/// RewriteSingleStoreAlloca - If there is only a single store to this value,
/// replace any loads of it that are directly dominated by the definition with
/// the value stored.
void PromoteMem2Reg::RewriteSingleStoreAlloca(AllocaInst *AI,
AllocaInfo &Info,
LargeBlockInfo &LBI) {
StoreInst *OnlyStore = Info.OnlyStore;
bool StoringGlobalVal = !isa<Instruction>(OnlyStore->getOperand(0));
BasicBlock *StoreBB = OnlyStore->getParent();
int StoreIndex = -1;
// Clear out UsingBlocks. We will reconstruct it here if needed.
Info.UsingBlocks.clear();
for (Value::use_iterator UI = AI->use_begin(), E = AI->use_end(); UI != E; ) {
Instruction *UserInst = cast<Instruction>(*UI++);
if (!isa<LoadInst>(UserInst)) {
assert(UserInst == OnlyStore && "Should only have load/stores");
continue;
}
LoadInst *LI = cast<LoadInst>(UserInst);
// Okay, if we have a load from the alloca, we want to replace it with the
// only value stored to the alloca. We can do this if the value is
// dominated by the store. If not, we use the rest of the mem2reg machinery
// to insert the phi nodes as needed.
if (!StoringGlobalVal) { // Non-instructions are always dominated.
if (LI->getParent() == StoreBB) {
// If we have a use that is in the same block as the store, compare the
// indices of the two instructions to see which one came first. If the
// load came before the store, we can't handle it.
if (StoreIndex == -1)
StoreIndex = LBI.getInstructionIndex(OnlyStore);
if (unsigned(StoreIndex) > LBI.getInstructionIndex(LI)) {
// Can't handle this load, bail out.
Info.UsingBlocks.push_back(StoreBB);
continue;
}
} else if (LI->getParent() != StoreBB &&
!dominates(StoreBB, LI->getParent())) {
// If the load and store are in different blocks, use BB dominance to
// check their relationships. If the store doesn't dom the use, bail
// out.
Info.UsingBlocks.push_back(LI->getParent());
continue;
}
}
// Otherwise, we *can* safely rewrite this load.
Value *ReplVal = OnlyStore->getOperand(0);
// If the replacement value is the load, this must occur in unreachable
// code.
if (ReplVal == LI)
ReplVal = UndefValue::get(LI->getType());
LI->replaceAllUsesWith(ReplVal);
if (AST && LI->getType()->isPointerTy())
AST->deleteValue(LI);
LI->eraseFromParent();
LBI.deleteValue(LI);
}
}
/// PromoteSingleBlockAlloca - Many allocas are only used within a single basic
/// block. If this is the case, avoid traversing the CFG and inserting a lot of
/// potentially useless PHI nodes by just performing a single linear pass over
/// the basic block using the Alloca.
///
/// If we cannot promote this alloca (because it is read before it is written),
/// return true. This is necessary in cases where, due to control flow, the
/// alloca is potentially undefined on some control flow paths. e.g. code like
/// this is potentially correct:
///
/// for (...) { if (c) { A = undef; undef = B; } }
///
/// ... so long as A is not used before undef is set.
///
void PromoteMem2Reg::PromoteSingleBlockAlloca(AllocaInst *AI, AllocaInfo &Info,
LargeBlockInfo &LBI) {
// The trickiest case to handle is when we have large blocks. Because of this,
// this code is optimized assuming that large blocks happen. This does not
// significantly pessimize the small block case. This uses LargeBlockInfo to
// make it efficient to get the index of various operations in the block.
// Clear out UsingBlocks. We will reconstruct it here if needed.
Info.UsingBlocks.clear();
// Walk the use-def list of the alloca, getting the locations of all stores.
typedef SmallVector<std::pair<unsigned, StoreInst*>, 64> StoresByIndexTy;
StoresByIndexTy StoresByIndex;
for (Value::use_iterator UI = AI->use_begin(), E = AI->use_end();
UI != E; ++UI)
if (StoreInst *SI = dyn_cast<StoreInst>(*UI))
StoresByIndex.push_back(std::make_pair(LBI.getInstructionIndex(SI), SI));
// If there are no stores to the alloca, just replace any loads with undef.
if (StoresByIndex.empty()) {
for (Value::use_iterator UI = AI->use_begin(), E = AI->use_end(); UI != E;)
if (LoadInst *LI = dyn_cast<LoadInst>(*UI++)) {
LI->replaceAllUsesWith(UndefValue::get(LI->getType()));
if (AST && LI->getType()->isPointerTy())
AST->deleteValue(LI);
LBI.deleteValue(LI);
LI->eraseFromParent();
}
return;
}
// Sort the stores by their index, making it efficient to do a lookup with a
// binary search.
std::sort(StoresByIndex.begin(), StoresByIndex.end());
// Walk all of the loads from this alloca, replacing them with the nearest
// store above them, if any.
for (Value::use_iterator UI = AI->use_begin(), E = AI->use_end(); UI != E;) {
LoadInst *LI = dyn_cast<LoadInst>(*UI++);
if (!LI) continue;
unsigned LoadIdx = LBI.getInstructionIndex(LI);
// Find the nearest store that has a lower than this load.
