void @load_store_factor3_i32(ptr %p) {
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 4, i1 true)
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[TMP7]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-; CHECK-NEXT:    [[TMP6:%.*]] = call <vscale x 4 x i32> @llvm.stepvector.nxv4i32()
-; CHECK-NEXT:    [[TMP14:%.*]] = icmp ult <vscale x 4 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; CHECK-NEXT:    [[TMP16:%.*]] = mul i64 [[INDEX]], 3
 ; CHECK-NEXT:    [[TMP17:%.*]] = getelementptr i32, ptr [[P:%.*]], i64 [[TMP16]]
-; CHECK-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 12 x i1> @llvm.vector.interleave3.nxv12i1(<vscale x 4 x i1> [[TMP14]], <vscale x 4 x i1> [[TMP14]], <vscale x 4 x i1> [[TMP14]])
-; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 12 x i32> @llvm.masked.load.nxv12i32.p0(ptr [[TMP17]], i32 4, <vscale x 12 x i1> [[INTERLEAVED_MASK]], <vscale x 12 x i32> poison)
+; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 12 x i32> @llvm.vp.load.nxv12i32.p0(ptr align 4 [[TMP17]], <vscale x 12 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.vector.deinterleave3.nxv12i32(<vscale x 12 x i32> [[WIDE_MASKED_VEC]])
 ; CHECK-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 0
 ; CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 1
@@ -383,8 +354,7 @@ define void @load_store_factor3_i32(ptr %p) {
 ; CHECK-NEXT:    [[TMP12:%.*]] = add <vscale x 4 x i32> [[TMP9]], splat (i32 2)
 ; CHECK-NEXT:    [[TMP13:%.*]] = add <vscale x 4 x i32> [[TMP10]], splat (i32 3)
 ; CHECK-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 12 x i32> @llvm.vector.interleave3.nxv12i32(<vscale x 4 x i32> [[TMP11]], <vscale x 4 x i32> [[TMP12]], <vscale x 4 x i32> [[TMP13]])
-; CHECK-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 12 x i1> @llvm.vector.interleave3.nxv12i1(<vscale x 4 x i1> [[TMP14]], <vscale x 4 x i1> [[TMP14]], <vscale x 4 x i1> [[TMP14]])
-; CHECK-NEXT:    call void @llvm.masked.store.nxv12i32.p0(<vscale x 12 x i32> [[INTERLEAVED_VEC]], ptr [[TMP17]], i32 4, <vscale x 12 x i1> [[INTERLEAVED_MASK1]])
+; CHECK-NEXT:    call void @llvm.vp.store.nxv12i32.p0(<vscale x 12 x i32> [[INTERLEAVED_VEC]], ptr align 4 [[TMP17]], <vscale x 12 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[TMP19:%.*]] = zext i32 [[TMP7]] to i64
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP19]], [[INDEX]]
 ; CHECK-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP19]]
@@ -480,14 +450,9 @@ define void @load_store_factor3_i32(ptr %p) {
 ; SCALABLE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 4, i1 true)
-; SCALABLE-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[TMP7]], i64 0
-; SCALABLE-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-; SCALABLE-NEXT:    [[TMP6:%.*]] = call <vscale x 4 x i32> @llvm.stepvector.nxv4i32()
-; SCALABLE-NEXT:    [[TMP14:%.*]] = icmp ult <vscale x 4 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; SCALABLE-NEXT:    [[TMP16:%.*]] = mul i64 [[INDEX]], 3
 ; SCALABLE-NEXT:    [[TMP17:%.*]] = getelementptr i32, ptr [[P:%.*]], i64 [[TMP16]]
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 12 x i1> @llvm.vector.interleave3.nxv12i1(<vscale x 4 x i1> [[TMP14]], <vscale x 4 x i1> [[TMP14]], <vscale x 4 x i1> [[TMP14]])
-; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 12 x i32> @llvm.masked.load.nxv12i32.p0(ptr [[TMP17]], i32 4, <vscale x 12 x i1> [[INTERLEAVED_MASK]], <vscale x 12 x i32> poison)
+; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 12 x i32> @llvm.vp.load.nxv12i32.p0(ptr align 4 [[TMP17]], <vscale x 12 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.vector.deinterleave3.nxv12i32(<vscale x 12 x i32> [[WIDE_MASKED_VEC]])
 ; SCALABLE-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 0
 ; SCALABLE-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 1
@@ -496,8 +461,7 @@ define void @load_store_factor3_i32(ptr %p) {
 ; SCALABLE-NEXT:    [[TMP12:%.*]] = add <vscale x 4 x i32> [[TMP9]], splat (i32 2)
 ; SCALABLE-NEXT:    [[TMP13:%.*]] = add <vscale x 4 x i32> [[TMP10]], splat (i32 3)
 ; SCALABLE-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 12 x i32> @llvm.vector.interleave3.nxv12i32(<vscale x 4 x i32> [[TMP11]], <vscale x 4 x i32> [[TMP12]], <vscale x 4 x i32> [[TMP13]])
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 12 x i1> @llvm.vector.interleave3.nxv12i1(<vscale x 4 x i1> [[TMP14]], <vscale x 4 x i1> [[TMP14]], <vscale x 4 x i1> [[TMP14]])
-; SCALABLE-NEXT:    call void @llvm.masked.store.nxv12i32.p0(<vscale x 12 x i32> [[INTERLEAVED_VEC]], ptr [[TMP17]], i32 4, <vscale x 12 x i1> [[INTERLEAVED_MASK1]])
+; SCALABLE-NEXT:    call void @llvm.vp.store.nxv12i32.p0(<vscale x 12 x i32> [[INTERLEAVED_VEC]], ptr align 4 [[TMP17]], <vscale x 12 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[TMP19:%.*]] = zext i32 [[TMP7]] to i64
 ; SCALABLE-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP19]], [[INDEX]]
 ; SCALABLE-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP19]]
@@ -572,14 +536,9 @@ define void @load_store_factor3_i64(ptr %p) {
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 2, i1 true)
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 2 x i32> poison, i32 [[TMP7]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 2 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 2 x i32> poison, <vscale x 2 x i32> zeroinitializer
-; CHECK-NEXT:    [[TMP6:%.*]] = call <vscale x 2 x i32> @llvm.stepvector.nxv2i32()
-; CHECK-NEXT:    [[TMP11:%.*]] = icmp ult <vscale x 2 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; CHECK-NEXT:    [[TMP8:%.*]] = mul i64 [[INDEX]], 3
 ; CHECK-NEXT:    [[TMP14:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP8]]
-; CHECK-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 6 x i1> @llvm.vector.interleave3.nxv6i1(<vscale x 2 x i1> [[TMP11]], <vscale x 2 x i1> [[TMP11]], <vscale x 2 x i1> [[TMP11]])
-; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 6 x i64> @llvm.masked.load.nxv6i64.p0(ptr [[TMP14]], i32 8, <vscale x 6 x i1> [[INTERLEAVED_MASK]], <vscale x 6 x i64> poison)
+; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 6 x i64> @llvm.vp.load.nxv6i64.p0(ptr align 8 [[TMP14]], <vscale x 6 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.vector.deinterleave3.nxv6i64(<vscale x 6 x i64> [[WIDE_MASKED_VEC]])
 ; CHECK-NEXT:    [[TMP23:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 0
 ; CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 1
@@ -588,8 +547,7 @@ define void @load_store_factor3_i64(ptr %p) {
 ; CHECK-NEXT:    [[TMP12:%.*]] = add <vscale x 2 x i64> [[TMP9]], splat (i64 2)
 ; CHECK-NEXT:    [[TMP13:%.*]] = add <vscale x 2 x i64> [[TMP10]], splat (i64 3)
 ; CHECK-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 6 x i64> @llvm.vector.interleave3.nxv6i64(<vscale x 2 x i64> [[TMP25]], <vscale x 2 x i64> [[TMP12]], <vscale x 2 x i64> [[TMP13]])
-; CHECK-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 6 x i1> @llvm.vector.interleave3.nxv6i1(<vscale x 2 x i1> [[TMP11]], <vscale x 2 x i1> [[TMP11]], <vscale x 2 x i1> [[TMP11]])
-; CHECK-NEXT:    call void @llvm.masked.store.nxv6i64.p0(<vscale x 6 x i64> [[INTERLEAVED_VEC]], ptr [[TMP14]], i32 8, <vscale x 6 x i1> [[INTERLEAVED_MASK1]])
+; CHECK-NEXT:    call void @llvm.vp.store.nxv6i64.p0(<vscale x 6 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP14]], <vscale x 6 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[TMP19:%.*]] = zext i32 [[TMP7]] to i64
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP19]], [[INDEX]]
 ; CHECK-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP19]]
@@ -685,14 +643,9 @@ define void @load_store_factor3_i64(ptr %p) {
 ; SCALABLE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 2, i1 true)
-; SCALABLE-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 2 x i32> poison, i32 [[TMP7]], i64 0
-; SCALABLE-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 2 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 2 x i32> poison, <vscale x 2 x i32> zeroinitializer
-; SCALABLE-NEXT:    [[TMP6:%.*]] = call <vscale x 2 x i32> @llvm.stepvector.nxv2i32()
