[LV] Changing the interface of ValueMap

[Ayal]

Summary

Following the review of D32871:

An alternative interface which may be simpler and clearer is to only support setting and getting a single Value per part, or per part-and-lane. Setting a value can assert that no value has already been set; this assert can be overridden when needed using a specialized "resetting" method. This keeps the implementation of *MapStorage internal to ValueMap, albeit potentially being less efficient as it avoids batching (if not inlined). If so, it may be better done in a separate patch. Sounds reasonable?

This is that separate patch.

[mssimpso]

Hi Ayal,

Sorry for the long turn-around on this. This looks mostly fine to me. I think it would make sense to commit this prior to the VPlan patch since it's conceptually a separate change. Committing first would also ensure that we keep the interface clean, throughout.

Michael, do you have any issues with moving to this new interface for the ValueMap?

lib/Transforms/Vectorize/LoopVectorize.cpp

685–687

Don't we still need to declare this functions, and the others, as friends?

698

Entry[Part].resize(VF) null initializes the values, right?

2448–2449

It might be more efficient in the common case if we pulled this out of the loop. Though, I don't really have a strong preference.

if (isa<TruncInst>(EntryVal))
  for (unsigned Part = 0; Part < UF; ++Part)
    addMetadata(getOrCreateVectorValue(EntryVal, Part);

Hopefully, I've understood the interface correctly...

2562–2563

Same as above.

2705

Do we need to check/assert that the map contains a non-null entry for each Lane? Is it possible that some of the entries would be null?

Maybe uniform-after-vectorization values have null entries for VF > 1? If this is still the case, it might be useful to re-think that choice. It might be simpler to just broadcast the VF == 1 value to the other lanes instead of leaving them null.

4037

Why do we use VectorLoopValueMap.getVectorValue() here instead of getOrCreateVectorValue()? Is it because we know the Phi must exist in the map already? This makes sense to me.

I think in the original code, we had to use VectorLoopValueMap.getVector() to get a non-const reference.

4216–4217

Can't we just use VectorLoopValueMap.getVectorValue(LoopExitInst, Part), etc.?

[mkuper]

In D34473#789254, @mssimpso wrote:

Michael, do you have any issues with moving to this new interface for the ValueMap?

Makes sense to me, modulo a couple of nits.

lib/Transforms/Vectorize/LoopVectorize.cpp

710

Since unset is only ever used before set, I think a nicer interface would be to have "reset" that combines an unset with a set.
But if you prefer this, I'm ok with it.

3009

This comment doesn't quite parse.

[Ayal]

lib/Transforms/Vectorize/LoopVectorize.cpp

685–687

Exposing this interface that manipulates individual Values only, keeps their storage internal to ValueMap, w/o providing any external befriended method access to it. This follows the thought of eventually moving this interface out of ILV and into a standalone storage class.

698

Ah, sure, if we remember to do Entry[Part].resize(VF, nullptr) ... Good catch, Thanks!

710

Having "reset" was our original thought too, will return to it. Having "unset" may also undermine hasAnyVectorValue() btw.

2448–2449

We could indeed unswitch this invariant "if" manually; but its also good to reuse LastInduction rather than retrieve it. The trip count is probably small for this to have noticeable impact. It may also be conceptually better to keep addMetadata() right next to setVectorValue().

2562–2563

ditto

2705

As originally documented, this check is essentially asking if the value has been scalarized, to any lane/part. For that we can (continue to) check if the map contains the value, w/o looking at any specific lane/parts. Perhaps it would be better instead to ask directly if the decision is to scalarize this value (going forward, by consulting the associated VPlan recipe), instead of asking ValueMap for this information.

The uniform-after-vectorization values are indeed entered per lane 0 only, keeping other lanes null. Another way to re-think this choice is to allocate only UF entries for such values, and separate the type (scalar/vector) of the values being stored from how many values are being stored (per part, or per part-and-lane). Can add a TODO comment.

3009

Right; will fix the comment as we move it.

4037

Yes, we use VectorLoopValueMap.getVectorValue(Phi, 0) here because we know the Phi must already exist in the map; the original code asserts so, and should have also asserted that PhiParts[0] is not null, as it is passed to SetInsertPoint().

Agreed, the original code uses VectorLoopValueMap.getVector(Phi) mainly to get a non-const reference for (re)setting PhiPart[Part] = Shuffle further below.

4216–4217

We could just use VectorLoopValueMap.getVectorValue(LoopExitInst, Part) and hold a single Value *RdxPart instead of holding the temporary VectorParts RdxParts(UF), if the for-part loop creating the SExt/Zext's is fused with the for-part loop creating the Trunc's; this requires moving the insertion point back and forth...

We should indeed do this in the for-part loop below which creates a BinOp or MinMaxOp. Will do so in a revised version to be uploaded.

[Ayal]

Updated version addresses review comments.

[mkuper]

LGTM, but please wait for Matt.

lib/Transforms/Vectorize/LoopVectorize.cpp

698

I'm pretty sure it null-initializes implicitly, but explicit is always good. :-)

[mssimpso]

LGTM as well. Thanks!

[v.g.vassilev]

Hey, it looks like this broke our modules builds (http://lab.llvm.org:8011/builders/clang-x86_64-linux-selfhost-modules/builds/6937/steps/compile.llvm.stage2/logs/stdio) Do you have an idea what might be going on?

lib/Transforms/Vectorize/LoopVectorize.cpp

protected:
 
   /// Returns true if an instruction \p I should be scalarized instead of
   /// vectorized for the chosen vectorization factor.
   bool shouldScalarizeInstruction(Instruction *I) const;
 
   /// Returns true if we should generate a scalar version of \p IV.
   bool needsScalarInduction(Instruction *IV) const;
 
-  /// Return a constant reference to the VectorParts corresponding to \p V from
-  /// the original loop. If the value has already been vectorized, the
-  /// corresponding vector entry in VectorLoopValueMap is returned. If,
+  /// getOrCreateVectorValue and getOrCreateScalarValue coordinate to generate a
+  /// vector or scalar value on-demand if one is not yet available. When
+  /// vectorizing a loop, we visit the definition of an instruction before its
+  /// uses. When visiting the definition, we either vectorize or scalarize the
+  /// instruction, creating an entry for it in the corresponding map. (In some
+  /// cases, such as induction variables, we will create both vector and scalar
+  /// entries.) Then, as we encounter uses of the definition, we derive values
+  /// for each scalar or vector use unless such a value is already available.
+  /// For example, if we scalarize a definition and one of its uses is vector,
+  /// we build the required vector on-demand with an insertelement sequence
+  /// when visiting the use. Otherwise, if the use is scalar, we can use the
+  /// existing scalar definition.
+  ///
+  /// Return a value in the new loop corresponding to \p V from the original
+  /// loop at unroll index \p Part. If the value has already been vectorized,
+  /// the corresponding vector entry in VectorLoopValueMap is returned. If,
   /// however, the value has a scalar entry in VectorLoopValueMap, we construct
-  /// new vector values on-demand by inserting the scalar values into vectors
+  /// a new vector value on-demand by inserting the scalar values into a vector
   /// with an insertelement sequence. If the value has been neither vectorized
   /// nor scalarized, it must be loop invariant, so we simply broadcast the
-  /// value into vectors.
-  const VectorParts &getVectorValue(Value *V);
+  /// value into a vector.
+  Value *getOrCreateVectorValue(Value *V, unsigned Part);
 
   /// Return a value in the new loop corresponding to \p V from the original
   /// loop at unroll index \p Part and vector index \p Lane. If the value has
   /// been vectorized but not scalarized, the necessary extractelement
   /// instruction will be generated.
-  Value *getScalarValue(Value *V, unsigned Part, unsigned Lane);
+  Value *getOrCreateScalarValue(Value *V, unsigned Part, unsigned Lane);
 
   /// Try to vectorize the interleaved access group that \p Instr belongs to.
   void vectorizeInterleaveGroup(Instruction *Instr);
 
   /// Generate a shuffle sequence that will reverse the vector Vec.
   virtual Value *reverseVector(Value *Vec);
 
   /// Returns (and creates if needed) the original loop trip count.
protected:
   /// This is a helper class for maintaining vectorization state. It's used for
   /// mapping values from the original loop to their corresponding values in
   /// the new loop. Two mappings are maintained: one for vectorized values and
   /// one for scalarized values. Vectorized values are represented with UF
   /// vector values in the new loop, and scalarized values are represented with
   /// UF x VF scalar values in the new loop. UF and VF are the unroll and
   /// vectorization factors, respectively.
   ///
-  /// Entries can be added to either map with initVector and initScalar, which
-  /// initialize and return a constant reference to the new entry. If a
-  /// non-constant reference to a vector entry is required, getVector can be
-  /// used to retrieve a mutable entry. We currently directly modify the mapped
-  /// values during "fix-up" operations that occur once the first phase of
-  /// widening is complete. These operations include type truncation and the
-  /// second phase of recurrence widening.
+  /// Entries can be added to either map with setVectorValue and setScalarValue,
+  /// which assert that an entry was not already added before. If an entry is to
+  /// replace an existing one, call resetVectorValue. This is currently needed
+  /// to modify the mapped values during "fix-up" operations that occur once the
+  /// first phase of widening is complete. These operations include type
+  /// truncation and the second phase of recurrence widening.
   ///
-  /// Otherwise, entries from either map should be accessed using the
-  /// getVectorValue or getScalarValue functions from InnerLoopVectorizer.
-  /// getVectorValue and getScalarValue coordinate to generate a vector or
-  /// scalar value on-demand if one is not yet available. When vectorizing a
-  /// loop, we visit the definition of an instruction before its uses. When
-  /// visiting the definition, we either vectorize or scalarize the
-  /// instruction, creating an entry for it in the corresponding map. (In some
-  /// cases, such as induction variables, we will create both vector and scalar
-  /// entries.) Then, as we encounter uses of the definition, we derive values
-  /// for each scalar or vector use unless such a value is already available.
-  /// For example, if we scalarize a definition and one of its uses is vector,
-  /// we build the required vector on-demand with an insertelement sequence
-  /// when visiting the use. Otherwise, if the use is scalar, we can use the
-  /// existing scalar definition.
+  /// Entries from either map can be retrieved using the getVectorValue and
+  /// getScalarValue functions, which assert that the desired value exists.
+
   struct ValueMap {
 
