maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity

[SLP] Replace NeedToGather variable with enum.

This commit is contained in:
Dinar Temirbulatov 2019-12-23 08:21:18 +01:00
parent 9d1071eac4
commit a755ccefe6
1 changed files with 34 additions and 22 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

56
llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp
View File

@ -1432,7 +1432,8 @@ private:
Value *VectorizedValue = nullptr;
/// Do we need to gather this sequence ?
bool NeedToGather = false;
enum EntryState { Vectorize, NeedToGather };
EntryState State;
/// Does this sequence require some shuffling?
SmallVector<unsigned, 4> ReuseShuffleIndices;
@ -1574,7 +1575,15 @@ private:
dbgs() << "Scalars: \n";
for (Value *V : Scalars)
dbgs().indent(2) << *V << "\n";
dbgs() << "NeedToGather: " << NeedToGather << "\n";
dbgs() << "State: ";
switch (State) {
case Vectorize:
dbgs() << "Vectorize\n";
break;
case NeedToGather:
dbgs() << "NeedToGather\n";
break;
}
dbgs() << "MainOp: ";
if (MainOp)
dbgs() << *MainOp << "\n";
@ -1620,7 +1629,7 @@ private:
TreeEntry *Last = VectorizableTree.back().get();
Last->Idx = VectorizableTree.size() - 1;
Last->Scalars.insert(Last->Scalars.begin(), VL.begin(), VL.end());
Last->NeedToGather = !Vectorized;
Last->State = Vectorized ? TreeEntry::Vectorize : TreeEntry::NeedToGather;
Last->ReuseShuffleIndices.append(ReuseShuffleIndices.begin(),
ReuseShuffleIndices.end());
Last->ReorderIndices = ReorderIndices;
@ -2291,7 +2300,7 @@ template <> struct DOTGraphTraits<BoUpSLP *> : public DefaultDOTGraphTraits {
static std::string getNodeAttributes(const TreeEntry *Entry,
const BoUpSLP *) {
if (Entry->NeedToGather)
if (Entry->State == TreeEntry::NeedToGather)
return "color=red";
return "";
}
@ -2343,7 +2352,7 @@ void BoUpSLP::buildTree(ArrayRef<Value *> Roots,
TreeEntry *Entry = TEPtr.get();
// No need to handle users of gathered values.
if (Entry->NeedToGather)
if (Entry->State == TreeEntry::NeedToGather)
continue;
// For each lane:
@ -2380,7 +2389,7 @@ void BoUpSLP::buildTree(ArrayRef<Value *> Roots,
!InTreeUserNeedToExtract(Scalar, UserInst, TLI)) {
LLVM_DEBUG(dbgs() << "SLP: \tInternal user will be removed:" << *U
<< ".\n");
assert(!UseEntry->NeedToGather && "Bad state");
assert(UseEntry->State != TreeEntry::NeedToGather && "Bad state");
continue;
}
}
@ -3212,7 +3221,7 @@ int BoUpSLP::getEntryCost(TreeEntry *E) {
ReuseShuffleCost =
TTI->getShuffleCost(TargetTransformInfo::SK_PermuteSingleSrc, VecTy);
}
if (E->NeedToGather) {
if (E->State == TreeEntry::NeedToGather) {
if (allConstant(VL))
return 0;
if (isSplat(VL)) {
@ -3280,7 +3289,7 @@ int BoUpSLP::getEntryCost(TreeEntry *E) {
TTI->getVectorInstrCost(Instruction::ExtractElement, VecTy, Idx);
}
}
if (!E->NeedToGather) {
if (E->State == TreeEntry::Vectorize) {
int DeadCost = ReuseShuffleCost;
if (!E->ReorderIndices.empty()) {
// TODO: Merge this shuffle with the ReuseShuffleCost.
@ -3566,20 +3575,22 @@ bool BoUpSLP::isFullyVectorizableTinyTree() const {
<< VectorizableTree.size() << " is fully vectorizable .\n");