-; SCALABLE-NEXT:    [[TMP11:%.*]] = icmp ult <vscale x 2 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; SCALABLE-NEXT:    [[TMP8:%.*]] = mul i64 [[INDEX]], 3
 ; SCALABLE-NEXT:    [[TMP14:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP8]]
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 6 x i1> @llvm.vector.interleave3.nxv6i1(<vscale x 2 x i1> [[TMP11]], <vscale x 2 x i1> [[TMP11]], <vscale x 2 x i1> [[TMP11]])
-; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 6 x i64> @llvm.masked.load.nxv6i64.p0(ptr [[TMP14]], i32 8, <vscale x 6 x i1> [[INTERLEAVED_MASK]], <vscale x 6 x i64> poison)
+; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 6 x i64> @llvm.vp.load.nxv6i64.p0(ptr align 8 [[TMP14]], <vscale x 6 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.vector.deinterleave3.nxv6i64(<vscale x 6 x i64> [[WIDE_MASKED_VEC]])
 ; SCALABLE-NEXT:    [[TMP23:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 0
 ; SCALABLE-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 1
@@ -701,8 +654,7 @@ define void @load_store_factor3_i64(ptr %p) {
 ; SCALABLE-NEXT:    [[TMP12:%.*]] = add <vscale x 2 x i64> [[TMP9]], splat (i64 2)
 ; SCALABLE-NEXT:    [[TMP13:%.*]] = add <vscale x 2 x i64> [[TMP10]], splat (i64 3)
 ; SCALABLE-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 6 x i64> @llvm.vector.interleave3.nxv6i64(<vscale x 2 x i64> [[TMP25]], <vscale x 2 x i64> [[TMP12]], <vscale x 2 x i64> [[TMP13]])
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 6 x i1> @llvm.vector.interleave3.nxv6i1(<vscale x 2 x i1> [[TMP11]], <vscale x 2 x i1> [[TMP11]], <vscale x 2 x i1> [[TMP11]])
-; SCALABLE-NEXT:    call void @llvm.masked.store.nxv6i64.p0(<vscale x 6 x i64> [[INTERLEAVED_VEC]], ptr [[TMP14]], i32 8, <vscale x 6 x i1> [[INTERLEAVED_MASK1]])
+; SCALABLE-NEXT:    call void @llvm.vp.store.nxv6i64.p0(<vscale x 6 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP14]], <vscale x 6 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[TMP19:%.*]] = zext i32 [[TMP7]] to i64
 ; SCALABLE-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP19]], [[INDEX]]
 ; SCALABLE-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP19]]
@@ -777,14 +729,9 @@ define void @load_store_factor4(ptr %p) {
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 2, i1 true)
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 2 x i32> poison, i32 [[TMP7]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 2 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 2 x i32> poison, <vscale x 2 x i32> zeroinitializer
-; CHECK-NEXT:    [[TMP6:%.*]] = call <vscale x 2 x i32> @llvm.stepvector.nxv2i32()
-; CHECK-NEXT:    [[TMP10:%.*]] = icmp ult <vscale x 2 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; CHECK-NEXT:    [[TMP8:%.*]] = mul i64 [[INDEX]], 4
 ; CHECK-NEXT:    [[TMP9:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP8]]
-; CHECK-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave4.nxv8i1(<vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]])
-; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i64> @llvm.masked.load.nxv8i64.p0(ptr [[TMP9]], i32 8, <vscale x 8 x i1> [[INTERLEAVED_MASK]], <vscale x 8 x i64> poison)
+; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i64> @llvm.vp.load.nxv8i64.p0(ptr align 8 [[TMP9]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.vector.deinterleave4.nxv8i64(<vscale x 8 x i64> [[WIDE_MASKED_VEC]])
 ; CHECK-NEXT:    [[TMP24:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 0
 ; CHECK-NEXT:    [[TMP11:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 1
@@ -795,8 +742,7 @@ define void @load_store_factor4(ptr %p) {
 ; CHECK-NEXT:    [[TMP16:%.*]] = add <vscale x 2 x i64> [[TMP12]], splat (i64 3)
 ; CHECK-NEXT:    [[TMP17:%.*]] = add <vscale x 2 x i64> [[TMP13]], splat (i64 4)
 ; CHECK-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 8 x i64> @llvm.vector.interleave4.nxv8i64(<vscale x 2 x i64> [[TMP26]], <vscale x 2 x i64> [[TMP15]], <vscale x 2 x i64> [[TMP16]], <vscale x 2 x i64> [[TMP17]])
-; CHECK-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave4.nxv8i1(<vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]])
-; CHECK-NEXT:    call void @llvm.masked.store.nxv8i64.p0(<vscale x 8 x i64> [[INTERLEAVED_VEC]], ptr [[TMP9]], i32 8, <vscale x 8 x i1> [[INTERLEAVED_MASK1]])
+; CHECK-NEXT:    call void @llvm.vp.store.nxv8i64.p0(<vscale x 8 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP9]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[TMP22:%.*]] = zext i32 [[TMP7]] to i64
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP22]], [[INDEX]]
 ; CHECK-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP22]]
@@ -904,14 +850,9 @@ define void @load_store_factor4(ptr %p) {
 ; SCALABLE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 2, i1 true)
-; SCALABLE-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 2 x i32> poison, i32 [[TMP7]], i64 0
-; SCALABLE-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 2 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 2 x i32> poison, <vscale x 2 x i32> zeroinitializer
-; SCALABLE-NEXT:    [[TMP6:%.*]] = call <vscale x 2 x i32> @llvm.stepvector.nxv2i32()
-; SCALABLE-NEXT:    [[TMP10:%.*]] = icmp ult <vscale x 2 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; SCALABLE-NEXT:    [[TMP8:%.*]] = mul i64 [[INDEX]], 4
 ; SCALABLE-NEXT:    [[TMP9:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP8]]
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave4.nxv8i1(<vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]])
-; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i64> @llvm.masked.load.nxv8i64.p0(ptr [[TMP9]], i32 8, <vscale x 8 x i1> [[INTERLEAVED_MASK]], <vscale x 8 x i64> poison)
+; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i64> @llvm.vp.load.nxv8i64.p0(ptr align 8 [[TMP9]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.vector.deinterleave4.nxv8i64(<vscale x 8 x i64> [[WIDE_MASKED_VEC]])
 ; SCALABLE-NEXT:    [[TMP24:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 0
 ; SCALABLE-NEXT:    [[TMP11:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 1
@@ -922,8 +863,7 @@ define void @load_store_factor4(ptr %p) {
 ; SCALABLE-NEXT:    [[TMP16:%.*]] = add <vscale x 2 x i64> [[TMP12]], splat (i64 3)
 ; SCALABLE-NEXT:    [[TMP17:%.*]] = add <vscale x 2 x i64> [[TMP13]], splat (i64 4)
 ; SCALABLE-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 8 x i64> @llvm.vector.interleave4.nxv8i64(<vscale x 2 x i64> [[TMP26]], <vscale x 2 x i64> [[TMP15]], <vscale x 2 x i64> [[TMP16]], <vscale x 2 x i64> [[TMP17]])
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave4.nxv8i1(<vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]], <vscale x 2 x i1> [[TMP10]])
-; SCALABLE-NEXT:    call void @llvm.masked.store.nxv8i64.p0(<vscale x 8 x i64> [[INTERLEAVED_VEC]], ptr [[TMP9]], i32 8, <vscale x 8 x i1> [[INTERLEAVED_MASK1]])
+; SCALABLE-NEXT:    call void @llvm.vp.store.nxv8i64.p0(<vscale x 8 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP9]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[TMP22:%.*]] = zext i32 [[TMP7]] to i64
 ; SCALABLE-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP22]], [[INDEX]]
 ; SCALABLE-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP22]]
@@ -1008,14 +948,9 @@ define void @load_store_factor5(ptr %p) {
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 1, i1 true)
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i32> poison, i32 [[TMP7]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i32> poison, <vscale x 1 x i32> zeroinitializer
-; CHECK-NEXT:    [[TMP18:%.*]] = call <vscale x 1 x i32> @llvm.stepvector.nxv1i32()
-; CHECK-NEXT:    [[TMP5:%.*]] = icmp ult <vscale x 1 x i32> [[TMP18]], [[BROADCAST_SPLAT]]
 ; CHECK-NEXT:    [[TMP6:%.*]] = mul i64 [[INDEX]], 5
 ; CHECK-NEXT:    [[TMP19:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP6]]
-; CHECK-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 5 x i1> @llvm.vector.interleave5.nxv5i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 5 x i64> @llvm.masked.load.nxv5i64.p0(ptr [[TMP19]], i32 8, <vscale x 5 x i1> [[INTERLEAVED_MASK]], <vscale x 5 x i64> poison)
+; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 5 x i64> @llvm.vp.load.nxv5i64.p0(ptr align 8 [[TMP19]], <vscale x 5 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } @llvm.vector.deinterleave5.nxv5i64(<vscale x 5 x i64> [[WIDE_MASKED_VEC]])
 ; CHECK-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 0
 ; CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 1
@@ -1028,8 +963,7 @@ define void @load_store_factor5(ptr %p) {
 ; CHECK-NEXT:    [[TMP16:%.*]] = add <vscale x 1 x i64> [[TMP11]], splat (i64 4)
 ; CHECK-NEXT:    [[TMP17:%.*]] = add <vscale x 1 x i64> [[TMP12]], splat (i64 5)
 ; CHECK-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 5 x i64> @llvm.vector.interleave5.nxv5i64(<vscale x 1 x i64> [[TMP13]], <vscale x 1 x i64> [[TMP14]], <vscale x 1 x i64> [[TMP15]], <vscale x 1 x i64> [[TMP16]], <vscale x 1 x i64> [[TMP17]])
-; CHECK-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 5 x i1> @llvm.vector.interleave5.nxv5i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; CHECK-NEXT:    call void @llvm.masked.store.nxv5i64.p0(<vscale x 5 x i64> [[INTERLEAVED_VEC]], ptr [[TMP19]], i32 8, <vscale x 5 x i1> [[INTERLEAVED_MASK1]])
+; CHECK-NEXT:    call void @llvm.vp.store.nxv5i64.p0(<vscale x 5 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP19]], <vscale x 5 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[TMP25:%.*]] = zext i32 [[TMP7]] to i64
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP25]], [[INDEX]]
 ; CHECK-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP25]]
@@ -1150,14 +1084,9 @@ define void @load_store_factor5(ptr %p) {
 ; SCALABLE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 1, i1 true)
-; SCALABLE-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i32> poison, i32 [[TMP7]], i64 0
-; SCALABLE-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i32> poison, <vscale x 1 x i32> zeroinitializer
-; SCALABLE-NEXT:    [[TMP18:%.*]] = call <vscale x 1 x i32> @llvm.stepvector.nxv1i32()
-; SCALABLE-NEXT:    [[TMP5:%.*]] = icmp ult <vscale x 1 x i32> [[TMP18]], [[BROADCAST_SPLAT]]
 ; SCALABLE-NEXT:    [[TMP6:%.*]] = mul i64 [[INDEX]], 5
 ; SCALABLE-NEXT:    [[TMP19:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP6]]
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 5 x i1> @llvm.vector.interleave5.nxv5i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 5 x i64> @llvm.masked.load.nxv5i64.p0(ptr [[TMP19]], i32 8, <vscale x 5 x i1> [[INTERLEAVED_MASK]], <vscale x 5 x i64> poison)
+; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 5 x i64> @llvm.vp.load.nxv5i64.p0(ptr align 8 [[TMP19]], <vscale x 5 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } @llvm.vector.deinterleave5.nxv5i64(<vscale x 5 x i64> [[WIDE_MASKED_VEC]])
 ; SCALABLE-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 0
 ; SCALABLE-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 1
@@ -1170,8 +1099,7 @@ define void @load_store_factor5(ptr %p) {
 ; SCALABLE-NEXT:    [[TMP16:%.*]] = add <vscale x 1 x i64> [[TMP11]], splat (i64 4)
 ; SCALABLE-NEXT:    [[TMP17:%.*]] = add <vscale x 1 x i64> [[TMP12]], splat (i64 5)
 ; SCALABLE-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 5 x i64> @llvm.vector.interleave5.nxv5i64(<vscale x 1 x i64> [[TMP13]], <vscale x 1 x i64> [[TMP14]], <vscale x 1 x i64> [[TMP15]], <vscale x 1 x i64> [[TMP16]], <vscale x 1 x i64> [[TMP17]])
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 5 x i1> @llvm.vector.interleave5.nxv5i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; SCALABLE-NEXT:    call void @llvm.masked.store.nxv5i64.p0(<vscale x 5 x i64> [[INTERLEAVED_VEC]], ptr [[TMP19]], i32 8, <vscale x 5 x i1> [[INTERLEAVED_MASK1]])
+; SCALABLE-NEXT:    call void @llvm.vp.store.nxv5i64.p0(<vscale x 5 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP19]], <vscale x 5 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[TMP25:%.*]] = zext i32 [[TMP7]] to i64
 ; SCALABLE-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP25]], [[INDEX]]
 ; SCALABLE-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP25]]
@@ -1267,14 +1195,9 @@ define void @load_store_factor6(ptr %p) {
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 1, i1 true)
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i32> poison, i32 [[TMP7]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i32> poison, <vscale x 1 x i32> zeroinitializer
-; CHECK-NEXT:    [[TMP20:%.*]] = call <vscale x 1 x i32> @llvm.stepvector.nxv1i32()
-; CHECK-NEXT:    [[TMP5:%.*]] = icmp ult <vscale x 1 x i32> [[TMP20]], [[BROADCAST_SPLAT]]
 ; CHECK-NEXT:    [[TMP6:%.*]] = mul i64 [[INDEX]], 6
 ; CHECK-NEXT:    [[TMP21:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP6]]
-; CHECK-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 6 x i1> @llvm.vector.interleave6.nxv6i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 6 x i64> @llvm.masked.load.nxv6i64.p0(ptr [[TMP21]], i32 8, <vscale x 6 x i1> [[INTERLEAVED_MASK]], <vscale x 6 x i64> poison)
+; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 6 x i64> @llvm.vp.load.nxv6i64.p0(ptr align 8 [[TMP21]], <vscale x 6 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } @llvm.vector.deinterleave6.nxv6i64(<vscale x 6 x i64> [[WIDE_MASKED_VEC]])
 ; CHECK-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 0
 ; CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 1
@@ -1289,8 +1212,7 @@ define void @load_store_factor6(ptr %p) {
 ; CHECK-NEXT:    [[TMP18:%.*]] = add <vscale x 1 x i64> [[TMP12]], splat (i64 5)
 ; CHECK-NEXT:    [[TMP19:%.*]] = add <vscale x 1 x i64> [[TMP13]], splat (i64 6)
 ; CHECK-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 6 x i64> @llvm.vector.interleave6.nxv6i64(<vscale x 1 x i64> [[TMP14]], <vscale x 1 x i64> [[TMP15]], <vscale x 1 x i64> [[TMP16]], <vscale x 1 x i64> [[TMP17]], <vscale x 1 x i64> [[TMP18]], <vscale x 1 x i64> [[TMP19]])
-; CHECK-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 6 x i1> @llvm.vector.interleave6.nxv6i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; CHECK-NEXT:    call void @llvm.masked.store.nxv6i64.p0(<vscale x 6 x i64> [[INTERLEAVED_VEC]], ptr [[TMP21]], i32 8, <vscale x 6 x i1> [[INTERLEAVED_MASK1]])
+; CHECK-NEXT:    call void @llvm.vp.store.nxv6i64.p0(<vscale x 6 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP21]], <vscale x 6 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[TMP28:%.*]] = zext i32 [[TMP7]] to i64
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP28]], [[INDEX]]
 ; CHECK-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP28]]
@@ -1424,14 +1346,9 @@ define void @load_store_factor6(ptr %p) {
 ; SCALABLE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 1, i1 true)
-; SCALABLE-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i32> poison, i32 [[TMP7]], i64 0
-; SCALABLE-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i32> poison, <vscale x 1 x i32> zeroinitializer
-; SCALABLE-NEXT:    [[TMP20:%.*]] = call <vscale x 1 x i32> @llvm.stepvector.nxv1i32()
-; SCALABLE-NEXT:    [[TMP5:%.*]] = icmp ult <vscale x 1 x i32> [[TMP20]], [[BROADCAST_SPLAT]]
 ; SCALABLE-NEXT:    [[TMP6:%.*]] = mul i64 [[INDEX]], 6
 ; SCALABLE-NEXT:    [[TMP21:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP6]]
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 6 x i1> @llvm.vector.interleave6.nxv6i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 6 x i64> @llvm.masked.load.nxv6i64.p0(ptr [[TMP21]], i32 8, <vscale x 6 x i1> [[INTERLEAVED_MASK]], <vscale x 6 x i64> poison)
+; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 6 x i64> @llvm.vp.load.nxv6i64.p0(ptr align 8 [[TMP21]], <vscale x 6 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } @llvm.vector.deinterleave6.nxv6i64(<vscale x 6 x i64> [[WIDE_MASKED_VEC]])
 ; SCALABLE-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 0
 ; SCALABLE-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 1
@@ -1446,8 +1363,7 @@ define void @load_store_factor6(ptr %p) {
 ; SCALABLE-NEXT:    [[TMP18:%.*]] = add <vscale x 1 x i64> [[TMP12]], splat (i64 5)
 ; SCALABLE-NEXT:    [[TMP19:%.*]] = add <vscale x 1 x i64> [[TMP13]], splat (i64 6)