     /// Construct an empty map with the given unroll and vectorization factors.
-    ValueMap(unsigned UnrollFactor, unsigned VecWidth)
-        : UF(UnrollFactor), VF(VecWidth) {
-      // The unroll and vectorization factors are only used in asserts builds
-      // to verify map entries are sized appropriately.
-      (void)UF;
-      (void)VF;
-    }
-
-    /// \return True if the map has a vector entry for \p Key.
-    bool hasVector(Value *Key) const { return VectorMapStorage.count(Key); }
-
-    /// \return True if the map has a scalar entry for \p Key.
-    bool hasScalar(Value *Key) const { return ScalarMapStorage.count(Key); }
-
-    /// \brief Map \p Key to the given VectorParts \p Entry, and return a
-    /// constant reference to the new vector map entry. The given key should
-    /// not already be in the map, and the given VectorParts should be
-    /// correctly sized for the current unroll factor.
-    const VectorParts &initVector(Value *Key, const VectorParts &Entry) {
-      assert(!hasVector(Key) && "Vector entry already initialized");
-      assert(Entry.size() == UF && "VectorParts has wrong dimensions");
+    ValueMap(unsigned UF, unsigned VF) : UF(UF), VF(VF) {}
+
+    /// \return True if the map has any vector entry for \p Key.
+    bool hasAnyVectorValue(Value *Key) const {
+      return VectorMapStorage.count(Key);
+    }
+
+    /// \return True if the map has a vector entry for \p Key and \p Part.
+    bool hasVectorValue(Value *Key, unsigned Part) const {
+      assert(Part < UF && "Queried Vector Part is too large.");
+      if (!hasAnyVectorValue(Key))
+        return false;
+      const VectorParts &Entry = VectorMapStorage.find(Key)->second;
+      assert(Entry.size() == UF && "VectorParts has wrong dimensions.");
+      return Entry[Part] != nullptr;
+    }
+
+    /// \return True if the map has any scalar entry for \p Key.
+    bool hasAnyScalarValue(Value *Key) const {
+      return ScalarMapStorage.count(Key);
+    }
+
+    /// \return True if the map has a scalar entry for \p Key, \p Part and
+    /// \p Part.
+    bool hasScalarValue(Value *Key, unsigned Part, unsigned Lane) const {
+      assert(Part < UF && "Queried Scalar Part is too large.");
+      assert(Lane < VF && "Queried Scalar Lane is too large.");
+      if (!hasAnyScalarValue(Key))
+        return false;
+      const ScalarParts &Entry = ScalarMapStorage.find(Key)->second;
+      assert(Entry.size() == UF && "ScalarParts has wrong dimensions.");
+      assert(Entry[Part].size() == VF && "ScalarParts has wrong dimensions.");
+      return Entry[Part][Lane] != nullptr;
+    }
+
+    /// Retrieve the existing vector value that corresponds to \p Key and
+    /// \p Part.
+    Value *getVectorValue(Value *Key, unsigned Part) {
+      assert(hasVectorValue(Key, Part) && "Getting non-existent value.");
+      return VectorMapStorage[Key][Part];
+    }
+
+    /// Retrieve the existing scalar value that corresponds to \p Key, \p Part
+    /// and \p Lane.
+    Value *getScalarValue(Value *Key, unsigned Part, unsigned Lane) {
+      assert(hasScalarValue(Key, Part, Lane) && "Getting non-existent value.");
+      return ScalarMapStorage[Key][Part][Lane];
+    }
+
+    /// Set a vector value associated with \p Key and \p Part. Assumes such a
+    /// value is not already set. If it is, use resetVectorValue() instead.
+    void setVectorValue(Value *Key, unsigned Part, Value *Vector) {
+      assert(!hasVectorValue(Key, Part) && "Vector value already set for part");
+      if (!VectorMapStorage.count(Key)) {
+        VectorParts Entry(UF);
         VectorMapStorage[Key] = Entry;
-      return VectorMapStorage[Key];
+      }
+      VectorMapStorage[Key][Part] = Vector;
     }
 
-    /// \brief Map \p Key to the given ScalarParts \p Entry, and return a
-    /// constant reference to the new scalar map entry. The given key should
-    /// not already be in the map, and the given ScalarParts should be
-    /// correctly sized for the current unroll and vectorization factors.
-    const ScalarParts &initScalar(Value *Key, const ScalarParts &Entry) {
-      assert(!hasScalar(Key) && "Scalar entry already initialized");
-      assert(Entry.size() == UF &&
-             all_of(make_range(Entry.begin(), Entry.end()),
-                    [&](const SmallVectorImpl<Value *> &Values) -> bool {
-                      return Values.size() == VF;
-                    }) &&
-             "ScalarParts has wrong dimensions");
+    /// Set a scalar value associated with \p Key for \p Part and \p Lane.
+    /// Assumes such a value is not already set.
+    void setScalarValue(Value *Key, unsigned Part, unsigned Lane,
+                        Value *Scalar) {
+      assert(!hasScalarValue(Key, Part, Lane) && "Scalar value already set");
+      if (!ScalarMapStorage.count(Key)) {
+        ScalarParts Entry(UF);
+        for (unsigned Part = 0; Part < UF; ++Part)
+          Entry[Part].resize(VF, nullptr);
+          // TODO: Consider storing uniform values only per-part, as they occupy
+          //       lane 0 only, keeping the other VF-1 redundant entries null.
         ScalarMapStorage[Key] = Entry;
-      return ScalarMapStorage[Key];
+      }
+      ScalarMapStorage[Key][Part][Lane] = Scalar;
     }
 
-    /// \return A reference to the vector map entry corresponding to \p Key.
-    /// The key should already be in the map. This function should only be used
-    /// when it's necessary to update values that have already been vectorized.
-    /// This is the case for "fix-up" operations including type truncation and
-    /// the second phase of recurrence vectorization. If a non-const reference
-    /// isn't required, getVectorValue should be used instead.
-    VectorParts &getVector(Value *Key) {
-      assert(hasVector(Key) && "Vector entry not initialized");
-      return VectorMapStorage.find(Key)->second;
-    }
-
-    /// Retrieve an entry from the vector or scalar maps. The preferred way to
-    /// access an existing mapped entry is with getVectorValue or
-    /// getScalarValue from InnerLoopVectorizer. Until those functions can be
-    /// moved inside ValueMap, we have to declare them as friends.
-    friend const VectorParts &InnerLoopVectorizer::getVectorValue(Value *V);
-    friend Value *InnerLoopVectorizer::getScalarValue(Value *V, unsigned Part,
-                                                      unsigned Lane);
+    /// Reset the vector value associated with \p Key for the given \p Part.
+    /// This function can be used to update values that have already been
+    /// vectorized. This is the case for "fix-up" operations including type
+    /// truncation and the second phase of recurrence vectorization.
+    void resetVectorValue(Value *Key, unsigned Part, Value *Vector) {
+      assert(hasVectorValue(Key, Part) && "Vector value not set for part");
+      VectorMapStorage[Key][Part] = Vector;
+    }
 
   private:
     /// The unroll factor. Each entry in the vector map contains UF vector
     /// values.
     unsigned UF;
 
     /// The vectorization factor. Each entry in the scalar map contains UF x VF
     /// scalar values.
Value *SplatVF = isa<Constant>(Mul)
                        : Builder.CreateVectorSplat(VF, Mul);
   Builder.restoreIP(CurrIP);
 
   // We may need to add the step a number of times, depending on the unroll
   // factor. The last of those goes into the PHI.
   PHINode *VecInd = PHINode::Create(SteppedStart->getType(), 2, "vec.ind",
                                     &*LoopVectorBody->getFirstInsertionPt());
   Instruction *LastInduction = VecInd;
-  VectorParts Entry(UF);
   for (unsigned Part = 0; Part < UF; ++Part) {
-    Entry[Part] = LastInduction;
+    VectorLoopValueMap.setVectorValue(EntryVal, Part, LastInduction);
+    if (isa<TruncInst>(EntryVal))
+      addMetadata(LastInduction, EntryVal);
     LastInduction = cast<Instruction>(addFastMathFlag(
         Builder.CreateBinOp(AddOp, LastInduction, SplatVF, "step.add")));
   }
-  VectorLoopValueMap.initVector(EntryVal, Entry);
-  if (isa<TruncInst>(EntryVal))
-    addMetadata(Entry, EntryVal);
 
   // Move the last step to the end of the latch block. This ensures consistent
   // placement of all induction updates.
   auto *LoopVectorLatch = LI->getLoopFor(LoopVectorBody)->getLoopLatch();
   auto *Br = cast<BranchInst>(LoopVectorLatch->getTerminator());
   auto *ICmp = cast<Instruction>(Br->getCondition());
   LastInduction->moveBefore(ICmp);
   LastInduction->setName("vec.ind.next");
if (Trunc) {
       Step = Builder.CreateTrunc(Step, TruncType);
     }
   }
 