StoresByIndexTy::iterator I =
std::lower_bound(StoresByIndex.begin(), StoresByIndex.end(),
std::pair<unsigned, StoreInst*>(LoadIdx, static_cast<StoreInst*>(0)),
StoreIndexSearchPredicate());
// If there is no store before this load, then we can't promote this load.
if (I == StoresByIndex.begin()) {
// Can't handle this load, bail out.
Info.UsingBlocks.push_back(LI->getParent());
continue;
}
// Otherwise, there was a store before this load, the load takes its value.
--I;
LI->replaceAllUsesWith(I->second->getOperand(0));
if (AST && LI->getType()->isPointerTy())
AST->deleteValue(LI);
LI->eraseFromParent();
LBI.deleteValue(LI);
}
}
void PromoteMem2Reg::run() {
Function &F = *DF.getRoot()->getParent();
if (AST) PointerAllocaValues.resize(Allocas.size());
AllocaDbgDeclares.resize(Allocas.size());
AllocaInfo Info;
LargeBlockInfo LBI;
for (unsigned AllocaNum = 0; AllocaNum != Allocas.size(); ++AllocaNum) {
AllocaInst *AI = Allocas[AllocaNum];
assert(isAllocaPromotable(AI) &&
"Cannot promote non-promotable alloca!");
assert(AI->getParent()->getParent() == &F &&
"All allocas should be in the same function, which is same as DF!");
if (AI->use_empty()) {
// If there are no uses of the alloca, just delete it now.
if (AST) AST->deleteValue(AI);
AI->eraseFromParent();
// Remove the alloca from the Allocas list, since it has been processed
RemoveFromAllocasList(AllocaNum);
++NumDeadAlloca;
continue;
}
// Calculate the set of read and write-locations for each alloca. This is
// analogous to finding the 'uses' and 'definitions' of each variable.
Info.AnalyzeAlloca(AI);
// If there is only a single store to this value, replace any loads of
// it that are directly dominated by the definition with the value stored.
if (Info.DefiningBlocks.size() == 1) {
RewriteSingleStoreAlloca(AI, Info, LBI);
// Finally, after the scan, check to see if the store is all that is left.
if (Info.UsingBlocks.empty()) {
// Record debuginfo for the store and remove the declaration's debuginfo.
if (DbgDeclareInst *DDI = Info.DbgDeclare) {
ConvertDebugDeclareToDebugValue(DDI, Info.OnlyStore);
DDI->eraseFromParent();
}
// Remove the (now dead) store and alloca.
Info.OnlyStore->eraseFromParent();
LBI.deleteValue(Info.OnlyStore);
if (AST) AST->deleteValue(AI);
AI->eraseFromParent();
LBI.deleteValue(AI);
// The alloca has been processed, move on.
RemoveFromAllocasList(AllocaNum);
++NumSingleStore;
continue;
}
}
// If the alloca is only read and written in one basic block, just perform a
// linear sweep over the block to eliminate it.
if (Info.OnlyUsedInOneBlock) {
PromoteSingleBlockAlloca(AI, Info, LBI);
// Finally, after the scan, check to see if the stores are all that is
// left.
if (Info.UsingBlocks.empty()) {
// Remove the (now dead) stores and alloca.
while (!AI->use_empty()) {
StoreInst *SI = cast<StoreInst>(AI->use_back());
// Record debuginfo for the store before removing it.
if (DbgDeclareInst *DDI = Info.DbgDeclare)
ConvertDebugDeclareToDebugValue(DDI, SI);
SI->eraseFromParent();
LBI.deleteValue(SI);
}
if (AST) AST->deleteValue(AI);
AI->eraseFromParent();
LBI.deleteValue(AI);
// The alloca has been processed, move on.
RemoveFromAllocasList(AllocaNum);
// The alloca's debuginfo can be removed as well.
if (DbgDeclareInst *DDI = Info.DbgDeclare)
DDI->eraseFromParent();
++NumLocalPromoted;
continue;
}
}
// If we haven't computed a numbering for the BB's in the function, do so
// now.
if (BBNumbers.empty()) {
unsigned ID = 0;
for (Function::iterator I = F.begin(), E = F.end(); I != E; ++I)
BBNumbers[I] = ID++;
}
// If we have an AST to keep updated, remember some pointer value that is
// stored into the alloca.
if (AST)
PointerAllocaValues[AllocaNum] = Info.AllocaPointerVal;
// Remember the dbg.declare intrinsic describing this alloca, if any.
if (Info.DbgDeclare) AllocaDbgDeclares[AllocaNum] = Info.DbgDeclare;
// Keep the reverse mapping of the 'Allocas' array for the rename pass.
AllocaLookup[Allocas[AllocaNum]] = AllocaNum;
// At this point, we're committed to promoting the alloca using IDF's, and
// the standard SSA construction algorithm. Determine which blocks need PHI
// nodes and see if we can optimize out some work by avoiding insertion of
// dead phi nodes.
