aboutsummaryrefslogtreecommitdiff
path: root/lib/CodeGen
diff options
context:
space:
mode:
Diffstat (limited to 'lib/CodeGen')
-rw-r--r--lib/CodeGen/BranchFolding.cpp237
-rw-r--r--lib/CodeGen/BranchFolding.h8
-rw-r--r--lib/CodeGen/CMakeLists.txt1
-rw-r--r--lib/CodeGen/LLVMTargetMachine.cpp8
-rw-r--r--lib/CodeGen/MachineBasicBlock.cpp45
-rw-r--r--lib/CodeGen/TailDuplication.cpp250
6 files changed, 310 insertions, 239 deletions
diff --git a/lib/CodeGen/BranchFolding.cpp b/lib/CodeGen/BranchFolding.cpp
index 0fd3c23085..8a62eb20bb 100644
--- a/lib/CodeGen/BranchFolding.cpp
+++ b/lib/CodeGen/BranchFolding.cpp
@@ -41,8 +41,6 @@ using namespace llvm;
STATISTIC(NumDeadBlocks, "Number of dead blocks removed");
STATISTIC(NumBranchOpts, "Number of branches optimized");
STATISTIC(NumTailMerge , "Number of block tails merged");
-STATISTIC(NumTailDups , "Number of tail duplicated blocks");
-STATISTIC(NumInstrDups , "Additional instructions due to tail duplication");
static cl::opt<cl::boolOrDefault> FlagEnableTailMerge("enable-tail-merge",
cl::init(cl::BOU_UNSET), cl::Hidden);
@@ -205,16 +203,6 @@ bool BranchFolder::OptimizeFunction(MachineFunction &MF,
MadeChange |= MadeChangeThisIteration;
}
- // Do tail duplication after tail merging is done. Otherwise it is
- // tough to avoid situations where tail duplication and tail merging undo
- // each other's transformations ad infinitum.
- MadeChangeThisIteration = true;
- while (MadeChangeThisIteration) {
- MadeChangeThisIteration = false;
- MadeChangeThisIteration |= TailDuplicateBlocks(MF);
- MadeChange |= MadeChangeThisIteration;
- }
-
// See if any jump tables have become mergable or dead as the code generator
// did its thing.
MachineJumpTableInfo *JTI = MF.getJumpTableInfo();
@@ -918,71 +906,6 @@ bool BranchFolder::OptimizeBranches(MachineFunction &MF) {
}
-/// CanFallThrough - Return true if the specified block (with the specified
-/// branch condition) can implicitly transfer control to the block after it by
-/// falling off the end of it. This should return false if it can reach the
-/// block after it, but it uses an explicit branch to do so (e.g. a table jump).
-///
-/// True is a conservative answer.
-///
-bool BranchFolder::CanFallThrough(MachineBasicBlock *CurBB,
- bool BranchUnAnalyzable,
- MachineBasicBlock *TBB,
- MachineBasicBlock *FBB,
- const SmallVectorImpl<MachineOperand> &Cond) {
- MachineFunction::iterator Fallthrough = CurBB;
- ++Fallthrough;
- // If FallthroughBlock is off the end of the function, it can't fall through.
- if (Fallthrough == CurBB->getParent()->end())
- return false;
-
- // If FallthroughBlock isn't a successor of CurBB, no fallthrough is possible.
- if (!CurBB->isSuccessor(Fallthrough))
- return false;
-
- // If we couldn't analyze the branch, examine the last instruction.
- // If the block doesn't end in a known control barrier, assume fallthrough
- // is possible. The isPredicable check is needed because this code can be
- // called during IfConversion, where an instruction which is normally a
- // Barrier is predicated and thus no longer an actual control barrier. This
- // is over-conservative though, because if an instruction isn't actually
- // predicated we could still treat it like a barrier.