   // If we haven't yet vectorized the induction variable, splat the scalar
   // induction variable, and build the necessary step vectors.
   if (!VectorizedIV) {
     Value *Broadcasted = getBroadcastInstrs(ScalarIV);
-    VectorParts Entry(UF);
-    for (unsigned Part = 0; Part < UF; ++Part)
-      Entry[Part] =
+    for (unsigned Part = 0; Part < UF; ++Part) {
+      Value *EntryPart =
           getStepVector(Broadcasted, VF * Part, Step, ID.getInductionOpcode());
-    VectorLoopValueMap.initVector(EntryVal, Entry);
+      VectorLoopValueMap.setVectorValue(EntryVal, Part, EntryPart);
       if (Trunc)
-      addMetadata(Entry, Trunc);
+        addMetadata(EntryPart, Trunc);
+    }
   }
 
   // If an induction variable is only used for counting loop iterations or
   // calculating addresses, it doesn't need to be widened. Create scalar steps
   // that can be used by instructions we will later scalarize. Note that the
   // addition of the scalar steps will not increase the number of instructions
   // in the loop in the common case prior to InstCombine. We will be trading
   // one vector extract for each scalar step.
void InnerLoopVectorizer::buildScalarSteps(Value *ScalarIV, Value *Step,
 
   // Determine the number of scalars we need to generate for each unroll
   // iteration. If EntryVal is uniform, we only need to generate the first
   // lane. Otherwise, we generate all VF values.
   unsigned Lanes =
     Cost->isUniformAfterVectorization(cast<Instruction>(EntryVal), VF) ? 1 : VF;
 
   // Compute the scalar steps and save the results in VectorLoopValueMap.
-  ScalarParts Entry(UF);
   for (unsigned Part = 0; Part < UF; ++Part) {
-    Entry[Part].resize(VF);
     for (unsigned Lane = 0; Lane < Lanes; ++Lane) {
       auto *StartIdx = getSignedIntOrFpConstant(ScalarIVTy, VF * Part + Lane);
       auto *Mul = addFastMathFlag(Builder.CreateBinOp(MulOp, StartIdx, Step));
       auto *Add = addFastMathFlag(Builder.CreateBinOp(AddOp, ScalarIV, Mul));
-      Entry[Part][Lane] = Add;
+      VectorLoopValueMap.setScalarValue(EntryVal, Part, Lane, Add);
     }
   }
-  VectorLoopValueMap.initScalar(EntryVal, Entry);
 }
 
 int LoopVectorizationLegality::isConsecutivePtr(Value *Ptr) {
 
   const ValueToValueMap &Strides = getSymbolicStrides() ? *getSymbolicStrides() :
     ValueToValueMap();
 
   int Stride = getPtrStride(PSE, Ptr, TheLoop, Strides, true, false);
   if (Stride == 1 || Stride == -1)
     return Stride;
   return 0;
 }
 
 bool LoopVectorizationLegality::isUniform(Value *V) {
   return LAI->isUniform(V);
 }
 
-const InnerLoopVectorizer::VectorParts &
-InnerLoopVectorizer::getVectorValue(Value *V) {
+Value *InnerLoopVectorizer::getOrCreateVectorValue(Value *V, unsigned Part) {
   assert(V != Induction && "The new induction variable should not be used.");
   assert(!V->getType()->isVectorTy() && "Can't widen a vector");
   assert(!V->getType()->isVoidTy() && "Type does not produce a value");
 
   // If we have a stride that is replaced by one, do it here.
   if (Legal->hasStride(V))
     V = ConstantInt::get(V->getType(), 1);
 
-  // If we have this scalar in the map, return it.
-  if (VectorLoopValueMap.hasVector(V))
-    return VectorLoopValueMap.VectorMapStorage[V];
+  // If we have a vector mapped to this value, return it.
+  if (VectorLoopValueMap.hasVectorValue(V, Part))
+    return VectorLoopValueMap.getVectorValue(V, Part);
 
   // If the value has not been vectorized, check if it has been scalarized
   // instead. If it has been scalarized, and we actually need the value in
   // vector form, we will construct the vector values on demand.
-  if (VectorLoopValueMap.hasScalar(V)) {
+  if (VectorLoopValueMap.hasAnyScalarValue(V)) {
 
-    // Initialize a new vector map entry.
-    VectorParts Entry(UF);
+    Value *ScalarValue = VectorLoopValueMap.getScalarValue(V, Part, 0);
 
     // If we've scalarized a value, that value should be an instruction.
     auto *I = cast<Instruction>(V);
 
     // If we aren't vectorizing, we can just copy the scalar map values over to
     // the vector map.
     if (VF == 1) {
-      for (unsigned Part = 0; Part < UF; ++Part)
-        Entry[Part] = getScalarValue(V, Part, 0);
-      return VectorLoopValueMap.initVector(V, Entry);
+      VectorLoopValueMap.setVectorValue(V, Part, ScalarValue);
+      return ScalarValue;
     }
 
     // Get the last scalar instruction we generated for V. If the value is
     // known to be uniform after vectorization, this corresponds to lane zero
     // of the last unroll iteration. Otherwise, the last instruction is the one
     // we created for the last vector lane of the last unroll iteration.
     unsigned LastLane = Cost->isUniformAfterVectorization(I, VF) ? 0 : VF - 1;
-    auto *LastInst = cast<Instruction>(getScalarValue(V, UF - 1, LastLane));
+    auto *LastInst =
+        cast<Instruction>(getOrCreateScalarValue(V, UF - 1, LastLane));
 
     // Set the insert point after the last scalarized instruction. This ensures
     // the insertelement sequence will directly follow the scalar definitions.
     auto OldIP = Builder.saveIP();
     auto NewIP = std::next(BasicBlock::iterator(LastInst));
     Builder.SetInsertPoint(&*NewIP);
 
     // However, if we are vectorizing, we need to construct the vector values.
     // If the value is known to be uniform after vectorization, we can just
     // broadcast the scalar value corresponding to lane zero for each unroll
     // iteration. Otherwise, we construct the vector values using insertelement
     // instructions. Since the resulting vectors are stored in
     // VectorLoopValueMap, we will only generate the insertelements once.
-    for (unsigned Part = 0; Part < UF; ++Part) {
-      Value *VectorValue = nullptr;
-      if (Cost->isUniformAfterVectorization(I, VF)) {
-        VectorValue = getBroadcastInstrs(getScalarValue(V, Part, 0));
-      } else {
-        VectorValue = UndefValue::get(VectorType::get(V->getType(), VF));
-        for (unsigned Lane = 0; Lane < VF; ++Lane)
-          VectorValue = Builder.CreateInsertElement(
-              VectorValue, getScalarValue(V, Part, Lane),
-              Builder.getInt32(Lane));
-      }
-      Entry[Part] = VectorValue;
-    }
+      VectorValue = getBroadcastInstrs(ScalarValue);
+            VectorValue, getOrCreateScalarValue(V, Part, Lane),
+    VectorLoopValueMap.setVectorValue(V, Part, VectorValue);
     Builder.restoreIP(OldIP);
-    return VectorLoopValueMap.initVector(V, Entry);
+    return VectorValue;
   }
 
   // If this scalar is unknown, assume that it is a constant or that it is
   // loop invariant. Broadcast V and save the value for future uses.
   Value *B = getBroadcastInstrs(V);
-  return VectorLoopValueMap.initVector(V, VectorParts(UF, B));
+  VectorLoopValueMap.setVectorValue(V, Part, B);
+  return B;
 }
 
-Value *InnerLoopVectorizer::getScalarValue(Value *V, unsigned Part,
+Value *InnerLoopVectorizer::getOrCreateScalarValue(Value *V, unsigned Part,
                                                    unsigned Lane) {
 
   // If the value is not an instruction contained in the loop, it should
   // already be scalar.
   if (OrigLoop->isLoopInvariant(V))
     return V;
 
-  assert(Lane > 0 ?
-         !Cost->isUniformAfterVectorization(cast<Instruction>(V), VF)
+  assert(Lane > 0 ? !Cost->isUniformAfterVectorization(cast<Instruction>(V), VF)
                   : true && "Uniform values only have lane zero");
 
   // If the value from the original loop has not been vectorized, it is
   // represented by UF x VF scalar values in the new loop. Return the requested
   // scalar value.
-  if (VectorLoopValueMap.hasScalar(V))
-    return VectorLoopValueMap.ScalarMapStorage[V][Part][Lane];
+  if (VectorLoopValueMap.hasScalarValue(V, Part, Lane))
+    return VectorLoopValueMap.getScalarValue(V, Part, Lane);
 
   // If the value has not been scalarized, get its entry in VectorLoopValueMap
   // for the given unroll part. If this entry is not a vector type (i.e., the
   // vectorization factor is one), there is no need to generate an
   // extractelement instruction.
-  auto *U = getVectorValue(V)[Part];
+  auto *U = getOrCreateVectorValue(V, Part);
   if (!U->getType()->isVectorTy()) {
     assert(VF == 1 && "Value not scalarized has non-vector type");
     return U;
   }
 
   // Otherwise, the value from the original loop has been vectorized and is
   // represented by UF vector values. Extract and return the requested scalar
   // value from the appropriate vector lane.
void InnerLoopVectorizer::vectorizeInterleaveGroup(Instruction *Instr) {
   // rather than directly getting the pointer for lane VF - 1, because the
   // pointer operand of the interleaved access is supposed to be uniform. For
   // uniform instructions, we're only required to generate a value for the
   // first vector lane in each unroll iteration.
   if (Group->isReverse())
     Index += (VF - 1) * Group->getFactor();
 
   for (unsigned Part = 0; Part < UF; Part++) {
-    Value *NewPtr = getScalarValue(Ptr, Part, 0);
+    Value *NewPtr = getOrCreateScalarValue(Ptr, Part, 0);
 