// We only handle trees of heights 1 and 2.
if (VectorizableTree.size() == 1 && !VectorizableTree[0]->NeedToGather)
if (VectorizableTree.size() == 1 &&
VectorizableTree[0]->State == TreeEntry::Vectorize)
return true;
if (VectorizableTree.size() != 2)
return false;
// Handle splat and all-constants stores.
if (!VectorizableTree[0]->NeedToGather &&
if (VectorizableTree[0]->State == TreeEntry::Vectorize &&
(allConstant(VectorizableTree[1]->Scalars) ||
isSplat(VectorizableTree[1]->Scalars)))
return true;
// Gathering cost would be too much for tiny trees.
if (VectorizableTree[0]->NeedToGather || VectorizableTree[1]->NeedToGather)
if (VectorizableTree[0]->State == TreeEntry::NeedToGather ||
VectorizableTree[1]->State == TreeEntry::NeedToGather)
return false;
return true;
@ -3733,12 +3744,13 @@ int BoUpSLP::getTreeCost() {
// their uses. Since such an approach results in fewer total entries,
// existing heuristics based on tree size may yield different results.
//
if (TE.NeedToGather &&
std::any_of(
std::next(VectorizableTree.begin(), I + 1), VectorizableTree.end(),
[TE](const std::unique_ptr<TreeEntry> &EntryPtr) {
return EntryPtr->NeedToGather && EntryPtr->isSame(TE.Scalars);
}))
if (TE.State == TreeEntry::NeedToGather &&
std::any_of(std::next(VectorizableTree.begin(), I + 1),
VectorizableTree.end(),
[TE](const std::unique_ptr<TreeEntry> &EntryPtr) {
return EntryPtr->State == TreeEntry::NeedToGather &&
EntryPtr->isSame(TE.Scalars);
}))
continue;
int C = getEntryCost(&TE);
@ -4029,7 +4041,7 @@ Value *BoUpSLP::vectorizeTree(TreeEntry *E) {
bool NeedToShuffleReuses = !E->ReuseShuffleIndices.empty();
if (E->NeedToGather) {
if (E->State == TreeEntry::NeedToGather) {
setInsertPointAfterBundle(E);
auto *V = Gather(E->Scalars, VecTy);
if (NeedToShuffleReuses) {
@ -4084,7 +4096,7 @@ Value *BoUpSLP::vectorizeTree(TreeEntry *E) {
}
case Instruction::ExtractElement: {
if (!E->NeedToGather) {
if (E->State == TreeEntry::Vectorize) {
Value *V = E->getSingleOperand(0);
if (!E->ReorderIndices.empty()) {
OrdersType Mask;
@ -4117,7 +4129,7 @@ Value *BoUpSLP::vectorizeTree(TreeEntry *E) {
return V;
}
case Instruction::ExtractValue: {
if (!E->NeedToGather) {
if (E->State == TreeEntry::Vectorize) {
LoadInst *LI = cast<LoadInst>(E->getSingleOperand(0));
Builder.SetInsertPoint(LI);
PointerType *PtrTy = PointerType::get(VecTy, LI->getPointerAddressSpace());
@ -4603,7 +4615,7 @@ BoUpSLP::vectorizeTree(ExtraValueToDebugLocsMap &ExternallyUsedValues) {
continue;
TreeEntry *E = getTreeEntry(Scalar);
assert(E && "Invalid scalar");
assert(!E->NeedToGather && "Extracting from a gather list");
assert(E->State == TreeEntry::Vectorize && "Extracting from a gather list");
Value *Vec = E->VectorizedValue;
assert(Vec && "Can't find vectorizable value");
@ -4676,7 +4688,7 @@ BoUpSLP::vectorizeTree(ExtraValueToDebugLocsMap &ExternallyUsedValues) {
TreeEntry *Entry = TEPtr.get();
// No need to handle users of gathered values.
if (Entry->NeedToGather)
if (Entry->State == TreeEntry::NeedToGather)
continue;
assert(Entry->VectorizedValue && "Can't find vectorizable value");