DetermineInsertionPoint(AI, AllocaNum, Info);
}
if (Allocas.empty())
return; // All of the allocas must have been trivial!
LBI.clear();
// Set the incoming values for the basic block to be null values for all of
// the alloca's. We do this in case there is a load of a value that has not
// been stored yet. In this case, it will get this null value.
//
RenamePassData::ValVector Values(Allocas.size());
for (unsigned i = 0, e = Allocas.size(); i != e; ++i)
Values[i] = UndefValue::get(Allocas[i]->getAllocatedType());
// Walks all basic blocks in the function performing the SSA rename algorithm
// and inserting the phi nodes we marked as necessary
//
std::vector<RenamePassData> RenamePassWorkList;
RenamePassWorkList.push_back(RenamePassData(F.begin(), 0, Values));
do {
RenamePassData RPD;
RPD.swap(RenamePassWorkList.back());
RenamePassWorkList.pop_back();
// RenamePass may add new worklist entries.
RenamePass(RPD.BB, RPD.Pred, RPD.Values, RenamePassWorkList);
} while (!RenamePassWorkList.empty());
// The renamer uses the Visited set to avoid infinite loops. Clear it now.
Visited.clear();
// Remove the allocas themselves from the function.
for (unsigned i = 0, e = Allocas.size(); i != e; ++i) {
Instruction *A = Allocas[i];
// If there are any uses of the alloca instructions left, they must be in
// sections of dead code that were not processed on the dominance frontier.
// Just delete the users now.
//
if (!A->use_empty())
A->replaceAllUsesWith(UndefValue::get(A->getType()));
if (AST) AST->deleteValue(A);
A->eraseFromParent();
}
// Remove alloca's dbg.declare instrinsics from the function.
for (unsigned i = 0, e = AllocaDbgDeclares.size(); i != e; ++i)
if (DbgDeclareInst *DDI = AllocaDbgDeclares[i])
DDI->eraseFromParent();
// Loop over all of the PHI nodes and see if there are any that we can get
// rid of because they merge all of the same incoming values. This can
// happen due to undef values coming into the PHI nodes. This process is
// iterative, because eliminating one PHI node can cause others to be removed.
bool EliminatedAPHI = true;
while (EliminatedAPHI) {
EliminatedAPHI = false;
for (DenseMap<std::pair<BasicBlock*, unsigned>, PHINode*>::iterator I =
NewPhiNodes.begin(), E = NewPhiNodes.end(); I != E;) {
PHINode *PN = I->second;
// If this PHI node merges one value and/or undefs, get the value.
if (Value *V = PN->hasConstantValue(&DT)) {
if (AST && PN->getType()->isPointerTy())
AST->deleteValue(PN);
PN->replaceAllUsesWith(V);
PN->eraseFromParent();
NewPhiNodes.erase(I++);
EliminatedAPHI = true;
continue;
}
++I;
}
}
// At this point, the renamer has added entries to PHI nodes for all reachable
// code. Unfortunately, there may be unreachable blocks which the renamer
// hasn't traversed. If this is the case, the PHI nodes may not
// have incoming values for all predecessors. Loop over all PHI nodes we have
// created, inserting undef values if they are missing any incoming values.
//
for (DenseMap<std::pair<BasicBlock*, unsigned>, PHINode*>::iterator I =
NewPhiNodes.begin(), E = NewPhiNodes.end(); I != E; ++I) {
// We want to do this once per basic block. As such, only process a block
// when we find the PHI that is the first entry in the block.
PHINode *SomePHI = I->second;
BasicBlock *BB = SomePHI->getParent();
if (&BB->front() != SomePHI)
continue;
// Only do work here if there the PHI nodes are missing incoming values. We
// know that all PHI nodes that were inserted in a block will have the same
// number of incoming values, so we can just check any of them.
if (SomePHI->getNumIncomingValues() == getNumPreds(BB))
continue;
// Get the preds for BB.
SmallVector<BasicBlock*, 16> Preds(pred_begin(BB), pred_end(BB));
// Ok, now we know that all of the PHI nodes are missing entries for some
// basic blocks. Start by sorting the incoming predecessors for efficient
// access.
std::sort(Preds.begin(), Preds.end());
// Now we loop through all BB's which have entries in SomePHI and remove
// them from the Preds list.
for (unsigned i = 0, e = SomePHI->getNumIncomingValues(); i != e; ++i) {
// Do a log(n) search of the Preds list for the entry we want.