     // Notice current instruction could be any index. Need to adjust the address
     // to the member of index 0.
     //
     // E.g.  a = A[i+1];     // Member of index 1 (Current instruction)
     //       b = A[i];       // Member of index 0
     // Current pointer is pointed to A[i+1], adjust it to A[i].
     //
if (isa<LoadInst>(Instr)) {
     // wide loads.
     for (unsigned I = 0; I < InterleaveFactor; ++I) {
       Instruction *Member = Group->getMember(I);
 
       // Skip the gaps in the group.
       if (!Member)
         continue;
 
-      VectorParts Entry(UF);
       Constant *StrideMask = createStrideMask(Builder, I, InterleaveFactor, VF);
       for (unsigned Part = 0; Part < UF; Part++) {
         Value *StridedVec = Builder.CreateShuffleVector(
             NewLoads[Part], UndefVec, StrideMask, "strided.vec");
 
         // If this member has different type, cast the result type.
         if (Member->getType() != ScalarTy) {
           VectorType *OtherVTy = VectorType::get(Member->getType(), VF);
           StridedVec = Builder.CreateBitOrPointerCast(StridedVec, OtherVTy);
         }
 
-        Entry[Part] =
-            Group->isReverse() ? reverseVector(StridedVec) : StridedVec;
+        if (Group->isReverse())
+          StridedVec = reverseVector(StridedVec);
+
+        VectorLoopValueMap.setVectorValue(Member, Part, StridedVec);
       }
-      VectorLoopValueMap.initVector(Member, Entry);
     }
     return;
   }
 
   // The sub vector type for current instruction.
   VectorType *SubVT = VectorType::get(ScalarTy, VF);
 
   // Vectorize the interleaved store group.
   for (unsigned Part = 0; Part < UF; Part++) {
     // Collect the stored vector from each member.
     SmallVector<Value *, 4> StoredVecs;
     for (unsigned i = 0; i < InterleaveFactor; i++) {
       // Interleaved store group doesn't allow a gap, so each index has a member
       Instruction *Member = Group->getMember(i);
       assert(Member && "Fail to get a member from an interleaved store group");
 
-      Value *StoredVec =
-          getVectorValue(cast<StoreInst>(Member)->getValueOperand())[Part];
+      Value *StoredVec = getOrCreateVectorValue(
+          cast<StoreInst>(Member)->getValueOperand(), Part);
       if (Group->isReverse())
         StoredVec = reverseVector(StoredVec);
 
       // If this member has different type, cast it to an unified type.
       if (StoredVec->getType() != SubVT)
         StoredVec = Builder.CreateBitOrPointerCast(StoredVec, SubVT);
 
       StoredVecs.push_back(StoredVec);
void InnerLoopVectorizer::vectorizeMemoryInstruction(Instruction *Instr) {
 
   // Determine if the pointer operand of the access is either consecutive or
   // reverse consecutive.
   int ConsecutiveStride = Legal->isConsecutivePtr(Ptr);
   bool Reverse = ConsecutiveStride < 0;
   bool CreateGatherScatter =
       (Decision == LoopVectorizationCostModel::CM_GatherScatter);
 
-  VectorParts VectorGep;
+  // Either Ptr feeds a vector load/store, or a vector GEP should feed a vector
+  // gather/scatter. Otherwise Decision should have been to Scalarize.
+  assert((ConsecutiveStride || CreateGatherScatter) &&
+         "The instruction should be scalarized");
 
   // Handle consecutive loads/stores.
-  if (ConsecutiveStride) {
-    Ptr = getScalarValue(Ptr, 0, 0);
-  } else {
-    // At this point we should vector version of GEP for Gather or Scatter
-    assert(CreateGatherScatter && "The instruction should be scalarized");
-    VectorGep = getVectorValue(Ptr);
-  }
+  if (ConsecutiveStride)
+    Ptr = getOrCreateScalarValue(Ptr, 0, 0);
 
   VectorParts Mask = createBlockInMask(Instr->getParent());
   // Handle Stores:
   if (SI) {
     assert(!Legal->isUniform(SI->getPointerOperand()) &&
            "We do not allow storing to uniform addresses");
     setDebugLocFromInst(Builder, SI);
-    // We don't want to update the value in the map as it might be used in
-    // another expression. So don't use a reference type for "StoredVal".
-    VectorParts StoredVal = getVectorValue(SI->getValueOperand());
 
     for (unsigned Part = 0; Part < UF; ++Part) {
       Instruction *NewSI = nullptr;
+      Value *StoredVal = getOrCreateVectorValue(SI->getValueOperand(), Part);
       if (CreateGatherScatter) {
         Value *MaskPart = Legal->isMaskRequired(SI) ? Mask[Part] : nullptr;
-        NewSI = Builder.CreateMaskedScatter(StoredVal[Part], VectorGep[Part],
-                                            Alignment, MaskPart);
+        Value *VectorGep = getOrCreateVectorValue(Ptr, Part);
+        NewSI = Builder.CreateMaskedScatter(StoredVal, VectorGep, Alignment,
+                                            MaskPart);
       } else {
         // Calculate the pointer for the specific unroll-part.
         Value *PartPtr =
             Builder.CreateGEP(nullptr, Ptr, Builder.getInt32(Part * VF));
 
         if (Reverse) {
           // If we store to reverse consecutive memory locations, then we need
           // to reverse the order of elements in the stored value.
-          StoredVal[Part] = reverseVector(StoredVal[Part]);
+          StoredVal = reverseVector(StoredVal);
           // If the address is consecutive but reversed, then the
           // wide store needs to start at the last vector element.
           PartPtr =
               Builder.CreateGEP(nullptr, Ptr, Builder.getInt32(-Part * VF));
           PartPtr =
               Builder.CreateGEP(nullptr, PartPtr, Builder.getInt32(1 - VF));
           Mask[Part] = reverseVector(Mask[Part]);
         }
 
         Value *VecPtr =
             Builder.CreateBitCast(PartPtr, DataTy->getPointerTo(AddressSpace));
 
         if (Legal->isMaskRequired(SI))
-          NewSI = Builder.CreateMaskedStore(StoredVal[Part], VecPtr, Alignment,
+          NewSI = Builder.CreateMaskedStore(StoredVal, VecPtr, Alignment,
                                             Mask[Part]);
         else
-          NewSI =
-              Builder.CreateAlignedStore(StoredVal[Part], VecPtr, Alignment);
+          NewSI = Builder.CreateAlignedStore(StoredVal, VecPtr, Alignment);
       }
       addMetadata(NewSI, SI);
     }
     return;
   }
 
   // Handle loads.
   assert(LI && "Must have a load instruction");
   setDebugLocFromInst(Builder, LI);
-  VectorParts Entry(UF);
   for (unsigned Part = 0; Part < UF; ++Part) {
-    Instruction *NewLI;
+    Value *NewLI;
     if (CreateGatherScatter) {
       Value *MaskPart = Legal->isMaskRequired(LI) ? Mask[Part] : nullptr;
-      NewLI = Builder.CreateMaskedGather(VectorGep[Part], Alignment, MaskPart,
+      Value *VectorGep = getOrCreateVectorValue(Ptr, Part);
+      NewLI = Builder.CreateMaskedGather(VectorGep, Alignment, MaskPart,
                                          nullptr, "wide.masked.gather");
-      Entry[Part] = NewLI;
     } else {
       // Calculate the pointer for the specific unroll-part.
       Value *PartPtr =
           Builder.CreateGEP(nullptr, Ptr, Builder.getInt32(Part * VF));
 
       if (Reverse) {
         // If the address is consecutive but reversed, then the
         // wide load needs to start at the last vector element.
         PartPtr = Builder.CreateGEP(nullptr, Ptr, Builder.getInt32(-Part * VF));
         PartPtr = Builder.CreateGEP(nullptr, PartPtr, Builder.getInt32(1 - VF));
         Mask[Part] = reverseVector(Mask[Part]);
       }
 
       Value *VecPtr =
           Builder.CreateBitCast(PartPtr, DataTy->getPointerTo(AddressSpace));
       if (Legal->isMaskRequired(LI))
         NewLI = Builder.CreateMaskedLoad(VecPtr, Alignment, Mask[Part],
                                          UndefValue::get(DataTy),
                                          "wide.masked.load");
       else
         NewLI = Builder.CreateAlignedLoad(VecPtr, Alignment, "wide.load");
-      Entry[Part] = Reverse ? reverseVector(NewLI) : NewLI;
+      if (Reverse)
+        NewLI = reverseVector(NewLI);
     }
     addMetadata(NewLI, LI);
+    VectorLoopValueMap.setVectorValue(Instr, Part, NewLI);
   }
-  VectorLoopValueMap.initVector(Instr, Entry);
 }
 
 void InnerLoopVectorizer::scalarizeInstruction(Instruction *Instr,
                                                bool IfPredicateInstr) {
   assert(!Instr->getType()->isAggregateType() && "Can't handle vectors");
   DEBUG(dbgs() << "LV: Scalarizing"
                << (IfPredicateInstr ? " and predicating:" : ":") << *Instr
                << '\n');
   // Holds vector parameters or scalars, in case of uniform vals.
   SmallVector<VectorParts, 4> Params;
 
   setDebugLocFromInst(Builder, Instr);
 
   // Does this instruction return a value ?
   bool IsVoidRetTy = Instr->getType()->isVoidTy();
 
-  // Initialize a new scalar map entry.
-  ScalarParts Entry(UF);
-
   VectorParts Cond;
   if (IfPredicateInstr)
     Cond = createBlockInMask(Instr->getParent());
 
   // Determine the number of scalars we need to generate for each unroll
   // iteration. If the instruction is uniform, we only need to generate the
   // first lane. Otherwise, we generate all VF values.
   unsigned Lanes = Cost->isUniformAfterVectorization(Instr, VF) ? 1 : VF;
 