- if (BranchUnAnalyzable)
- return CurBB->empty() || !CurBB->back().getDesc().isBarrier() ||
- CurBB->back().getDesc().isPredicable();
-
- // If there is no branch, control always falls through.
- if (TBB == 0) return true;
-
- // If there is some explicit branch to the fallthrough block, it can obviously
- // reach, even though the branch should get folded to fall through implicitly.
- if (MachineFunction::iterator(TBB) == Fallthrough ||
- MachineFunction::iterator(FBB) == Fallthrough)
- return true;
-
- // If it's an unconditional branch to some block not the fall through, it
- // doesn't fall through.
- if (Cond.empty()) return false;
-
- // Otherwise, if it is conditional and has no explicit false block, it falls
- // through.
- return FBB == 0;
-}
-
-/// CanFallThrough - Return true if the specified can implicitly transfer
-/// control to the block after it by falling off the end of it. This should
-/// return false if it can reach the block after it, but it uses an explicit
-/// branch to do so (e.g. a table jump).
-///
-/// True is a conservative answer.
-///
-bool BranchFolder::CanFallThrough(MachineBasicBlock *CurBB) {
- MachineBasicBlock *TBB = 0, *FBB = 0;
- SmallVector<MachineOperand, 4> Cond;
- bool CurUnAnalyzable = TII->AnalyzeBranch(*CurBB, TBB, FBB, Cond, true);
- return CanFallThrough(CurBB, CurUnAnalyzable, TBB, FBB, Cond);
-}
-
/// IsBetterFallthrough - Return true if it would be clearly better to
/// fall-through to MBB1 than to fall through into MBB2. This has to return
/// a strict ordering, returning true for both (MBB1,MBB2) and (MBB2,MBB1) will
@@ -1005,152 +928,6 @@ static bool IsBetterFallthrough(MachineBasicBlock *MBB1,
return MBB2I->getDesc().isCall() && !MBB1I->getDesc().isCall();
}
-/// TailDuplicateBlocks - Look for small blocks that are unconditionally
-/// branched to and do not fall through. Tail-duplicate their instructions
-/// into their predecessors to eliminate (dynamic) branches.
-bool BranchFolder::TailDuplicateBlocks(MachineFunction &MF) {
- bool MadeChange = false;
-
- for (MachineFunction::iterator I = ++MF.begin(), E = MF.end(); I != E; ) {
- MachineBasicBlock *MBB = I++;
-
- // Only duplicate blocks that end with unconditional branches.
- if (CanFallThrough(MBB))
- continue;
-
- MadeChange |= TailDuplicate(MBB, MF);
-
- // If it is dead, remove it.
- if (MBB->pred_empty()) {
- NumInstrDups -= MBB->size();
- RemoveDeadBlock(MBB);
- MadeChange = true;
- ++NumDeadBlocks;
- }
- }
- return MadeChange;
-}
-
-/// TailDuplicate - If it is profitable, duplicate TailBB's contents in each
-/// of its predecessors.
-bool BranchFolder::TailDuplicate(MachineBasicBlock *TailBB,
- MachineFunction &MF) {
- // Don't try to tail-duplicate single-block loops.
- if (TailBB->isSuccessor(TailBB))
- return false;
-
- // Set the limit on the number of instructions to duplicate, with a default
- // of one less than the tail-merge threshold. When optimizing for size,
- // duplicate only one, because one branch instruction can be eliminated to
- // compensate for the duplication.
- unsigned MaxDuplicateCount;
- if (MF.getFunction()->hasFnAttr(Attribute::OptimizeForSize))
- MaxDuplicateCount = 1;
- else if (TII->isProfitableToDuplicateIndirectBranch() &&
- !TailBB->empty() && TailBB->back().getDesc().isIndirectBranch())
- // If the target has hardware branch prediction that can handle indirect
- // branches, duplicating them can often make them predictable when there
- // are common paths through the code. The limit needs to be high enough
- // to allow undoing the effects of tail merging.