 ; SCALABLE-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 6 x i64> @llvm.vector.interleave6.nxv6i64(<vscale x 1 x i64> [[TMP14]], <vscale x 1 x i64> [[TMP15]], <vscale x 1 x i64> [[TMP16]], <vscale x 1 x i64> [[TMP17]], <vscale x 1 x i64> [[TMP18]], <vscale x 1 x i64> [[TMP19]])
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 6 x i1> @llvm.vector.interleave6.nxv6i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; SCALABLE-NEXT:    call void @llvm.masked.store.nxv6i64.p0(<vscale x 6 x i64> [[INTERLEAVED_VEC]], ptr [[TMP21]], i32 8, <vscale x 6 x i1> [[INTERLEAVED_MASK1]])
+; SCALABLE-NEXT:    call void @llvm.vp.store.nxv6i64.p0(<vscale x 6 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP21]], <vscale x 6 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[TMP28:%.*]] = zext i32 [[TMP7]] to i64
 ; SCALABLE-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP28]], [[INDEX]]
 ; SCALABLE-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP28]]
@@ -1554,14 +1470,9 @@ define void @load_store_factor7(ptr %p) {
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 1, i1 true)
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i32> poison, i32 [[TMP7]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i32> poison, <vscale x 1 x i32> zeroinitializer
-; CHECK-NEXT:    [[TMP22:%.*]] = call <vscale x 1 x i32> @llvm.stepvector.nxv1i32()
-; CHECK-NEXT:    [[TMP5:%.*]] = icmp ult <vscale x 1 x i32> [[TMP22]], [[BROADCAST_SPLAT]]
 ; CHECK-NEXT:    [[TMP6:%.*]] = mul i64 [[INDEX]], 7
 ; CHECK-NEXT:    [[TMP23:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP6]]
-; CHECK-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 7 x i1> @llvm.vector.interleave7.nxv7i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 7 x i64> @llvm.masked.load.nxv7i64.p0(ptr [[TMP23]], i32 8, <vscale x 7 x i1> [[INTERLEAVED_MASK]], <vscale x 7 x i64> poison)
+; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 7 x i64> @llvm.vp.load.nxv7i64.p0(ptr align 8 [[TMP23]], <vscale x 7 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } @llvm.vector.deinterleave7.nxv7i64(<vscale x 7 x i64> [[WIDE_MASKED_VEC]])
 ; CHECK-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 0
 ; CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 1
@@ -1578,8 +1489,7 @@ define void @load_store_factor7(ptr %p) {
 ; CHECK-NEXT:    [[TMP20:%.*]] = add <vscale x 1 x i64> [[TMP13]], splat (i64 6)
 ; CHECK-NEXT:    [[TMP21:%.*]] = add <vscale x 1 x i64> [[TMP14]], splat (i64 7)
 ; CHECK-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 7 x i64> @llvm.vector.interleave7.nxv7i64(<vscale x 1 x i64> [[TMP15]], <vscale x 1 x i64> [[TMP16]], <vscale x 1 x i64> [[TMP17]], <vscale x 1 x i64> [[TMP18]], <vscale x 1 x i64> [[TMP19]], <vscale x 1 x i64> [[TMP20]], <vscale x 1 x i64> [[TMP21]])
-; CHECK-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 7 x i1> @llvm.vector.interleave7.nxv7i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; CHECK-NEXT:    call void @llvm.masked.store.nxv7i64.p0(<vscale x 7 x i64> [[INTERLEAVED_VEC]], ptr [[TMP23]], i32 8, <vscale x 7 x i1> [[INTERLEAVED_MASK1]])
+; CHECK-NEXT:    call void @llvm.vp.store.nxv7i64.p0(<vscale x 7 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP23]], <vscale x 7 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[TMP31:%.*]] = zext i32 [[TMP7]] to i64
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP31]], [[INDEX]]
 ; CHECK-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP31]]
@@ -1727,14 +1637,9 @@ define void @load_store_factor7(ptr %p) {
 ; SCALABLE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 1, i1 true)
-; SCALABLE-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i32> poison, i32 [[TMP7]], i64 0
-; SCALABLE-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i32> poison, <vscale x 1 x i32> zeroinitializer
-; SCALABLE-NEXT:    [[TMP22:%.*]] = call <vscale x 1 x i32> @llvm.stepvector.nxv1i32()
-; SCALABLE-NEXT:    [[TMP5:%.*]] = icmp ult <vscale x 1 x i32> [[TMP22]], [[BROADCAST_SPLAT]]
 ; SCALABLE-NEXT:    [[TMP6:%.*]] = mul i64 [[INDEX]], 7
 ; SCALABLE-NEXT:    [[TMP23:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP6]]
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 7 x i1> @llvm.vector.interleave7.nxv7i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 7 x i64> @llvm.masked.load.nxv7i64.p0(ptr [[TMP23]], i32 8, <vscale x 7 x i1> [[INTERLEAVED_MASK]], <vscale x 7 x i64> poison)
+; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 7 x i64> @llvm.vp.load.nxv7i64.p0(ptr align 8 [[TMP23]], <vscale x 7 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } @llvm.vector.deinterleave7.nxv7i64(<vscale x 7 x i64> [[WIDE_MASKED_VEC]])
 ; SCALABLE-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 0
 ; SCALABLE-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 1
@@ -1751,8 +1656,7 @@ define void @load_store_factor7(ptr %p) {
 ; SCALABLE-NEXT:    [[TMP20:%.*]] = add <vscale x 1 x i64> [[TMP13]], splat (i64 6)
 ; SCALABLE-NEXT:    [[TMP21:%.*]] = add <vscale x 1 x i64> [[TMP14]], splat (i64 7)
 ; SCALABLE-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 7 x i64> @llvm.vector.interleave7.nxv7i64(<vscale x 1 x i64> [[TMP15]], <vscale x 1 x i64> [[TMP16]], <vscale x 1 x i64> [[TMP17]], <vscale x 1 x i64> [[TMP18]], <vscale x 1 x i64> [[TMP19]], <vscale x 1 x i64> [[TMP20]], <vscale x 1 x i64> [[TMP21]])
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 7 x i1> @llvm.vector.interleave7.nxv7i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; SCALABLE-NEXT:    call void @llvm.masked.store.nxv7i64.p0(<vscale x 7 x i64> [[INTERLEAVED_VEC]], ptr [[TMP23]], i32 8, <vscale x 7 x i1> [[INTERLEAVED_MASK1]])
+; SCALABLE-NEXT:    call void @llvm.vp.store.nxv7i64.p0(<vscale x 7 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP23]], <vscale x 7 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[TMP31:%.*]] = zext i32 [[TMP7]] to i64
 ; SCALABLE-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP31]], [[INDEX]]
 ; SCALABLE-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP31]]
@@ -1870,14 +1774,9 @@ define void @load_store_factor8(ptr %p) {
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 1, i1 true)
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i32> poison, i32 [[TMP7]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i32> poison, <vscale x 1 x i32> zeroinitializer
-; CHECK-NEXT:    [[TMP4:%.*]] = call <vscale x 1 x i32> @llvm.stepvector.nxv1i32()
-; CHECK-NEXT:    [[TMP5:%.*]] = icmp ult <vscale x 1 x i32> [[TMP4]], [[BROADCAST_SPLAT]]
 ; CHECK-NEXT:    [[TMP6:%.*]] = shl i64 [[INDEX]], 3
 ; CHECK-NEXT:    [[TMP24:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP6]]
-; CHECK-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave8.nxv8i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i64> @llvm.masked.load.nxv8i64.p0(ptr [[TMP24]], i32 8, <vscale x 8 x i1> [[INTERLEAVED_MASK]], <vscale x 8 x i64> poison)
+; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i64> @llvm.vp.load.nxv8i64.p0(ptr align 8 [[TMP24]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } @llvm.vector.deinterleave8.nxv8i64(<vscale x 8 x i64> [[WIDE_MASKED_VEC]])
 ; CHECK-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 0
 ; CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 1
@@ -1896,8 +1795,7 @@ define void @load_store_factor8(ptr %p) {
 ; CHECK-NEXT:    [[TMP22:%.*]] = add <vscale x 1 x i64> [[TMP14]], splat (i64 7)
 ; CHECK-NEXT:    [[TMP23:%.*]] = add <vscale x 1 x i64> [[TMP15]], splat (i64 8)
 ; CHECK-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 8 x i64> @llvm.vector.interleave8.nxv8i64(<vscale x 1 x i64> [[TMP16]], <vscale x 1 x i64> [[TMP17]], <vscale x 1 x i64> [[TMP18]], <vscale x 1 x i64> [[TMP19]], <vscale x 1 x i64> [[TMP20]], <vscale x 1 x i64> [[TMP21]], <vscale x 1 x i64> [[TMP22]], <vscale x 1 x i64> [[TMP23]])
-; CHECK-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave8.nxv8i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; CHECK-NEXT:    call void @llvm.masked.store.nxv8i64.p0(<vscale x 8 x i64> [[INTERLEAVED_VEC]], ptr [[TMP24]], i32 8, <vscale x 8 x i1> [[INTERLEAVED_MASK1]])