SmallVector<BasicBlock*, 16>::iterator EntIt =
std::lower_bound(Preds.begin(), Preds.end(),
SomePHI->getIncomingBlock(i));
assert(EntIt != Preds.end() && *EntIt == SomePHI->getIncomingBlock(i)&&
"PHI node has entry for a block which is not a predecessor!");
// Remove the entry
Preds.erase(EntIt);
}
// At this point, the blocks left in the preds list must have dummy
// entries inserted into every PHI nodes for the block. Update all the phi
// nodes in this block that we are inserting (there could be phis before
// mem2reg runs).
unsigned NumBadPreds = SomePHI->getNumIncomingValues();
BasicBlock::iterator BBI = BB->begin();
while ((SomePHI = dyn_cast<PHINode>(BBI++)) &&
SomePHI->getNumIncomingValues() == NumBadPreds) {
Value *UndefVal = UndefValue::get(SomePHI->getType());
for (unsigned pred = 0, e = Preds.size(); pred != e; ++pred)
SomePHI->addIncoming(UndefVal, Preds[pred]);
}
}
NewPhiNodes.clear();
}
/// Many allocas are only used within a single basic block. If this is the
/// case, avoid traversing the CFG and inserting a lot of potentially useless
/// PHI nodes by just performing a single linear pass over the basic block
/// using the Alloca.
///
/// If we cannot promote this alloca (because it is read before it is written),
/// return true. This is necessary in cases where, due to control flow, the
/// alloca is potentially undefined on some control flow paths. e.g. code like
/// this is potentially correct:
///
/// for (...) { if (c) { A = undef; undef = B; } }
///
/// ... so long as A is not used before undef is set.
static void promoteSingleBlockAlloca(AllocaInst *AI, const AllocaInfo &Info,
LargeBlockInfo &LBI,
AliasSetTracker *AST) {
// The trickiest case to handle is when we have large blocks. Because of this,
// this code is optimized assuming that large blocks happen. This does not
// significantly pessimize the small block case. This uses LargeBlockInfo to
// make it efficient to get the index of various operations in the block.
// Walk the use-def list of the alloca, getting the locations of all stores.
typedef SmallVector<std::pair<unsigned, StoreInst *>, 64> StoresByIndexTy;
StoresByIndexTy StoresByIndex;
for (Value::use_iterator UI = AI->use_begin(), E = AI->use_end(); UI != E;
++UI)
if (StoreInst *SI = dyn_cast<StoreInst>(*UI))
StoresByIndex.push_back(std::make_pair(LBI.getInstructionIndex(SI), SI));
// Sort the stores by their index, making it efficient to do a lookup with a
// binary search.
std::sort(StoresByIndex.begin(), StoresByIndex.end(),
StoreIndexSearchPredicate());
// Walk all of the loads from this alloca, replacing them with the nearest
// store above them, if any.
for (Value::use_iterator UI = AI->use_begin(), E = AI->use_end(); UI != E;) {
LoadInst *LI = dyn_cast<LoadInst>(*UI++);
if (!LI)
continue;
unsigned LoadIdx = LBI.getInstructionIndex(LI);
// Find the nearest store that has a lower index than this load.
StoresByIndexTy::iterator I =
std::lower_bound(StoresByIndex.begin(), StoresByIndex.end(),
std::make_pair(LoadIdx, static_cast<StoreInst *>(0)),
StoreIndexSearchPredicate());
if (I == StoresByIndex.begin())
// If there is no store before this load, the load takes the undef value.
LI->replaceAllUsesWith(UndefValue::get(LI->getType()));
else
// Otherwise, there was a store before this load, the load takes its value.
LI->replaceAllUsesWith(llvm::prior(I)->second->getOperand(0));
if (AST && LI->getType()->isPointerTy())
AST->deleteValue(LI);
LI->eraseFromParent();
LBI.deleteValue(LI);
}
// Remove the (now dead) stores and alloca.
while (!AI->use_empty()) {
StoreInst *SI = cast<StoreInst>(AI->use_back());
// Record debuginfo for the store before removing it.
if (DbgDeclareInst *DDI = Info.DbgDeclare) {
DIBuilder DIB(*AI->getParent()->getParent()->getParent());
ConvertDebugDeclareToDebugValue(DDI, SI, DIB);
}
SI->eraseFromParent();
LBI.deleteValue(SI);
}
if (AST)
AST->deleteValue(AI);
AI->eraseFromParent();
LBI.deleteValue(AI);
// The alloca's debuginfo can be removed as well.
if (DbgDeclareInst *DDI = Info.DbgDeclare)
DDI->eraseFromParent();
++NumLocalPromoted;
}
/// \brief Rewrite as many loads as possible given a single store.
///
/// When there is only a single store, we can use the domtree to trivially
/// replace all of the dominated loads with the stored value. Do so, and return
/// true if this has successfully promoted the alloca entirely. If this returns
/// false there were some loads which were not dominated by the single store
/// and thus must be phi-ed with undef. We fall back to the standard alloca
/// promotion algorithm in that case.
static bool rewriteSingleStoreAlloca(AllocaInst *AI, AllocaInfo &Info,
LargeBlockInfo &LBI,
DominatorTree &DT,
AliasSetTracker *AST) {
StoreInst *OnlyStore = Info.OnlyStore;
bool StoringGlobalVal = !isa<Instruction>(OnlyStore->getOperand(0));
BasicBlock *StoreBB = OnlyStore->getParent();
int StoreIndex = -1;
// Clear out UsingBlocks. We will reconstruct it here if needed.