   // For each vector unroll 'part':
   for (unsigned Part = 0; Part < UF; ++Part) {
-    Entry[Part].resize(VF);
     // For each scalar that we create:
     for (unsigned Lane = 0; Lane < Lanes; ++Lane) {
 
       // Start if-block.
       Value *Cmp = nullptr;
       if (IfPredicateInstr) {
         Cmp = Cond[Part];
         if (Cmp->getType()->isVectorTy())
           Cmp = Builder.CreateExtractElement(Cmp, Builder.getInt32(Lane));
         Cmp = Builder.CreateICmp(ICmpInst::ICMP_EQ, Cmp,
                                  ConstantInt::get(Cmp->getType(), 1));
       }
 
       Instruction *Cloned = Instr->clone();
       if (!IsVoidRetTy)
         Cloned->setName(Instr->getName() + ".cloned");
 
       // Replace the operands of the cloned instructions with their scalar
       // equivalents in the new loop.
       for (unsigned op = 0, e = Instr->getNumOperands(); op != e; ++op) {
-        auto *NewOp = getScalarValue(Instr->getOperand(op), Part, Lane);
+        auto *NewOp = getOrCreateScalarValue(Instr->getOperand(op), Part, Lane);
         Cloned->setOperand(op, NewOp);
       }
       addNewMetadata(Cloned, Instr);
 
       // Place the cloned scalar in the new loop.
       Builder.Insert(Cloned);
 
       // Add the cloned scalar to the scalar map entry.
-      Entry[Part][Lane] = Cloned;
+      VectorLoopValueMap.setScalarValue(Instr, Part, Lane, Cloned);
 
       // If we just cloned a new assumption, add it the assumption cache.
       if (auto *II = dyn_cast<IntrinsicInst>(Cloned))
         if (II->getIntrinsicID() == Intrinsic::assume)
           AC->registerAssumption(II);
 
       // End if-block.
       if (IfPredicateInstr)
         PredicatedInstructions.push_back(std::make_pair(Cloned, Cmp));
     }
   }
-  VectorLoopValueMap.initScalar(Instr, Entry);
 }
 
 PHINode *InnerLoopVectorizer::createInductionVariable(Loop *L, Value *Start,
                                                       Value *End, Value *Step,
                                                       Instruction *DL) {
   BasicBlock *Header = L->getHeader();
   BasicBlock *Latch = L->getLoopLatch();
   // As we're just creating this loop, it's possible no latch exists
void InnerLoopVectorizer::truncateToMinimalBitwidths() {
   // truncated version of `I` and reextend its result. InstCombine runs
   // later and will remove any ext/trunc pairs.
   //
   SmallPtrSet<Value *, 4> Erased;
   for (const auto &KV : Cost->getMinimalBitwidths()) {
     // If the value wasn't vectorized, we must maintain the original scalar
     // type. The absence of the value from VectorLoopValueMap indicates that it
     // wasn't vectorized.
-    if (!VectorLoopValueMap.hasVector(KV.first))
+    if (!VectorLoopValueMap.hasAnyVectorValue(KV.first))
       continue;
-    VectorParts &Parts = VectorLoopValueMap.getVector(KV.first);
-    for (Value *&I : Parts) {
+    for (unsigned Part = 0; Part < UF; ++Part) {
+      Value *I = getOrCreateVectorValue(KV.first, Part);
       if (Erased.count(I) || I->use_empty() || !isa<Instruction>(I))
         continue;
       Type *OriginalTy = I->getType();
       Type *ScalarTruncatedTy =
           IntegerType::get(OriginalTy->getContext(), KV.second);
       Type *TruncatedTy = VectorType::get(ScalarTruncatedTy,
                                           OriginalTy->getVectorNumElements());
       if (TruncatedTy == OriginalTy)
for (unsigned Part = 0; Part < UF; ++Part) {
       }
 
       // Lastly, extend the result.
       NewI->takeName(cast<Instruction>(I));
       Value *Res = B.CreateZExtOrTrunc(NewI, OriginalTy);
       I->replaceAllUsesWith(Res);
       cast<Instruction>(I)->eraseFromParent();
       Erased.insert(I);
-      I = Res;
+      VectorLoopValueMap.resetVectorValue(KV.first, Part, Res);
     }
   }
 
   // We'll have created a bunch of ZExts that are now parentless. Clean up.
   for (const auto &KV : Cost->getMinimalBitwidths()) {
     // If the value wasn't vectorized, we must maintain the original scalar
     // type. The absence of the value from VectorLoopValueMap indicates that it
     // wasn't vectorized.
-    if (!VectorLoopValueMap.hasVector(KV.first))
+    if (!VectorLoopValueMap.hasAnyVectorValue(KV.first))
       continue;
-    VectorParts &Parts = VectorLoopValueMap.getVector(KV.first);
-    for (Value *&I : Parts) {
+    for (unsigned Part = 0; Part < UF; ++Part) {
+      Value *I = getOrCreateVectorValue(KV.first, Part);
       ZExtInst *Inst = dyn_cast<ZExtInst>(I);
       if (Inst && Inst->use_empty()) {
         Value *NewI = Inst->getOperand(0);
         Inst->eraseFromParent();
-        I = NewI;
+        VectorLoopValueMap.resetVectorValue(KV.first, Part, NewI);
       }
     }
   }
 }
 
 void InnerLoopVectorizer::fixVectorizedLoop() {
   // Insert truncates and extends for any truncated instructions as hints to
   // InstCombine.
if (VF > 1) {
     Builder.SetInsertPoint(LoopVectorPreHeader->getTerminator());
     VectorInit = Builder.CreateInsertElement(
         UndefValue::get(VectorType::get(VectorInit->getType(), VF)), VectorInit,
         Builder.getInt32(VF - 1), "vector.recur.init");
   }
 
   // We constructed a temporary phi node in the first phase of vectorization.
   // This phi node will eventually be deleted.
-  VectorParts &PhiParts = VectorLoopValueMap.getVector(Phi);
-  Builder.SetInsertPoint(cast<Instruction>(PhiParts[0]));
+  Builder.SetInsertPoint(
+      cast<Instruction>(VectorLoopValueMap.getVectorValue(Phi, 0)));
 
   // Create a phi node for the new recurrence. The current value will either be
   // the initial value inserted into a vector or loop-varying vector value.
   auto *VecPhi = Builder.CreatePHI(VectorInit->getType(), 2, "vector.recur");
   VecPhi->addIncoming(VectorInit, LoopVectorPreHeader);
 
   // Get the vectorized previous value.
-  auto &PreviousParts = getVectorValue(Previous);
+  Value *PreviousLastPart = getOrCreateVectorValue(Previous, UF - 1);
 
   // Set the insertion point after the previous value if it is an instruction.
   // Note that the previous value may have been constant-folded so it is not
   // guaranteed to be an instruction in the vector loop. Also, if the previous
   // value is a phi node, we should insert after all the phi nodes to avoid
   // breaking basic block verification.
-  if (LI->getLoopFor(LoopVectorBody)->isLoopInvariant(PreviousParts[UF - 1]) ||
-      isa<PHINode>(PreviousParts[UF - 1]))
+  if (LI->getLoopFor(LoopVectorBody)->isLoopInvariant(PreviousLastPart) ||
+      isa<PHINode>(PreviousLastPart))
     Builder.SetInsertPoint(&*LoopVectorBody->getFirstInsertionPt());
   else
     Builder.SetInsertPoint(
-        &*++BasicBlock::iterator(cast<Instruction>(PreviousParts[UF - 1])));
+        &*++BasicBlock::iterator(cast<Instruction>(PreviousLastPart)));
 
   // We will construct a vector for the recurrence by combining the values for
   // the current and previous iterations. This is the required shuffle mask.
   SmallVector<Constant *, 8> ShuffleMask(VF);
   ShuffleMask[0] = Builder.getInt32(VF - 1);
   for (unsigned I = 1; I < VF; ++I)
     ShuffleMask[I] = Builder.getInt32(I + VF - 1);
 
   // The vector from which to take the initial value for the current iteration
   // (actual or unrolled). Initially, this is the vector phi node.
   Value *Incoming = VecPhi;
 
   // Shuffle the current and previous vector and update the vector parts.
   for (unsigned Part = 0; Part < UF; ++Part) {
+    Value *PreviousPart = getOrCreateVectorValue(Previous, Part);
+    Value *PhiPart = VectorLoopValueMap.getVectorValue(Phi, Part);
     auto *Shuffle =
-        VF > 1
-            ? Builder.CreateShuffleVector(Incoming, PreviousParts[Part],
+        VF > 1 ? Builder.CreateShuffleVector(Incoming, PreviousPart,
                                              ConstantVector::get(ShuffleMask))
                : Incoming;
-    PhiParts[Part]->replaceAllUsesWith(Shuffle);
-    cast<Instruction>(PhiParts[Part])->eraseFromParent();
-    PhiParts[Part] = Shuffle;
-    Incoming = PreviousParts[Part];
+    PhiPart->replaceAllUsesWith(Shuffle);
+    cast<Instruction>(PhiPart)->eraseFromParent();
+    VectorLoopValueMap.resetVectorValue(Phi, Part, Shuffle);
+    Incoming = PreviousPart;
   }
 
   // Fix the latch value of the new recurrence in the vector loop.
   VecPhi->addIncoming(Incoming, LI->getLoopFor(LoopVectorBody)->getLoopLatch());
 