+; CHECK-NEXT:    call void @llvm.vp.store.nxv8i64.p0(<vscale x 8 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP24]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[TMP34:%.*]] = zext i32 [[TMP7]] to i64
 ; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP34]], [[INDEX]]
 ; CHECK-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP34]]
@@ -2056,14 +1954,9 @@ define void @load_store_factor8(ptr %p) {
 ; SCALABLE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 1, i1 true)
-; SCALABLE-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 1 x i32> poison, i32 [[TMP7]], i64 0
-; SCALABLE-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 1 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 1 x i32> poison, <vscale x 1 x i32> zeroinitializer
-; SCALABLE-NEXT:    [[TMP4:%.*]] = call <vscale x 1 x i32> @llvm.stepvector.nxv1i32()
-; SCALABLE-NEXT:    [[TMP5:%.*]] = icmp ult <vscale x 1 x i32> [[TMP4]], [[BROADCAST_SPLAT]]
 ; SCALABLE-NEXT:    [[TMP6:%.*]] = shl i64 [[INDEX]], 3
 ; SCALABLE-NEXT:    [[TMP24:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP6]]
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave8.nxv8i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i64> @llvm.masked.load.nxv8i64.p0(ptr [[TMP24]], i32 8, <vscale x 8 x i1> [[INTERLEAVED_MASK]], <vscale x 8 x i64> poison)
+; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i64> @llvm.vp.load.nxv8i64.p0(ptr align 8 [[TMP24]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } @llvm.vector.deinterleave8.nxv8i64(<vscale x 8 x i64> [[WIDE_MASKED_VEC]])
 ; SCALABLE-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 0
 ; SCALABLE-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64>, <vscale x 1 x i64> } [[STRIDED_VEC]], 1
@@ -2082,8 +1975,7 @@ define void @load_store_factor8(ptr %p) {
 ; SCALABLE-NEXT:    [[TMP22:%.*]] = add <vscale x 1 x i64> [[TMP14]], splat (i64 7)
 ; SCALABLE-NEXT:    [[TMP23:%.*]] = add <vscale x 1 x i64> [[TMP15]], splat (i64 8)
 ; SCALABLE-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 8 x i64> @llvm.vector.interleave8.nxv8i64(<vscale x 1 x i64> [[TMP16]], <vscale x 1 x i64> [[TMP17]], <vscale x 1 x i64> [[TMP18]], <vscale x 1 x i64> [[TMP19]], <vscale x 1 x i64> [[TMP20]], <vscale x 1 x i64> [[TMP21]], <vscale x 1 x i64> [[TMP22]], <vscale x 1 x i64> [[TMP23]])
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK1:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave8.nxv8i1(<vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]], <vscale x 1 x i1> [[TMP5]])
-; SCALABLE-NEXT:    call void @llvm.masked.store.nxv8i64.p0(<vscale x 8 x i64> [[INTERLEAVED_VEC]], ptr [[TMP24]], i32 8, <vscale x 8 x i1> [[INTERLEAVED_MASK1]])
+; SCALABLE-NEXT:    call void @llvm.vp.store.nxv8i64.p0(<vscale x 8 x i64> [[INTERLEAVED_VEC]], ptr align 8 [[TMP24]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[TMP34:%.*]] = zext i32 [[TMP7]] to i64
 ; SCALABLE-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[TMP34]], [[INDEX]]
 ; SCALABLE-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP34]]
@@ -2213,14 +2105,9 @@ define void @combine_load_factor2_i32(ptr noalias %p, ptr noalias %q) {
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 4, i1 true)
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[TMP7]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-; CHECK-NEXT:    [[TMP6:%.*]] = call <vscale x 4 x i32> @llvm.stepvector.nxv4i32()
-; CHECK-NEXT:    [[TMP12:%.*]] = icmp ult <vscale x 4 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; CHECK-NEXT:    [[TMP13:%.*]] = shl i64 [[INDEX]], 1
 ; CHECK-NEXT:    [[TMP15:%.*]] = getelementptr i32, ptr [[P:%.*]], i64 [[TMP13]]
-; CHECK-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave2.nxv8i1(<vscale x 4 x i1> [[TMP12]], <vscale x 4 x i1> [[TMP12]])
-; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i32> @llvm.masked.load.nxv8i32.p0(ptr [[TMP15]], i32 4, <vscale x 8 x i1> [[INTERLEAVED_MASK]], <vscale x 8 x i32> poison)
+; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i32> @llvm.vp.load.nxv8i32.p0(ptr align 4 [[TMP15]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.vector.deinterleave2.nxv8i32(<vscale x 8 x i32> [[WIDE_MASKED_VEC]])
 ; CHECK-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 0
 ; CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 1
@@ -2304,14 +2191,9 @@ define void @combine_load_factor2_i32(ptr noalias %p, ptr noalias %q) {
 ; SCALABLE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 4, i1 true)
-; SCALABLE-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[TMP7]], i64 0
-; SCALABLE-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-; SCALABLE-NEXT:    [[TMP6:%.*]] = call <vscale x 4 x i32> @llvm.stepvector.nxv4i32()
-; SCALABLE-NEXT:    [[TMP12:%.*]] = icmp ult <vscale x 4 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; SCALABLE-NEXT:    [[TMP13:%.*]] = shl i64 [[INDEX]], 1
 ; SCALABLE-NEXT:    [[TMP15:%.*]] = getelementptr i32, ptr [[P:%.*]], i64 [[TMP13]]
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave2.nxv8i1(<vscale x 4 x i1> [[TMP12]], <vscale x 4 x i1> [[TMP12]])
-; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i32> @llvm.masked.load.nxv8i32.p0(ptr [[TMP15]], i32 4, <vscale x 8 x i1> [[INTERLEAVED_MASK]], <vscale x 8 x i32> poison)
+; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 8 x i32> @llvm.vp.load.nxv8i32.p0(ptr align 4 [[TMP15]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.vector.deinterleave2.nxv8i32(<vscale x 8 x i32> [[WIDE_MASKED_VEC]])
 ; SCALABLE-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 0
 ; SCALABLE-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 1
@@ -2381,14 +2263,9 @@ define void @combine_load_factor2_i64(ptr noalias %p, ptr noalias %q) {
 ; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; CHECK-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 2, i1 true)
-; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 2 x i32> poison, i32 [[TMP7]], i64 0
-; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 2 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 2 x i32> poison, <vscale x 2 x i32> zeroinitializer
-; CHECK-NEXT:    [[TMP6:%.*]] = call <vscale x 2 x i32> @llvm.stepvector.nxv2i32()
-; CHECK-NEXT:    [[TMP12:%.*]] = icmp ult <vscale x 2 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; CHECK-NEXT:    [[TMP13:%.*]] = shl i64 [[INDEX]], 1
 ; CHECK-NEXT:    [[TMP15:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP13]]
-; CHECK-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 4 x i1> @llvm.vector.interleave2.nxv4i1(<vscale x 2 x i1> [[TMP12]], <vscale x 2 x i1> [[TMP12]])
-; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 4 x i64> @llvm.masked.load.nxv4i64.p0(ptr [[TMP15]], i32 8, <vscale x 4 x i1> [[INTERLEAVED_MASK]], <vscale x 4 x i64> poison)
+; CHECK-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 4 x i64> @llvm.vp.load.nxv4i64.p0(ptr align 8 [[TMP15]], <vscale x 4 x i1> splat (i1 true), i32 [[TMP7]])
 ; CHECK-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.vector.deinterleave2.nxv4i64(<vscale x 4 x i64> [[WIDE_MASKED_VEC]])
 ; CHECK-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 0
 ; CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 1
@@ -2472,14 +2349,9 @@ define void @combine_load_factor2_i64(ptr noalias %p, ptr noalias %q) {
 ; SCALABLE-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[AVL:%.*]] = phi i64 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; SCALABLE-NEXT:    [[TMP7:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 2, i1 true)
-; SCALABLE-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 2 x i32> poison, i32 [[TMP7]], i64 0
-; SCALABLE-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 2 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 2 x i32> poison, <vscale x 2 x i32> zeroinitializer
-; SCALABLE-NEXT:    [[TMP6:%.*]] = call <vscale x 2 x i32> @llvm.stepvector.nxv2i32()
-; SCALABLE-NEXT:    [[TMP12:%.*]] = icmp ult <vscale x 2 x i32> [[TMP6]], [[BROADCAST_SPLAT]]
 ; SCALABLE-NEXT:    [[TMP13:%.*]] = shl i64 [[INDEX]], 1
 ; SCALABLE-NEXT:    [[TMP15:%.*]] = getelementptr i64, ptr [[P:%.*]], i64 [[TMP13]]