Info.UsingBlocks.clear();
for (Value::use_iterator UI = AI->use_begin(), E = AI->use_end(); UI != E;) {
Instruction *UserInst = cast<Instruction>(*UI++);
if (!isa<LoadInst>(UserInst)) {
assert(UserInst == OnlyStore && "Should only have load/stores");
continue;
}
LoadInst *LI = cast<LoadInst>(UserInst);
// Okay, if we have a load from the alloca, we want to replace it with the
// only value stored to the alloca. We can do this if the value is
// dominated by the store. If not, we use the rest of the mem2reg machinery
// to insert the phi nodes as needed.
if (!StoringGlobalVal) { // Non-instructions are always dominated.
if (LI->getParent() == StoreBB) {
// If we have a use that is in the same block as the store, compare the
// indices of the two instructions to see which one came first. If the
// load came before the store, we can't handle it.
if (StoreIndex == -1)
StoreIndex = LBI.getInstructionIndex(OnlyStore);
if (unsigned(StoreIndex) > LBI.getInstructionIndex(LI)) {
// Can't handle this load, bail out.
Info.UsingBlocks.push_back(StoreBB);
continue;
}
} else if (LI->getParent() != StoreBB &&
!DT.dominates(StoreBB, LI->getParent())) {
// If the load and store are in different blocks, use BB dominance to
// check their relationships. If the store doesn't dom the use, bail
// out.
Info.UsingBlocks.push_back(LI->getParent());
continue;
}
}
// Otherwise, we *can* safely rewrite this load.
Value *ReplVal = OnlyStore->getOperand(0);
// If the replacement value is the load, this must occur in unreachable
// code.
if (ReplVal == LI)
ReplVal = UndefValue::get(LI->getType());
LI->replaceAllUsesWith(ReplVal);
if (AST && LI->getType()->isPointerTy())
AST->deleteValue(LI);
LI->eraseFromParent();
LBI.deleteValue(LI);
}
// Finally, after the scan, check to see if the store is all that is left.
if (!Info.UsingBlocks.empty())
return false; // If not, we'll have to fall back for the remainder.
// Record debuginfo for the store and remove the declaration's
// debuginfo.
if (DbgDeclareInst *DDI = Info.DbgDeclare) {
DIBuilder DIB(*AI->getParent()->getParent()->getParent());
ConvertDebugDeclareToDebugValue(DDI, Info.OnlyStore, DIB);
DDI->eraseFromParent();
}
// Remove the (now dead) store and alloca.
Info.OnlyStore->eraseFromParent();
LBI.deleteValue(Info.OnlyStore);
if (AST)
AST->deleteValue(AI);
AI->eraseFromParent();
LBI.deleteValue(AI);
return true;
}
void PromoteMem2Reg::run() {
Function &F = *DT.getRoot()->getParent();
AllocaDbgDeclares.resize(Allocas.size());
AllocaInfo Info;
LargeBlockInfo LBI;
ForwardIDFCalculator IDF(DT);
for (unsigned AllocaNum = 0; AllocaNum != Allocas.size(); ++AllocaNum) {
AllocaInst *AI = Allocas[AllocaNum];
assert(isAllocaPromotable(AI) && "Cannot promote non-promotable alloca!");
assert(AI->getParent()->getParent() == &F &&
"All allocas should be in the same function, which is same as DF!");
removeLifetimeIntrinsicUsers(AI);
if (AI->use_empty()) {
// If there are no uses of the alloca, just delete it now.
AI->eraseFromParent();
// Remove the alloca from the Allocas list, since it has been processed
RemoveFromAllocasList(AllocaNum);
++NumDeadAlloca;
continue;
}
// Calculate the set of read and write-locations for each alloca. This is
// analogous to finding the 'uses' and 'definitions' of each variable.
Info.AnalyzeAlloca(AI);
// If there is only a single store to this value, replace any loads of
// it that are directly dominated by the definition with the value stored.
if (Info.DefiningBlocks.size() == 1) {
if (rewriteSingleStoreAlloca(AI, Info, LBI, SQ.DL, DT, AC)) {
// The alloca has been processed, move on.
RemoveFromAllocasList(AllocaNum);
++NumSingleStore;
continue;
}
}
// If the alloca is only read and written in one basic block, just perform a
// linear sweep over the block to eliminate it.
if (Info.OnlyUsedInOneBlock &&
promoteSingleBlockAlloca(AI, Info, LBI, SQ.DL, DT, AC)) {
// The alloca has been processed, move on.
RemoveFromAllocasList(AllocaNum);
continue;
}
// If we haven't computed a numbering for the BB's in the function, do so
// now.
if (BBNumbers.empty()) {
unsigned ID = 0;
for (auto &BB : F)
BBNumbers[&BB] = ID++;
}
// Remember the dbg.declare intrinsic describing this alloca, if any.
if (!Info.DbgDeclares.empty())
AllocaDbgDeclares[AllocaNum] = Info.DbgDeclares;
// Keep the reverse mapping of the 'Allocas' array for the rename pass.