   // Extract the last vector element in the middle block. This will be the
   // initial value for the recurrence when jumping to the scalar loop.
   auto *ExtractForScalar = Incoming;
void InnerLoopVectorizer::fixFirstOrderRecurrence(PHINode *Phi) {
   if (VF > 1)
     ExtractForPhiUsedOutsideLoop = Builder.CreateExtractElement(
         Incoming, Builder.getInt32(VF - 2), "vector.recur.extract.for.phi");
   // When loop is unrolled without vectorizing, initialize
   // ExtractForPhiUsedOutsideLoop with the value just prior to unrolled value of
   // `Incoming`. This is analogous to the vectorized case above: extracting the
   // second last element when VF > 1.
   else if (UF > 1)
-    ExtractForPhiUsedOutsideLoop = PreviousParts[UF - 2];
+    ExtractForPhiUsedOutsideLoop = getOrCreateVectorValue(Previous, UF - 2);
 
   // Fix the initial value of the original recurrence in the scalar loop.
   Builder.SetInsertPoint(&*LoopScalarPreHeader->begin());
   auto *Start = Builder.CreatePHI(Phi->getType(), 2, "scalar.recur.init");
   for (auto *BB : predecessors(LoopScalarPreHeader)) {
     auto *Incoming = BB == LoopMiddleBlock ? ExtractForScalar : ScalarInit;
     Start->addIncoming(Incoming, BB);
   }
void InnerLoopVectorizer::fixReduction(PHINode *Phi) {
 
   // We need to generate a reduction vector from the incoming scalar.
   // To do so, we need to generate the 'identity' vector and override
   // one of the elements with the incoming scalar reduction. We need
   // to do it in the vector-loop preheader.
   Builder.SetInsertPoint(LoopBypassBlocks[1]->getTerminator());
 
   // This is the vector-clone of the value that leaves the loop.
-  const VectorParts &VectorExit = getVectorValue(LoopExitInst);
-  Type *VecTy = VectorExit[0]->getType();
+  Type *VecTy = getOrCreateVectorValue(LoopExitInst, 0)->getType();
 
   // Find the reduction identity variable. Zero for addition, or, xor,
   // one for multiplication, -1 for And.
   Value *Identity;
   Value *VectorStart;
   if (RK == RecurrenceDescriptor::RK_IntegerMinMax ||
       RK == RecurrenceDescriptor::RK_FloatMinMax) {
     // MinMax reduction have the start value as their identify.
if (VF == 1) {
         Builder.CreateInsertElement(Identity, ReductionStartValue, Zero);
     }
   }
 
   // Fix the vector-loop phi.
 
   // Reductions do not have to start at zero. They can start with
   // any loop invariant values.
-  const VectorParts &VecRdxPhi = getVectorValue(Phi);
   BasicBlock *Latch = OrigLoop->getLoopLatch();
   Value *LoopVal = Phi->getIncomingValueForBlock(Latch);
-  const VectorParts &Val = getVectorValue(LoopVal);
-  for (unsigned part = 0; part < UF; ++part) {
+  for (unsigned Part = 0; Part < UF; ++Part) {
+    Value *VecRdxPhi = getOrCreateVectorValue(Phi, Part);
+    Value *Val = getOrCreateVectorValue(LoopVal, Part);
     // Make sure to add the reduction stat value only to the
     // first unroll part.
-    Value *StartVal = (part == 0) ? VectorStart : Identity;
-    cast<PHINode>(VecRdxPhi[part])
-      ->addIncoming(StartVal, LoopVectorPreHeader);
-    cast<PHINode>(VecRdxPhi[part])
-      ->addIncoming(Val[part], LI->getLoopFor(LoopVectorBody)->getLoopLatch());
+    Value *StartVal = (Part == 0) ? VectorStart : Identity;
+    cast<PHINode>(VecRdxPhi)->addIncoming(StartVal, LoopVectorPreHeader);
+    cast<PHINode>(VecRdxPhi)
+      ->addIncoming(Val, LI->getLoopFor(LoopVectorBody)->getLoopLatch());
   }
 
   // Before each round, move the insertion point right between
   // the PHIs and the values we are going to write.
   // This allows us to write both PHINodes and the extractelement
   // instructions.
   Builder.SetInsertPoint(&*LoopMiddleBlock->getFirstInsertionPt());
 
-  VectorParts &RdxParts = VectorLoopValueMap.getVector(LoopExitInst);
   setDebugLocFromInst(Builder, LoopExitInst);
 
   // If the vector reduction can be performed in a smaller type, we truncate
   // then extend the loop exit value to enable InstCombine to evaluate the
   // entire expression in the smaller type.
   if (VF > 1 && Phi->getType() != RdxDesc.getRecurrenceType()) {
     Type *RdxVecTy = VectorType::get(RdxDesc.getRecurrenceType(), VF);
     Builder.SetInsertPoint(LoopVectorBody->getTerminator());
-    for (unsigned part = 0; part < UF; ++part) {
-      Value *Trunc = Builder.CreateTrunc(RdxParts[part], RdxVecTy);
+    VectorParts RdxParts(UF);
+    for (unsigned Part = 0; Part < UF; ++Part) {
+      RdxParts[Part] = VectorLoopValueMap.getVectorValue(LoopExitInst, Part);
+      Value *Trunc = Builder.CreateTrunc(RdxParts[Part], RdxVecTy);
       Value *Extnd = RdxDesc.isSigned() ? Builder.CreateSExt(Trunc, VecTy)
                                         : Builder.CreateZExt(Trunc, VecTy);
-      for (Value::user_iterator UI = RdxParts[part]->user_begin();
-           UI != RdxParts[part]->user_end();)
+      for (Value::user_iterator UI = RdxParts[Part]->user_begin();
+           UI != RdxParts[Part]->user_end();)
         if (*UI != Trunc) {
-          (*UI++)->replaceUsesOfWith(RdxParts[part], Extnd);
-          RdxParts[part] = Extnd;
+          (*UI++)->replaceUsesOfWith(RdxParts[Part], Extnd);
+          RdxParts[Part] = Extnd;
         } else {
           ++UI;
         }
     }
     Builder.SetInsertPoint(&*LoopMiddleBlock->getFirstInsertionPt());
-    for (unsigned part = 0; part < UF; ++part)
-      RdxParts[part] = Builder.CreateTrunc(RdxParts[part], RdxVecTy);
+    for (unsigned Part = 0; Part < UF; ++Part) {
+      RdxParts[Part] = Builder.CreateTrunc(RdxParts[Part], RdxVecTy);
+      VectorLoopValueMap.resetVectorValue(LoopExitInst, Part, RdxParts[Part]);
+    }
   }
 
   // Reduce all of the unrolled parts into a single vector.
-  Value *ReducedPartRdx = RdxParts[0];
+  Value *ReducedPartRdx = VectorLoopValueMap.getVectorValue(LoopExitInst, 0);
   unsigned Op = RecurrenceDescriptor::getRecurrenceBinOp(RK);
   setDebugLocFromInst(Builder, ReducedPartRdx);
-  for (unsigned part = 1; part < UF; ++part) {
+  for (unsigned Part = 1; Part < UF; ++Part) {
+    Value *RdxPart = VectorLoopValueMap.getVectorValue(LoopExitInst, Part);
     if (Op != Instruction::ICmp && Op != Instruction::FCmp)
       // Floating point operations had to be 'fast' to enable the reduction.
       ReducedPartRdx = addFastMathFlag(
-          Builder.CreateBinOp((Instruction::BinaryOps)Op, RdxParts[part],
+          Builder.CreateBinOp((Instruction::BinaryOps)Op, RdxPart,
                               ReducedPartRdx, "bin.rdx"));
     else
       ReducedPartRdx = RecurrenceDescriptor::createMinMaxOp(
-          Builder, MinMaxKind, ReducedPartRdx, RdxParts[part]);
+          Builder, MinMaxKind, ReducedPartRdx, RdxPart);
   }
 
   if (VF > 1) {
     bool NoNaN = Legal->hasFunNoNaNAttr();
     ReducedPartRdx =
         createTargetReduction(Builder, TTI, RdxDesc, ReducedPartRdx, NoNaN);
     // If the reduction can be performed in a smaller type, we need to extend
     // the reduction to the wider type before we branch to the original loop.
InnerLoopVectorizer::createEdgeMask(BasicBlock *Src, BasicBlock *Dst) {
 
   VectorParts SrcMask = createBlockInMask(Src);
 
   // The terminator has to be a branch inst!
   BranchInst *BI = dyn_cast<BranchInst>(Src->getTerminator());
   assert(BI && "Unexpected terminator found");
 
   if (BI->isConditional()) {
-    VectorParts EdgeMask = getVectorValue(BI->getCondition());
 
+    VectorParts EdgeMask(UF);
+    for (unsigned Part = 0; Part < UF; ++Part) {
+      auto *EdgeMaskPart = getOrCreateVectorValue(BI->getCondition(), Part);
       if (BI->getSuccessor(0) != Dst)
-      for (unsigned part = 0; part < UF; ++part)
-        EdgeMask[part] = Builder.CreateNot(EdgeMask[part]);
+        EdgeMaskPart = Builder.CreateNot(EdgeMaskPart);
 
-    for (unsigned part = 0; part < UF; ++part)
-      EdgeMask[part] = Builder.CreateAnd(EdgeMask[part], SrcMask[part]);
+      EdgeMaskPart = Builder.CreateAnd(EdgeMaskPart, SrcMask[Part]);
+      EdgeMask[Part] = EdgeMaskPart;
+    }
 
     EdgeMaskCache[Edge] = EdgeMask;
     return EdgeMask;
   }
 
   EdgeMaskCache[Edge] = SrcMask;
   return SrcMask;
 }
 
 InnerLoopVectorizer::VectorParts
 InnerLoopVectorizer::createBlockInMask(BasicBlock *BB) {
   assert(OrigLoop->contains(BB) && "Block is not a part of a loop");
 