-; SCALABLE-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 4 x i1> @llvm.vector.interleave2.nxv4i1(<vscale x 2 x i1> [[TMP12]], <vscale x 2 x i1> [[TMP12]])
-; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 4 x i64> @llvm.masked.load.nxv4i64.p0(ptr [[TMP15]], i32 8, <vscale x 4 x i1> [[INTERLEAVED_MASK]], <vscale x 4 x i64> poison)
+; SCALABLE-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 4 x i64> @llvm.vp.load.nxv4i64.p0(ptr align 8 [[TMP15]], <vscale x 4 x i1> splat (i1 true), i32 [[TMP7]])
 ; SCALABLE-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.vector.deinterleave2.nxv4i64(<vscale x 4 x i64> [[WIDE_MASKED_VEC]])
 ; SCALABLE-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 0
 ; SCALABLE-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64> } [[STRIDED_VEC]], 1
diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/interleaved-masked-access.ll b/llvm/test/Transforms/LoopVectorize/RISCV/interleaved-masked-access.ll
index 1bceb871bd999..f581442112eea 100644
--- a/llvm/test/Transforms/LoopVectorize/RISCV/interleaved-masked-access.ll
+++ b/llvm/test/Transforms/LoopVectorize/RISCV/interleaved-masked-access.ll
@@ -113,34 +113,29 @@ define void @masked_strided_factor2(ptr noalias nocapture readonly %p, ptr noali
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[VEC_IND:%.*]] = phi <vscale x 16 x i32> [ [[TMP0]], [[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[AVL:%.*]] = phi i32 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP1:%.*]] = call i32 @llvm.experimental.get.vector.length.i32(i32 [[AVL]], i32 16, i1 true)
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[BROADCAST_SPLATINSERT3:%.*]] = insertelement <vscale x 16 x i32> poison, i32 [[TMP1]], i64 0
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[BROADCAST_SPLAT4:%.*]] = shufflevector <vscale x 16 x i32> [[BROADCAST_SPLATINSERT3]], <vscale x 16 x i32> poison, <vscale x 16 x i32> zeroinitializer
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <vscale x 16 x i32> poison, i32 [[TMP1]], i64 0
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[BROADCAST_SPLAT2:%.*]] = shufflevector <vscale x 16 x i32> [[BROADCAST_SPLATINSERT1]], <vscale x 16 x i32> poison, <vscale x 16 x i32> zeroinitializer
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP2:%.*]] = call <vscale x 16 x i32> @llvm.stepvector.nxv16i32()
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP3:%.*]] = icmp ult <vscale x 16 x i32> [[TMP2]], [[BROADCAST_SPLAT4]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP4:%.*]] = icmp ugt <vscale x 16 x i32> [[VEC_IND]], [[BROADCAST_SPLAT]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP5:%.*]] = select <vscale x 16 x i1> [[TMP3]], <vscale x 16 x i1> [[TMP4]], <vscale x 16 x i1> zeroinitializer
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP6:%.*]] = shl i32 [[EVL_BASED_IV]], 1
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP7:%.*]] = sext i32 [[TMP6]] to i64
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP8:%.*]] = getelementptr i8, ptr [[P]], i64 [[TMP7]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 32 x i1> @llvm.vector.interleave2.nxv32i1(<vscale x 16 x i1> [[TMP5]], <vscale x 16 x i1> [[TMP5]])
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 32 x i8> @llvm.masked.load.nxv32i8.p0(ptr [[TMP8]], i32 1, <vscale x 32 x i1> [[INTERLEAVED_MASK]], <vscale x 32 x i8> poison)
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.vector.deinterleave2.nxv32i8(<vscale x 32 x i8> [[WIDE_MASKED_VEC]])
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8> } [[STRIDED_VEC]], 0
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP10:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8> } [[STRIDED_VEC]], 1
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP11:%.*]] = call <vscale x 16 x i8> @llvm.smax.nxv16i8(<vscale x 16 x i8> [[TMP9]], <vscale x 16 x i8> [[TMP10]])
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP12:%.*]] = sext i32 [[TMP6]] to i64
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP13:%.*]] = getelementptr i8, ptr [[Q]], i64 [[TMP12]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP14:%.*]] = sub <vscale x 16 x i8> zeroinitializer, [[TMP11]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 32 x i8> @llvm.vector.interleave2.nxv32i8(<vscale x 16 x i8> [[TMP11]], <vscale x 16 x i8> [[TMP14]])
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[INTERLEAVED_MASK5:%.*]] = call <vscale x 32 x i1> @llvm.vector.interleave2.nxv32i1(<vscale x 16 x i1> [[TMP5]], <vscale x 16 x i1> [[TMP5]])
-; PREDICATED_DATA-WITH-EVL-NEXT:    call void @llvm.masked.store.nxv32i8.p0(<vscale x 32 x i8> [[INTERLEAVED_VEC]], ptr [[TMP13]], i32 1, <vscale x 32 x i1> [[INTERLEAVED_MASK5]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP2:%.*]] = icmp ugt <vscale x 16 x i32> [[VEC_IND]], [[BROADCAST_SPLAT]]
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP3:%.*]] = shl i32 [[EVL_BASED_IV]], 1
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP4:%.*]] = sext i32 [[TMP3]] to i64
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP5:%.*]] = getelementptr i8, ptr [[P]], i64 [[TMP4]]
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 32 x i1> @llvm.vector.interleave2.nxv32i1(<vscale x 16 x i1> [[TMP2]], <vscale x 16 x i1> [[TMP2]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[WIDE_VP_LOAD:%.*]] = call <vscale x 32 x i8> @llvm.vp.load.nxv32i8.p0(ptr align 1 [[TMP5]], <vscale x 32 x i1> [[INTERLEAVED_MASK]], i32 [[TMP1]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.vector.deinterleave2.nxv32i8(<vscale x 32 x i8> [[WIDE_VP_LOAD]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP6:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8> } [[STRIDED_VEC]], 0
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP7:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8> } [[STRIDED_VEC]], 1
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP8:%.*]] = call <vscale x 16 x i8> @llvm.smax.nxv16i8(<vscale x 16 x i8> [[TMP6]], <vscale x 16 x i8> [[TMP7]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP9:%.*]] = sext i32 [[TMP3]] to i64
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP10:%.*]] = getelementptr i8, ptr [[Q]], i64 [[TMP9]]
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP11:%.*]] = sub <vscale x 16 x i8> zeroinitializer, [[TMP8]]
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[INTERLEAVED_MASK3:%.*]] = call <vscale x 32 x i1> @llvm.vector.interleave2.nxv32i1(<vscale x 16 x i1> [[TMP2]], <vscale x 16 x i1> [[TMP2]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 32 x i8> @llvm.vector.interleave2.nxv32i8(<vscale x 16 x i8> [[TMP8]], <vscale x 16 x i8> [[TMP11]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    call void @llvm.vp.store.nxv32i8.p0(<vscale x 32 x i8> [[INTERLEAVED_VEC]], ptr align 1 [[TMP10]], <vscale x 32 x i1> [[INTERLEAVED_MASK3]], i32 [[TMP1]])
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[INDEX_EVL_NEXT]] = add nuw i32 [[TMP1]], [[EVL_BASED_IV]]
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[AVL_NEXT]] = sub nuw i32 [[AVL]], [[TMP1]]
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[VEC_IND_NEXT]] = add <vscale x 16 x i32> [[VEC_IND]], [[BROADCAST_SPLAT2]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP15:%.*]] = icmp eq i32 [[INDEX_EVL_NEXT]], 1024
-; PREDICATED_DATA-WITH-EVL-NEXT:    br i1 [[TMP15]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP12:%.*]] = icmp eq i32 [[INDEX_EVL_NEXT]], 1024
+; PREDICATED_DATA-WITH-EVL-NEXT:    br i1 [[TMP12]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
 ; PREDICATED_DATA-WITH-EVL:       middle.block:
 ; PREDICATED_DATA-WITH-EVL-NEXT:    br label [[FOR_END:%.*]]
 ; PREDICATED_DATA-WITH-EVL:       scalar.ph:
@@ -296,38 +291,33 @@ define void @masked_strided_factor4(ptr noalias nocapture readonly %p, ptr noali
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[VEC_IND:%.*]] = phi <vscale x 16 x i32> [ [[TMP0]], [[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[AVL:%.*]] = phi i32 [ 1024, [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP1:%.*]] = call i32 @llvm.experimental.get.vector.length.i32(i32 [[AVL]], i32 16, i1 true)