AllocaLookup[Allocas[AllocaNum]] = AllocaNum;
// At this point, we're committed to promoting the alloca using IDF's, and
// the standard SSA construction algorithm. Determine which blocks need PHI
// nodes and see if we can optimize out some work by avoiding insertion of
// dead phi nodes.
// Unique the set of defining blocks for efficient lookup.
SmallPtrSet<BasicBlock *, 32> DefBlocks;
DefBlocks.insert(Info.DefiningBlocks.begin(), Info.DefiningBlocks.end());
// Determine which blocks the value is live in. These are blocks which lead
// to uses.
SmallPtrSet<BasicBlock *, 32> LiveInBlocks;
ComputeLiveInBlocks(AI, Info, DefBlocks, LiveInBlocks);
// At this point, we're committed to promoting the alloca using IDF's, and
// the standard SSA construction algorithm. Determine which blocks need phi
// nodes and see if we can optimize out some work by avoiding insertion of
// dead phi nodes.
IDF.setLiveInBlocks(LiveInBlocks);
IDF.setDefiningBlocks(DefBlocks);
SmallVector<BasicBlock *, 32> PHIBlocks;
IDF.calculate(PHIBlocks);
if (PHIBlocks.size() > 1)
llvm::sort(PHIBlocks, [this](BasicBlock *A, BasicBlock *B) {
return BBNumbers.lookup(A) < BBNumbers.lookup(B);
});
unsigned CurrentVersion = 0;
for (BasicBlock *BB : PHIBlocks)
QueuePhiNode(BB, AllocaNum, CurrentVersion);
}
if (Allocas.empty())
return; // All of the allocas must have been trivial!
LBI.clear();
// Set the incoming values for the basic block to be null values for all of
// the alloca's. We do this in case there is a load of a value that has not
// been stored yet. In this case, it will get this null value.
RenamePassData::ValVector Values(Allocas.size());
for (unsigned i = 0, e = Allocas.size(); i != e; ++i)
Values[i] = UndefValue::get(Allocas[i]->getAllocatedType());
// When handling debug info, treat all incoming values as if they have unknown
// locations until proven otherwise.
RenamePassData::LocationVector Locations(Allocas.size());
// Walks all basic blocks in the function performing the SSA rename algorithm
// and inserting the phi nodes we marked as necessary
std::vector<RenamePassData> RenamePassWorkList;
RenamePassWorkList.emplace_back(&F.front(), nullptr, std::move(Values),
std::move(Locations));
do {
RenamePassData RPD = std::move(RenamePassWorkList.back());
RenamePassWorkList.pop_back();
// RenamePass may add new worklist entries.
RenamePass(RPD.BB, RPD.Pred, RPD.Values, RPD.Locations, RenamePassWorkList);
} while (!RenamePassWorkList.empty());
// The renamer uses the Visited set to avoid infinite loops. Clear it now.
Visited.clear();
// Remove the allocas themselves from the function.
for (Instruction *A : Allocas) {
// If there are any uses of the alloca instructions left, they must be in
// unreachable basic blocks that were not processed by walking the dominator
// tree. Just delete the users now.
if (!A->use_empty())
A->replaceAllUsesWith(UndefValue::get(A->getType()));
A->eraseFromParent();
}
// Remove alloca's dbg.declare instrinsics from the function.
for (auto &Declares : AllocaDbgDeclares)
for (auto *DII : Declares)
DII->eraseFromParent();
// Loop over all of the PHI nodes and see if there are any that we can get
// rid of because they merge all of the same incoming values. This can
// happen due to undef values coming into the PHI nodes. This process is
// iterative, because eliminating one PHI node can cause others to be removed.
bool EliminatedAPHI = true;
while (EliminatedAPHI) {
EliminatedAPHI = false;
// Iterating over NewPhiNodes is deterministic, so it is safe to try to
// simplify and RAUW them as we go. If it was not, we could add uses to
// the values we replace with in a non-deterministic order, thus creating
// non-deterministic def->use chains.
for (DenseMap<std::pair<unsigned, unsigned>, PHINode *>::iterator
I = NewPhiNodes.begin(),
E = NewPhiNodes.end();
I != E;) {
PHINode *PN = I->second;
// If this PHI node merges one value and/or undefs, get the value.
if (Value *V = SimplifyInstruction(PN, SQ)) {
PN->replaceAllUsesWith(V);
PN->eraseFromParent();
NewPhiNodes.erase(I++);
EliminatedAPHI = true;
continue;
}
++I;
}
}
// At this point, the renamer has added entries to PHI nodes for all reachable
// code. Unfortunately, there may be unreachable blocks which the renamer
// hasn't traversed. If this is the case, the PHI nodes may not
// have incoming values for all predecessors. Loop over all PHI nodes we have
// created, inserting undef values if they are missing any incoming values.
for (DenseMap<std::pair<unsigned, unsigned>, PHINode *>::iterator
I = NewPhiNodes.begin(),
E = NewPhiNodes.end();
I != E; ++I) {
// We want to do this once per basic block. As such, only process a block
// when we find the PHI that is the first entry in the block.