   // Look for cached value.
   BlockMaskCacheTy::iterator BCEntryIt = BlockMaskCache.find(BB);
   if (BCEntryIt != BlockMaskCache.end())
     return BCEntryIt->second;
 
+  VectorParts BlockMask(UF);
+
   // Loop incoming mask is all-one.
   if (OrigLoop->getHeader() == BB) {
     Value *C = ConstantInt::get(IntegerType::getInt1Ty(BB->getContext()), 1);
-    const VectorParts &BlockMask = getVectorValue(C);
+    for (unsigned Part = 0; Part < UF; ++Part)
+      BlockMask[Part] = getOrCreateVectorValue(C, Part);
     BlockMaskCache[BB] = BlockMask;
     return BlockMask;
   }
 
   // This is the block mask. We OR all incoming edges, and with zero.
   Value *Zero = ConstantInt::get(IntegerType::getInt1Ty(BB->getContext()), 0);
-  VectorParts BlockMask = getVectorValue(Zero);
+  for (unsigned Part = 0; Part < UF; ++Part)
+    BlockMask[Part] = getOrCreateVectorValue(Zero, Part);
 
   // For each pred:
-  for (pred_iterator it = pred_begin(BB), e = pred_end(BB); it != e; ++it) {
-    VectorParts EM = createEdgeMask(*it, BB);
-    for (unsigned part = 0; part < UF; ++part)
-      BlockMask[part] = Builder.CreateOr(BlockMask[part], EM[part]);
+  for (pred_iterator It = pred_begin(BB), E = pred_end(BB); It != E; ++It) {
+    VectorParts EM = createEdgeMask(*It, BB);
+    for (unsigned Part = 0; Part < UF; ++Part)
+      BlockMask[Part] = Builder.CreateOr(BlockMask[Part], EM[Part]);
   }
 
   BlockMaskCache[BB] = BlockMask;
   return BlockMask;
 }
 
 void InnerLoopVectorizer::widenPHIInstruction(Instruction *PN, unsigned UF,
                                               unsigned VF) {
   PHINode *P = cast<PHINode>(PN);
   // In order to support recurrences we need to be able to vectorize Phi nodes.
   // Phi nodes have cycles, so we need to vectorize them in two stages. This is
   // stage #1: We create a new vector PHI node with no incoming edges. We'll use
   // this value when we vectorize all of the instructions that use the PHI.
   if (Legal->isReductionVariable(P) || Legal->isFirstOrderRecurrence(P)) {
-    VectorParts Entry(UF);
-    for (unsigned part = 0; part < UF; ++part) {
+    for (unsigned Part = 0; Part < UF; ++Part) {
       // This is phase one of vectorizing PHIs.
       Type *VecTy =
           (VF == 1) ? PN->getType() : VectorType::get(PN->getType(), VF);
-      Entry[part] = PHINode::Create(
+      Value *EntryPart = PHINode::Create(
           VecTy, 2, "vec.phi", &*LoopVectorBody->getFirstInsertionPt());
+      VectorLoopValueMap.setVectorValue(P, Part, EntryPart);
     }
-    VectorLoopValueMap.initVector(P, Entry);
     return;
   }
 
   setDebugLocFromInst(Builder, P);
   // Check for PHI nodes that are lowered to vector selects.
   if (P->getParent() != OrigLoop->getHeader()) {
     // We know that all PHIs in non-header blocks are converted into
     // selects, so we don't have to worry about the insertion order and we
     // can just use the builder.
     // At this point we generate the predication tree. There may be
     // duplications since this is a simple recursive scan, but future
     // optimizations will clean it up.
 
     unsigned NumIncoming = P->getNumIncomingValues();
 
     // Generate a sequence of selects of the form:
     // SELECT(Mask3, In3,
     //      SELECT(Mask2, In2,
     //                   ( ...)))
     VectorParts Entry(UF);
     for (unsigned In = 0; In < NumIncoming; In++) {
       VectorParts Cond =
           createEdgeMask(P->getIncomingBlock(In), P->getParent());
-      const VectorParts &In0 = getVectorValue(P->getIncomingValue(In));
 
-      for (unsigned part = 0; part < UF; ++part) {
+      for (unsigned Part = 0; Part < UF; ++Part) {
+        Value *In0 = getOrCreateVectorValue(P->getIncomingValue(In), Part);
         // We might have single edge PHIs (blocks) - use an identity
         // 'select' for the first PHI operand.
         if (In == 0)
-          Entry[part] = Builder.CreateSelect(Cond[part], In0[part], In0[part]);
+          Entry[Part] = Builder.CreateSelect(Cond[Part], In0, In0);
         else
           // Select between the current value and the previous incoming edge
           // based on the incoming mask.
-          Entry[part] = Builder.CreateSelect(Cond[part], In0[part], Entry[part],
+          Entry[Part] = Builder.CreateSelect(Cond[Part], In0, Entry[Part],
                                              "predphi");
       }
     }
-    VectorLoopValueMap.initVector(P, Entry);
+    for (unsigned Part = 0; Part < UF; ++Part)
+      VectorLoopValueMap.setVectorValue(P, Part, Entry[Part]);
     return;
   }
 
   // This PHINode must be an induction variable.
   // Make sure that we know about it.
   assert(Legal->getInductionVars()->count(P) && "Not an induction variable");
 
   InductionDescriptor II = Legal->getInductionVars()->lookup(P);
case InductionDescriptor::IK_PtrInduction: {
     Value *PtrInd = Induction;
     PtrInd = Builder.CreateSExtOrTrunc(PtrInd, II.getStep()->getType());
     // Determine the number of scalars we need to generate for each unroll
     // iteration. If the instruction is uniform, we only need to generate the
     // first lane. Otherwise, we generate all VF values.
     unsigned Lanes = Cost->isUniformAfterVectorization(P, VF) ? 1 : VF;
     // These are the scalar results. Notice that we don't generate vector GEPs
     // because scalar GEPs result in better code.
-    ScalarParts Entry(UF);
     for (unsigned Part = 0; Part < UF; ++Part) {
-      Entry[Part].resize(VF);
       for (unsigned Lane = 0; Lane < Lanes; ++Lane) {
         Constant *Idx = ConstantInt::get(PtrInd->getType(), Lane + Part * VF);
         Value *GlobalIdx = Builder.CreateAdd(PtrInd, Idx);
         Value *SclrGep = II.transform(Builder, GlobalIdx, PSE.getSE(), DL);
         SclrGep->setName("next.gep");
-        Entry[Part][Lane] = SclrGep;
+        VectorLoopValueMap.setScalarValue(P, Part, Lane, SclrGep);
       }
     }
-    VectorLoopValueMap.initScalar(P, Entry);
     return;
   }
   }
 }
 
 /// A helper function for checking whether an integer division-related
 /// instruction may divide by zero (in which case it must be predicated if
 /// executed conditionally in the scalar code).
void InnerLoopVectorizer::vectorizeInstruction(Instruction &I) {
   } // End of PHI.
   case Instruction::GetElementPtr: {
     // Construct a vector GEP by widening the operands of the scalar GEP as
     // necessary. We mark the vector GEP 'inbounds' if appropriate. A GEP
     // results in a vector of pointers when at least one operand of the GEP
     // is vector-typed. Thus, to keep the representation compact, we only use
     // vector-typed operands for loop-varying values.
     auto *GEP = cast<GetElementPtrInst>(&I);
-    VectorParts Entry(UF);
 
     if (VF > 1 && OrigLoop->hasLoopInvariantOperands(GEP)) {
       // If we are vectorizing, but the GEP has only loop-invariant operands,
       // the GEP we build (by only using vector-typed operands for
       // loop-varying values) would be a scalar pointer. Thus, to ensure we
       // produce a vector of pointers, we need to either arbitrarily pick an
       // operand to broadcast, or broadcast a clone of the original GEP.
       // Here, we broadcast a clone of the original.
       //
       // TODO: If at some point we decide to scalarize instructions having
       //       loop-invariant operands, this special case will no longer be
       //       required. We would add the scalarization decision to
       //       collectLoopScalars() and teach getVectorValue() to broadcast
       //       the lane-zero scalar value.
       auto *Clone = Builder.Insert(GEP->clone());
-      for (unsigned Part = 0; Part < UF; ++Part)
-        Entry[Part] = Builder.CreateVectorSplat(VF, Clone);
+      for (unsigned Part = 0; Part < UF; ++Part) {
+        Value *EntryPart = Builder.CreateVectorSplat(VF, Clone);
+        VectorLoopValueMap.setVectorValue(&I, Part, EntryPart);
+        addMetadata(EntryPart, GEP);
+      }
     } else {
       // If the GEP has at least one loop-varying operand, we are sure to
       // produce a vector of pointers. But if we are only unrolling, we want
       // to produce a scalar GEP for each unroll part. Thus, the GEP we
       // produce with the code below will be scalar (if VF == 1) or vector
       // (otherwise). Note that for the unroll-only case, we still maintain
       // values in the vector mapping with initVector, as we do for other
       // instructions.
       for (unsigned Part = 0; Part < UF; ++Part) {
 
         // The pointer operand of the new GEP. If it's loop-invariant, we
         // won't broadcast it.
-        auto *Ptr = OrigLoop->isLoopInvariant(GEP->getPointerOperand())
-                        ? GEP->getPointerOperand()
-                        : getVectorValue(GEP->getPointerOperand())[Part];
+        auto *Ptr =
+            OrigLoop->isLoopInvariant(GEP->getPointerOperand())
+                : getOrCreateVectorValue(GEP->getPointerOperand(), Part);
 