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[BROADCAST_SPLATINSERT3:%.*]] = insertelement <vscale x 16 x i32> poison, i32 [[TMP1]], i64 0
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[BROADCAST_SPLAT4:%.*]] = shufflevector <vscale x 16 x i32> [[BROADCAST_SPLATINSERT3]], <vscale x 16 x i32> poison, <vscale x 16 x i32> zeroinitializer
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <vscale x 16 x i32> poison, i32 [[TMP1]], i64 0
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[BROADCAST_SPLAT2:%.*]] = shufflevector <vscale x 16 x i32> [[BROADCAST_SPLATINSERT1]], <vscale x 16 x i32> poison, <vscale x 16 x i32> zeroinitializer
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP2:%.*]] = call <vscale x 16 x i32> @llvm.stepvector.nxv16i32()
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP3:%.*]] = icmp ult <vscale x 16 x i32> [[TMP2]], [[BROADCAST_SPLAT4]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP4:%.*]] = icmp ugt <vscale x 16 x i32> [[VEC_IND]], [[BROADCAST_SPLAT]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP5:%.*]] = select <vscale x 16 x i1> [[TMP3]], <vscale x 16 x i1> [[TMP4]], <vscale x 16 x i1> zeroinitializer
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP6:%.*]] = shl i32 [[EVL_BASED_IV]], 2
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP7:%.*]] = sext i32 [[TMP6]] to i64
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP8:%.*]] = getelementptr i8, ptr [[P]], i64 [[TMP7]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 64 x i1> @llvm.vector.interleave4.nxv64i1(<vscale x 16 x i1> [[TMP5]], <vscale x 16 x i1> [[TMP5]], <vscale x 16 x i1> [[TMP5]], <vscale x 16 x i1> [[TMP5]])
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[WIDE_MASKED_VEC:%.*]] = call <vscale x 64 x i8> @llvm.masked.load.nxv64i8.p0(ptr [[TMP8]], i32 1, <vscale x 64 x i1> [[INTERLEAVED_MASK]], <vscale x 64 x i8> poison)
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.vector.deinterleave4.nxv64i8(<vscale x 64 x i8> [[WIDE_MASKED_VEC]])
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[STRIDED_VEC]], 0
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP10:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[STRIDED_VEC]], 1
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP11:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[STRIDED_VEC]], 2
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP12:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[STRIDED_VEC]], 3
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP13:%.*]] = call <vscale x 16 x i8> @llvm.smax.nxv16i8(<vscale x 16 x i8> [[TMP9]], <vscale x 16 x i8> [[TMP10]])
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP14:%.*]] = sub <vscale x 16 x i8> zeroinitializer, [[TMP13]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP15:%.*]] = call <vscale x 16 x i8> @llvm.smax.nxv16i8(<vscale x 16 x i8> [[TMP11]], <vscale x 16 x i8> [[TMP12]])
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP16:%.*]] = sub <vscale x 16 x i8> zeroinitializer, [[TMP15]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP17:%.*]] = sext i32 [[TMP6]] to i64
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP18:%.*]] = getelementptr i8, ptr [[Q]], i64 [[TMP17]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 64 x i8> @llvm.vector.interleave4.nxv64i8(<vscale x 16 x i8> [[TMP13]], <vscale x 16 x i8> [[TMP14]], <vscale x 16 x i8> [[TMP15]], <vscale x 16 x i8> [[TMP16]])
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[INTERLEAVED_MASK5:%.*]] = call <vscale x 64 x i1> @llvm.vector.interleave4.nxv64i1(<vscale x 16 x i1> [[TMP5]], <vscale x 16 x i1> [[TMP5]], <vscale x 16 x i1> [[TMP5]], <vscale x 16 x i1> [[TMP5]])
-; PREDICATED_DATA-WITH-EVL-NEXT:    call void @llvm.masked.store.nxv64i8.p0(<vscale x 64 x i8> [[INTERLEAVED_VEC]], ptr [[TMP18]], i32 1, <vscale x 64 x i1> [[INTERLEAVED_MASK5]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP2:%.*]] = icmp ugt <vscale x 16 x i32> [[VEC_IND]], [[BROADCAST_SPLAT]]
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP3:%.*]] = shl i32 [[EVL_BASED_IV]], 2
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP4:%.*]] = sext i32 [[TMP3]] to i64
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP5:%.*]] = getelementptr i8, ptr [[P]], i64 [[TMP4]]
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 64 x i1> @llvm.vector.interleave4.nxv64i1(<vscale x 16 x i1> [[TMP2]], <vscale x 16 x i1> [[TMP2]], <vscale x 16 x i1> [[TMP2]], <vscale x 16 x i1> [[TMP2]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[WIDE_VP_LOAD:%.*]] = call <vscale x 64 x i8> @llvm.vp.load.nxv64i8.p0(ptr align 1 [[TMP5]], <vscale x 64 x i1> [[INTERLEAVED_MASK]], i32 [[TMP1]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.vector.deinterleave4.nxv64i8(<vscale x 64 x i8> [[WIDE_VP_LOAD]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP6:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[STRIDED_VEC]], 0
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP7:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[STRIDED_VEC]], 1
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP8:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[STRIDED_VEC]], 2
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[STRIDED_VEC]], 3
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP10:%.*]] = call <vscale x 16 x i8> @llvm.smax.nxv16i8(<vscale x 16 x i8> [[TMP6]], <vscale x 16 x i8> [[TMP7]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP11:%.*]] = sub <vscale x 16 x i8> zeroinitializer, [[TMP10]]
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP12:%.*]] = call <vscale x 16 x i8> @llvm.smax.nxv16i8(<vscale x 16 x i8> [[TMP8]], <vscale x 16 x i8> [[TMP9]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP13:%.*]] = sub <vscale x 16 x i8> zeroinitializer, [[TMP12]]
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP14:%.*]] = sext i32 [[TMP3]] to i64
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP15:%.*]] = getelementptr i8, ptr [[Q]], i64 [[TMP14]]
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[INTERLEAVED_MASK3:%.*]] = call <vscale x 64 x i1> @llvm.vector.interleave4.nxv64i1(<vscale x 16 x i1> [[TMP2]], <vscale x 16 x i1> [[TMP2]], <vscale x 16 x i1> [[TMP2]], <vscale x 16 x i1> [[TMP2]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[INTERLEAVED_VEC:%.*]] = call <vscale x 64 x i8> @llvm.vector.interleave4.nxv64i8(<vscale x 16 x i8> [[TMP10]], <vscale x 16 x i8> [[TMP11]], <vscale x 16 x i8> [[TMP12]], <vscale x 16 x i8> [[TMP13]])
+; PREDICATED_DATA-WITH-EVL-NEXT:    call void @llvm.vp.store.nxv64i8.p0(<vscale x 64 x i8> [[INTERLEAVED_VEC]], ptr align 1 [[TMP15]], <vscale x 64 x i1> [[INTERLEAVED_MASK3]], i32 [[TMP1]])
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[INDEX_EVL_NEXT]] = add nuw i32 [[TMP1]], [[EVL_BASED_IV]]
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[AVL_NEXT]] = sub nuw i32 [[AVL]], [[TMP1]]
 ; PREDICATED_DATA-WITH-EVL-NEXT:    [[VEC_IND_NEXT]] = add <vscale x 16 x i32> [[VEC_IND]], [[BROADCAST_SPLAT2]]
-; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP19:%.*]] = icmp eq i32 [[INDEX_EVL_NEXT]], 1024
-; PREDICATED_DATA-WITH-EVL-NEXT:    br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]]
+; PREDICATED_DATA-WITH-EVL-NEXT:    [[TMP16:%.*]] = icmp eq i32 [[INDEX_EVL_NEXT]], 1024
+; PREDICATED_DATA-WITH-EVL-NEXT:    br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]]
 ; PREDICATED_DATA-WITH-EVL:       middle.block:
 ; PREDICATED_DATA-WITH-EVL-NEXT:    br label [[FOR_END:%.*]]
 ; PREDICATED_DATA-WITH-EVL:       scalar.ph:
diff --git a/llvm/test/Transforms/LoopVectorize/RISCV/tail-folding-interleave.ll b/llvm/test/Transforms/LoopVectorize/RISCV/tail-folding-interleave.ll
index f1ec586b65e03..dde5edc0a895f 100644
--- a/llvm/test/Transforms/LoopVectorize/RISCV/tail-folding-interleave.ll
+++ b/llvm/test/Transforms/LoopVectorize/RISCV/tail-folding-interleave.ll
@@ -7,7 +7,6 @@
 ; RUN: -prefer-predicate-over-epilogue=scalar-epilogue \
 ; RUN: -mtriple=riscv64 -mattr=+v -S < %s | FileCheck --check-prefix=NO-VP %s
 
-; FIXME: interleaved accesses are not supported yet with predicated vectorization.