PHINode *SomePHI = I->second;
BasicBlock *BB = SomePHI->getParent();
if (&BB->front() != SomePHI)
continue;
// Only do work here if there the PHI nodes are missing incoming values. We
// know that all PHI nodes that were inserted in a block will have the same
// number of incoming values, so we can just check any of them.
if (SomePHI->getNumIncomingValues() == getNumPreds(BB))
continue;
// Get the preds for BB.
SmallVector<BasicBlock *, 16> Preds(pred_begin(BB), pred_end(BB));
// Ok, now we know that all of the PHI nodes are missing entries for some
// basic blocks. Start by sorting the incoming predecessors for efficient
// access.
auto CompareBBNumbers = [this](BasicBlock *A, BasicBlock *B) {
return BBNumbers.lookup(A) < BBNumbers.lookup(B);
};
llvm::sort(Preds, CompareBBNumbers);
// Now we loop through all BB's which have entries in SomePHI and remove
// them from the Preds list.
for (unsigned i = 0, e = SomePHI->getNumIncomingValues(); i != e; ++i) {
// Do a log(n) search of the Preds list for the entry we want.
SmallVectorImpl<BasicBlock *>::iterator EntIt = std::lower_bound(
Preds.begin(), Preds.end(), SomePHI->getIncomingBlock(i),
CompareBBNumbers);
assert(EntIt != Preds.end() && *EntIt == SomePHI->getIncomingBlock(i) &&
"PHI node has entry for a block which is not a predecessor!");
// Remove the entry
Preds.erase(EntIt);
}
// At this point, the blocks left in the preds list must have dummy
// entries inserted into every PHI nodes for the block. Update all the phi
// nodes in this block that we are inserting (there could be phis before
// mem2reg runs).
unsigned NumBadPreds = SomePHI->getNumIncomingValues();
BasicBlock::iterator BBI = BB->begin();
while ((SomePHI = dyn_cast<PHINode>(BBI++)) &&
SomePHI->getNumIncomingValues() == NumBadPreds) {
Value *UndefVal = UndefValue::get(SomePHI->getType());
for (BasicBlock *Pred : Preds)
SomePHI->addIncoming(UndefVal, Pred);
}
}
NewPhiNodes.clear();
}
/// Many allocas are only used within a single basic block. If this is the
/// case, avoid traversing the CFG and inserting a lot of potentially useless
/// PHI nodes by just performing a single linear pass over the basic block
/// using the Alloca.
///
/// If we cannot promote this alloca (because it is read before it is written),
/// return false. This is necessary in cases where, due to control flow, the
/// alloca is undefined only on some control flow paths. e.g. code like
/// this is correct in LLVM IR:
/// // A is an alloca with no stores so far
/// for (...) {
/// int t = *A;
/// if (!first_iteration)
/// use(t);
/// *A = 42;
/// }
static bool promoteSingleBlockAlloca(AllocaInst *AI, const AllocaInfo &Info,
LargeBlockInfo &LBI,
const DataLayout &DL,
DominatorTree &DT,
AssumptionCache *AC) {
// The trickiest case to handle is when we have large blocks. Because of this,
// this code is optimized assuming that large blocks happen. This does not
// significantly pessimize the small block case. This uses LargeBlockInfo to
// make it efficient to get the index of various operations in the block.
// Walk the use-def list of the alloca, getting the locations of all stores.
using StoresByIndexTy = SmallVector<std::pair<unsigned, StoreInst *>, 64>;
StoresByIndexTy StoresByIndex;
for (User *U : AI->users())
if (StoreInst *SI = dyn_cast<StoreInst>(U))
StoresByIndex.push_back(std::make_pair(LBI.getInstructionIndex(SI), SI));
// Sort the stores by their index, making it efficient to do a lookup with a
// binary search.
llvm::sort(StoresByIndex, less_first());
// Walk all of the loads from this alloca, replacing them with the nearest
// store above them, if any.
for (auto UI = AI->user_begin(), E = AI->user_end(); UI != E;) {
LoadInst *LI = dyn_cast<LoadInst>(*UI++);
if (!LI)
continue;
unsigned LoadIdx = LBI.getInstructionIndex(LI);
// Find the nearest store that has a lower index than this load.
StoresByIndexTy::iterator I =
std::lower_bound(StoresByIndex.begin(), StoresByIndex.end(),
std::make_pair(LoadIdx,
static_cast<StoreInst *>(nullptr)),
less_first());
if (I == StoresByIndex.begin()) {
if (StoresByIndex.empty())
// If there are no stores, the load takes the undef value.
LI->replaceAllUsesWith(UndefValue::get(LI->getType()));
else
// There is no store before this load, bail out (load may be affected
// by the following stores - see main comment).
return false;
} else {
// Otherwise, there was a store before this load, the load takes its value.
// Note, if the load was marked as nonnull we don't want to lose that
// information when we erase it. So we preserve it with an assume.