         // Collect all the indices for the new GEP. If any index is
         // loop-invariant, we won't broadcast it.
         SmallVector<Value *, 4> Indices;
         for (auto &U : make_range(GEP->idx_begin(), GEP->idx_end())) {
           if (OrigLoop->isLoopInvariant(U.get()))
             Indices.push_back(U.get());
           else
-            Indices.push_back(getVectorValue(U.get())[Part]);
+            Indices.push_back(getOrCreateVectorValue(U.get(), Part));
         }
 
         // Create the new GEP. Note that this GEP may be a scalar if VF == 1,
         // but it should be a vector, otherwise.
         auto *NewGEP = GEP->isInBounds()
                            ? Builder.CreateInBoundsGEP(Ptr, Indices)
                            : Builder.CreateGEP(Ptr, Indices);
         assert((VF == 1 || NewGEP->getType()->isVectorTy()) &&
                "NewGEP is not a pointer vector");
-        Entry[Part] = NewGEP;
+        VectorLoopValueMap.setVectorValue(&I, Part, NewGEP);
+        addMetadata(NewGEP, GEP);
       }
     }
 
-    VectorLoopValueMap.initVector(&I, Entry);
-    addMetadata(Entry, GEP);
     break;
   }
   case Instruction::UDiv:
   case Instruction::SDiv:
   case Instruction::SRem:
   case Instruction::URem:
     // Scalarize with predication if this instruction may divide by zero and
     // block execution is conditional, otherwise fallthrough.
void InnerLoopVectorizer::vectorizeInstruction(Instruction &I) {
   case Instruction::LShr:
   case Instruction::AShr:
   case Instruction::And:
   case Instruction::Or:
   case Instruction::Xor: {
     // Just widen binops.
     auto *BinOp = cast<BinaryOperator>(&I);
     setDebugLocFromInst(Builder, BinOp);
-    const VectorParts &A = getVectorValue(BinOp->getOperand(0));
-    const VectorParts &B = getVectorValue(BinOp->getOperand(1));
 
-    // Use this vector value for all users of the original instruction.
-    VectorParts Entry(UF);
     for (unsigned Part = 0; Part < UF; ++Part) {
-      Value *V = Builder.CreateBinOp(BinOp->getOpcode(), A[Part], B[Part]);
+      Value *A = getOrCreateVectorValue(BinOp->getOperand(0), Part);
+      Value *B = getOrCreateVectorValue(BinOp->getOperand(1), Part);
+      Value *V = Builder.CreateBinOp(BinOp->getOpcode(), A, B);
 
       if (BinaryOperator *VecOp = dyn_cast<BinaryOperator>(V))
         VecOp->copyIRFlags(BinOp);
 
-      Entry[Part] = V;
+      // Use this vector value for all users of the original instruction.
+      VectorLoopValueMap.setVectorValue(&I, Part, V);
+      addMetadata(V, BinOp);
     }
 
-    VectorLoopValueMap.initVector(&I, Entry);
-    addMetadata(Entry, BinOp);
     break;
   }
   case Instruction::Select: {
     // Widen selects.
     // If the selector is loop invariant we can create a select
     // instruction with a scalar condition. Otherwise, use vector-select.
     auto *SE = PSE.getSE();
     bool InvariantCond =
         SE->isLoopInvariant(PSE.getSCEV(I.getOperand(0)), OrigLoop);
     setDebugLocFromInst(Builder, &I);
 
     // The condition can be loop invariant  but still defined inside the
     // loop. This means that we can't just use the original 'cond' value.
     // We have to take the 'vectorized' value and pick the first lane.
     // Instcombine will make this a no-op.
-    const VectorParts &Cond = getVectorValue(I.getOperand(0));
-    const VectorParts &Op0 = getVectorValue(I.getOperand(1));
-    const VectorParts &Op1 = getVectorValue(I.getOperand(2));
 
-    auto *ScalarCond = getScalarValue(I.getOperand(0), 0, 0);
+    auto *ScalarCond = getOrCreateScalarValue(I.getOperand(0), 0, 0);
 
-    VectorParts Entry(UF);
     for (unsigned Part = 0; Part < UF; ++Part) {
-      Entry[Part] = Builder.CreateSelect(
-          InvariantCond ? ScalarCond : Cond[Part], Op0[Part], Op1[Part]);
+      Value *Cond = getOrCreateVectorValue(I.getOperand(0), Part);
+      Value *Op0 = getOrCreateVectorValue(I.getOperand(1), Part);
+      Value *Op1 = getOrCreateVectorValue(I.getOperand(2), Part);
+      Value *Sel =
+          Builder.CreateSelect(InvariantCond ? ScalarCond : Cond, Op0, Op1);
+      VectorLoopValueMap.setVectorValue(&I, Part, Sel);
+      addMetadata(Sel, &I);
     }
 
-    VectorLoopValueMap.initVector(&I, Entry);
-    addMetadata(Entry, &I);
     break;
   }
 
   case Instruction::ICmp:
   case Instruction::FCmp: {
     // Widen compares. Generate vector compares.
     bool FCmp = (I.getOpcode() == Instruction::FCmp);
     auto *Cmp = dyn_cast<CmpInst>(&I);
     setDebugLocFromInst(Builder, Cmp);
-    const VectorParts &A = getVectorValue(Cmp->getOperand(0));
-    const VectorParts &B = getVectorValue(Cmp->getOperand(1));
-    VectorParts Entry(UF);
     for (unsigned Part = 0; Part < UF; ++Part) {
+      Value *A = getOrCreateVectorValue(Cmp->getOperand(0), Part);
+      Value *B = getOrCreateVectorValue(Cmp->getOperand(1), Part);
       Value *C = nullptr;
       if (FCmp) {
-        C = Builder.CreateFCmp(Cmp->getPredicate(), A[Part], B[Part]);
+        C = Builder.CreateFCmp(Cmp->getPredicate(), A, B);
         cast<FCmpInst>(C)->copyFastMathFlags(Cmp);
       } else {
-        C = Builder.CreateICmp(Cmp->getPredicate(), A[Part], B[Part]);
+        C = Builder.CreateICmp(Cmp->getPredicate(), A, B);
       }
-      Entry[Part] = C;
+      VectorLoopValueMap.setVectorValue(&I, Part, C);
+      addMetadata(C, &I);
     }
 
-    VectorLoopValueMap.initVector(&I, Entry);
-    addMetadata(Entry, &I);
     break;
   }
 
   case Instruction::Store:
   case Instruction::Load:
     vectorizeMemoryInstruction(&I);
     break;
   case Instruction::ZExt:
if (Cost->isOptimizableIVTruncate(CI, VF)) {
                             cast<TruncInst>(CI));
       break;
     }
 
     /// Vectorize casts.
     Type *DestTy =
         (VF == 1) ? CI->getType() : VectorType::get(CI->getType(), VF);
 
-    const VectorParts &A = getVectorValue(CI->getOperand(0));
-    VectorParts Entry(UF);
-    for (unsigned Part = 0; Part < UF; ++Part)
-      Entry[Part] = Builder.CreateCast(CI->getOpcode(), A[Part], DestTy);
-    VectorLoopValueMap.initVector(&I, Entry);
-    addMetadata(Entry, &I);
+    for (unsigned Part = 0; Part < UF; ++Part) {
+      Value *A = getOrCreateVectorValue(CI->getOperand(0), Part);
+      Value *Cast = Builder.CreateCast(CI->getOpcode(), A, DestTy);
+      VectorLoopValueMap.setVectorValue(&I, Part, Cast);
+      addMetadata(Cast, &I);
+    }
     break;
   }
 
   case Instruction::Call: {
     // Ignore dbg intrinsics.
     if (isa<DbgInfoIntrinsic>(I))
       break;
     setDebugLocFromInst(Builder, &I);
case Instruction::Call: {
     unsigned CallCost = getVectorCallCost(CI, VF, *TTI, TLI, NeedToScalarize);
     bool UseVectorIntrinsic =
         ID && getVectorIntrinsicCost(CI, VF, *TTI, TLI) <= CallCost;
     if (!UseVectorIntrinsic && NeedToScalarize) {
       scalarizeInstruction(&I);
       break;
     }
 
-    VectorParts Entry(UF);
     for (unsigned Part = 0; Part < UF; ++Part) {
       SmallVector<Value *, 4> Args;
       for (unsigned i = 0, ie = CI->getNumArgOperands(); i != ie; ++i) {
         Value *Arg = CI->getArgOperand(i);
         // Some intrinsics have a scalar argument - don't replace it with a
         // vector.
-        if (!UseVectorIntrinsic || !hasVectorInstrinsicScalarOpd(ID, i)) {
-          const VectorParts &VectorArg = getVectorValue(CI->getArgOperand(i));
-          Arg = VectorArg[Part];
-        }
+        if (!UseVectorIntrinsic || !hasVectorInstrinsicScalarOpd(ID, i))
+          Arg = getOrCreateVectorValue(CI->getArgOperand(i), Part);
         Args.push_back(Arg);
       }
 
       Function *VectorF;
       if (UseVectorIntrinsic) {
         // Use vector version of the intrinsic.
         Type *TysForDecl[] = {CI->getType()};
         if (VF > 1)
for (unsigned Part = 0; Part < UF; ++Part) {
 
       SmallVector<OperandBundleDef, 1> OpBundles;
       CI->getOperandBundlesAsDefs(OpBundles);
       CallInst *V = Builder.CreateCall(VectorF, Args, OpBundles);
 
       if (isa<FPMathOperator>(V))
         V->copyFastMathFlags(CI);
 
-      Entry[Part] = V;
+      VectorLoopValueMap.setVectorValue(&I, Part, V);
+      addMetadata(V, &I);
     }
 
-    VectorLoopValueMap.initVector(&I, Entry);
-    addMetadata(Entry, &I);
     break;
   }
 
   default:
     // All other instructions are unsupported. Scalarize them.
     scalarizeInstruction(&I);
     break;
   } // end of switch.