 define void @interleave(ptr noalias %a, ptr noalias %b, i64 %N) {
 ; IF-EVL-LABEL: @interleave(
 ; IF-EVL-NEXT:  entry:
@@ -19,25 +18,20 @@ define void @interleave(ptr noalias %a, ptr noalias %b, i64 %N) {
 ; IF-EVL:       vector.body:
 ; IF-EVL-NEXT:    [[EVL_BASED_IV:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_EVL_NEXT:%.*]], [[VECTOR_BODY]] ]
 ; IF-EVL-NEXT:    [[AVL:%.*]] = phi i64 [ [[N:%.*]], [[VECTOR_PH]] ], [ [[AVL_NEXT:%.*]], [[VECTOR_BODY]] ]
-; IF-EVL-NEXT:    [[TMP11:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 4, i1 true)
-; IF-EVL-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <vscale x 4 x i32> poison, i32 [[TMP11]], i64 0
-; IF-EVL-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <vscale x 4 x i32> [[BROADCAST_SPLATINSERT]], <vscale x 4 x i32> poison, <vscale x 4 x i32> zeroinitializer
-; IF-EVL-NEXT:    [[TMP16:%.*]] = call <vscale x 4 x i32> @llvm.stepvector.nxv4i32()
-; IF-EVL-NEXT:    [[TMP17:%.*]] = icmp ult <vscale x 4 x i32> [[TMP16]], [[BROADCAST_SPLAT]]
+; IF-EVL-NEXT:    [[TMP16:%.*]] = call i32 @llvm.experimental.get.vector.length.i64(i64 [[AVL]], i32 4, i1 true)
 ; IF-EVL-NEXT:    [[TMP6:%.*]] = getelementptr inbounds [2 x i32], ptr [[B:%.*]], i64 [[EVL_BASED_IV]], i32 0
-; IF-EVL-NEXT:    [[INTERLEAVED_MASK:%.*]] = call <vscale x 8 x i1> @llvm.vector.interleave2.nxv8i1(<vscale x 4 x i1> [[TMP17]], <vscale x 4 x i1> [[TMP17]])
-; IF-EVL-NEXT:    [[WIDE_VEC:%.*]] = call <vscale x 8 x i32> @llvm.masked.load.nxv8i32.p0(ptr [[TMP6]], i32 4, <vscale x 8 x i1> [[INTERLEAVED_MASK]], <vscale x 8 x i32> poison)
+; IF-EVL-NEXT:    [[WIDE_VEC:%.*]] = call <vscale x 8 x i32> @llvm.vp.load.nxv8i32.p0(ptr align 4 [[TMP6]], <vscale x 8 x i1> splat (i1 true), i32 [[TMP16]])
 ; IF-EVL-NEXT:    [[STRIDED_VEC:%.*]] = call { <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.vector.deinterleave2.nxv8i32(<vscale x 8 x i32> [[WIDE_VEC]])
 ; IF-EVL-NEXT:    [[TMP14:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 0
 ; IF-EVL-NEXT:    [[TMP15:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[STRIDED_VEC]], 1
 ; IF-EVL-NEXT:    [[TMP9:%.*]] = add nsw <vscale x 4 x i32> [[TMP15]], [[TMP14]]
 ; IF-EVL-NEXT:    [[TMP10:%.*]] = getelementptr inbounds i32, ptr [[A:%.*]], i64 [[EVL_BASED_IV]]
-; IF-EVL-NEXT:    call void @llvm.vp.store.nxv4i32.p0(<vscale x 4 x i32> [[TMP9]], ptr align 4 [[TMP10]], <vscale x 4 x i1> splat (i1 true), i32 [[TMP11]])
-; IF-EVL-NEXT:    [[TMP18:%.*]] = zext i32 [[TMP11]] to i64
-; IF-EVL-NEXT:    [[INDEX_EVL_NEXT]] = add i64 [[TMP18]], [[EVL_BASED_IV]]
-; IF-EVL-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP18]]
-; IF-EVL-NEXT:    [[TMP19:%.*]] = icmp eq i64 [[INDEX_EVL_NEXT]], [[N]]
-; IF-EVL-NEXT:    br i1 [[TMP19]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
+; IF-EVL-NEXT:    call void @llvm.vp.store.nxv4i32.p0(<vscale x 4 x i32> [[TMP9]], ptr align 4 [[TMP10]], <vscale x 4 x i1> splat (i1 true), i32 [[TMP16]])
+; IF-EVL-NEXT:    [[TMP11:%.*]] = zext i32 [[TMP16]] to i64
+; IF-EVL-NEXT:    [[INDEX_EVL_NEXT]] = add i64 [[TMP11]], [[EVL_BASED_IV]]
+; IF-EVL-NEXT:    [[AVL_NEXT]] = sub nuw i64 [[AVL]], [[TMP11]]
+; IF-EVL-NEXT:    [[TMP17:%.*]] = icmp eq i64 [[INDEX_EVL_NEXT]], [[N]]
+; IF-EVL-NEXT:    br i1 [[TMP17]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
 ; IF-EVL:       middle.block:
 ; IF-EVL-NEXT:    br label [[FOR_COND_CLEANUP:%.*]]
 ; IF-EVL:       scalar.ph:

>From 324ea8b53827f6716904942491b6e761694296cd Mon Sep 17 00:00:00 2001
From: Mel Chen <mel.chen at sifive.com>
Date: Mon, 11 Aug 2025 01:31:16 -0700
Subject: [PATCH 2/3] use base class to get type

---
 llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp b/llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp
index b5efe9b560c7e..513e0cb4cc4c1 100644
--- a/llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp
+++ b/llvm/lib/Transforms/Vectorize/VPlanAnalysis.cpp
@@ -297,7 +297,7 @@ Type *VPTypeAnalysis::inferScalarType(const VPValue *V) {
           .Case<VPBlendRecipe, VPInstruction, VPWidenRecipe, VPReplicateRecipe,
                 VPWidenCallRecipe, VPWidenMemoryRecipe, VPWidenSelectRecipe>(
               [this](const auto *R) { return inferScalarTypeForRecipe(R); })
-          .Case<VPInterleaveRecipe, VPInterleaveEVLRecipe>([V](const auto *R) {
+          .Case<VPInterleaveBase>([V](const auto *R) {
             // TODO: Use info from interleave group.
             return V->getUnderlyingValue()->getType();
           })

>From 9f448ad8041c3dd4c8275be5b84db300a251232a Mon Sep 17 00:00:00 2001
From: Mel Chen <mel.chen at sifive.com>
Date: Mon, 11 Aug 2025 03:35:11 -0700
Subject: [PATCH 3/3] Unify getStoredValues()

---
 llvm/lib/Transforms/Vectorize/VPlan.h | 20 +++++---------------
 1 file changed, 5 insertions(+), 15 deletions(-)

diff --git a/llvm/lib/Transforms/Vectorize/VPlan.h b/llvm/lib/Transforms/Vectorize/VPlan.h
index e297a38ceddda..33991f5ebbc50 100644
--- a/llvm/lib/Transforms/Vectorize/VPlan.h
+++ b/llvm/lib/Transforms/Vectorize/VPlan.h
@@ -2481,7 +2481,11 @@ class LLVM_ABI_FOR_TEST VPInterleaveBase : public VPRecipeBase {
 
   /// Return the VPValues stored by this interleave group. If it is a load
   /// interleave group, return an empty ArrayRef.
-  virtual ArrayRef<VPValue *> getStoredValues() const = 0;
+  ArrayRef<VPValue *> getStoredValues() const {
+    return ArrayRef<VPValue *>(op_end() -
+                                   (getNumStoreOperands() + (HasMask ? 1 : 0)),
+                               getNumStoreOperands());
+  }
 };
 
 /// VPInterleaveRecipe is a recipe for transforming an interleave group of load
@@ -2523,13 +2527,6 @@ class LLVM_ABI_FOR_TEST VPInterleaveRecipe final : public VPInterleaveBase {
   unsigned getNumStoreOperands() const override {
     return getNumOperands() - (HasMask ? 2 : 1);
   }
-
-  ArrayRef<VPValue *> getStoredValues() const override {
-    // The first operand is the address, followed by the stored values, followed
-    // by an optional mask.
-    return ArrayRef<VPValue *>(op_begin(), getNumOperands())
-        .slice(1, getNumStoreOperands());
-  }
 };
 
 /// A recipe for interleaved access operations with vector-predication
@@ -2579,13 +2576,6 @@ class LLVM_ABI_FOR_TEST VPInterleaveEVLRecipe final : public VPInterleaveBase {
   unsigned getNumStoreOperands() const override {
     return getNumOperands() - (HasMask ? 3 : 2);
   }
-
-  ArrayRef<VPValue *> getStoredValues() const override {
-    // The first operand is the address, and the second operand is EVL, followed
-    // by the stored values, followe by an optional mask.
-    return ArrayRef<VPValue *>(op_begin(), getNumOperands())
-        .slice(2, getNumStoreOperands());
-  }
 };
 
 /// A recipe to represent inloop reduction operations, performing a reduction on