Value *ReplVal = std::prev(I)->second->getOperand(0);
if (AC && LI->getMetadata(LLVMContext::MD_nonnull) &&
!isKnownNonZero(ReplVal, DL, 0, AC, LI, &DT))
addAssumeNonNull(AC, LI);
// If the replacement value is the load, this must occur in unreachable
// code.
if (ReplVal == LI)
ReplVal = UndefValue::get(LI->getType());
LI->replaceAllUsesWith(ReplVal);
}
LI->eraseFromParent();
LBI.deleteValue(LI);
}
// Remove the (now dead) stores and alloca.
while (!AI->use_empty()) {
StoreInst *SI = cast<StoreInst>(AI->user_back());
// Record debuginfo for the store before removing it.
for (DbgVariableIntrinsic *DII : Info.DbgDeclares) {
DIBuilder DIB(*AI->getModule(), /*AllowUnresolved*/ false);
ConvertDebugDeclareToDebugValue(DII, SI, DIB);
}
SI->eraseFromParent();
LBI.deleteValue(SI);
}
AI->eraseFromParent();
LBI.deleteValue(AI);
// The alloca's debuginfo can be removed as well.
for (DbgVariableIntrinsic *DII : Info.DbgDeclares) {
DII->eraseFromParent();
LBI.deleteValue(DII);
}
++NumLocalPromoted;
return true;
}
/// Rewrite as many loads as possible given a single store.
///
/// When there is only a single store, we can use the domtree to trivially
/// replace all of the dominated loads with the stored value. Do so, and return
/// true if this has successfully promoted the alloca entirely. If this returns
/// false there were some loads which were not dominated by the single store
/// and thus must be phi-ed with undef. We fall back to the standard alloca
/// promotion algorithm in that case.
static bool rewriteSingleStoreAlloca(AllocaInst *AI, AllocaInfo &Info,
LargeBlockInfo &LBI, const DataLayout &DL,
DominatorTree &DT, AssumptionCache *AC) {
StoreInst *OnlyStore = Info.OnlyStore;
bool StoringGlobalVal = !isa<Instruction>(OnlyStore->getOperand(0));
BasicBlock *StoreBB = OnlyStore->getParent();
int StoreIndex = -1;
// Clear out UsingBlocks. We will reconstruct it here if needed.
Info.UsingBlocks.clear();
for (auto UI = AI->user_begin(), E = AI->user_end(); UI != E;) {
Instruction *UserInst = cast<Instruction>(*UI++);
if (!isa<LoadInst>(UserInst)) {
assert(UserInst == OnlyStore && "Should only have load/stores");
continue;
}
LoadInst *LI = cast<LoadInst>(UserInst);
// Okay, if we have a load from the alloca, we want to replace it with the
// only value stored to the alloca. We can do this if the value is
// dominated by the store. If not, we use the rest of the mem2reg machinery
// to insert the phi nodes as needed.
if (!StoringGlobalVal) { // Non-instructions are always dominated.
if (LI->getParent() == StoreBB) {
// If we have a use that is in the same block as the store, compare the
// indices of the two instructions to see which one came first. If the
// load came before the store, we can't handle it.
if (StoreIndex == -1)
StoreIndex = LBI.getInstructionIndex(OnlyStore);
if (unsigned(StoreIndex) > LBI.getInstructionIndex(LI)) {
// Can't handle this load, bail out.
Info.UsingBlocks.push_back(StoreBB);
continue;
}
} else if (LI->getParent() != StoreBB &&
!DT.dominates(StoreBB, LI->getParent())) {
// If the load and store are in different blocks, use BB dominance to
// check their relationships. If the store doesn't dom the use, bail
// out.
Info.UsingBlocks.push_back(LI->getParent());
continue;
}
}
// Otherwise, we *can* safely rewrite this load.
Value *ReplVal = OnlyStore->getOperand(0);
// If the replacement value is the load, this must occur in unreachable
// code.
if (ReplVal == LI)
ReplVal = UndefValue::get(LI->getType());
// If the load was marked as nonnull we don't want to lose
// that information when we erase this Load. So we preserve
// it with an assume.
if (AC && LI->getMetadata(LLVMContext::MD_nonnull) &&
!isKnownNonZero(ReplVal, DL, 0, AC, LI, &DT))
addAssumeNonNull(AC, LI);
LI->replaceAllUsesWith(ReplVal);
LI->eraseFromParent();
LBI.deleteValue(LI);
}
// Finally, after the scan, check to see if the store is all that is left.
if (!Info.UsingBlocks.empty())
return false; // If not, we'll have to fall back for the remainder.
// Record debuginfo for the store and remove the declaration's
// debuginfo.
for (DbgVariableIntrinsic *DII : Info.DbgDeclares) {
DIBuilder DIB(*AI->getModule(), /*AllowUnresolved*/ false);
ConvertDebugDeclareToDebugValue(DII, Info.OnlyStore, DIB);
DII->eraseFromParent();
LBI.deleteValue(DII);
}
// Remove the (now dead) store and alloca.
Info.OnlyStore->eraseFromParent();
LBI.deleteValue(Info.OnlyStore);
AI->eraseFromParent();
LBI.deleteValue(AI);
return true;
}