- MaxDuplicateCount = 20;
- else
- MaxDuplicateCount = TailMergeSize - 1;
-
- // Check the instructions in the block to determine whether tail-duplication
- // is invalid or unlikely to be profitable.
- unsigned i = 0;
- bool HasCall = false;
- for (MachineBasicBlock::iterator I = TailBB->begin();
- I != TailBB->end(); ++I, ++i) {
- // Non-duplicable things shouldn't be tail-duplicated.
- if (I->getDesc().isNotDuplicable()) return false;
- // Don't duplicate more than the threshold.
- if (i == MaxDuplicateCount) return false;
- // Remember if we saw a call.
- if (I->getDesc().isCall()) HasCall = true;
- }
- // Heuristically, don't tail-duplicate calls if it would expand code size,
- // as it's less likely to be worth the extra cost.
- if (i > 1 && HasCall)
- return false;
-
- // Iterate through all the unique predecessors and tail-duplicate this
- // block into them, if possible. Copying the list ahead of time also
- // avoids trouble with the predecessor list reallocating.
- bool Changed = false;
- SmallSetVector<MachineBasicBlock *, 8> Preds(TailBB->pred_begin(),
- TailBB->pred_end());
- for (SmallSetVector<MachineBasicBlock *, 8>::iterator PI = Preds.begin(),
- PE = Preds.end(); PI != PE; ++PI) {
- MachineBasicBlock *PredBB = *PI;
-
- assert(TailBB != PredBB &&
- "Single-block loop should have been rejected earlier!");
- if (PredBB->succ_size() > 1) continue;
-
- MachineBasicBlock *PredTBB, *PredFBB;
- SmallVector<MachineOperand, 4> PredCond;
- if (TII->AnalyzeBranch(*PredBB, PredTBB, PredFBB, PredCond, true))
- continue;
- if (!PredCond.empty())
- continue;
- // EH edges are ignored by AnalyzeBranch.
- if (PredBB->succ_size() != 1)
- continue;
- // Don't duplicate into a fall-through predecessor (at least for now).
- if (PredBB->isLayoutSuccessor(TailBB) && CanFallThrough(PredBB))
- continue;
-
- DEBUG(errs() << "\nTail-duplicating into PredBB: " << *PredBB
- << "From Succ: " << *TailBB);
-
- // Remove PredBB's unconditional branch.
- TII->RemoveBranch(*PredBB);
- // Clone the contents of TailBB into PredBB.
- for (MachineBasicBlock::iterator I = TailBB->begin(), E = TailBB->end();
- I != E; ++I) {
- MachineInstr *NewMI = MF.CloneMachineInstr(I);
- PredBB->insert(PredBB->end(), NewMI);
- }
- NumInstrDups += TailBB->size() - 1; // subtract one for removed branch
-
- // Update the CFG.
- PredBB->removeSuccessor(PredBB->succ_begin());
- assert(PredBB->succ_empty() &&
- "TailDuplicate called on block with multiple successors!");
- for (MachineBasicBlock::succ_iterator I = TailBB->succ_begin(),
- E = TailBB->succ_end(); I != E; ++I)
- PredBB->addSuccessor(*I);
-
- Changed = true;
- ++NumTailDups;
- }
-
- // If TailBB was duplicated into all its predecessors except for the prior
- // block, which falls through unconditionally, move the contents of this
- // block into the prior block.
- MachineBasicBlock &PrevBB = *prior(MachineFunction::iterator(TailBB));
- MachineBasicBlock *PriorTBB = 0, *PriorFBB = 0;
- SmallVector<MachineOperand, 4> PriorCond;
- bool PriorUnAnalyzable =
- TII->AnalyzeBranch(PrevBB, PriorTBB, PriorFBB, PriorCond, true);
- // This has to check PrevBB->succ_size() because EH edges are ignored by
- // AnalyzeBranch.
- if (!PriorUnAnalyzable && PriorCond.empty() && !PriorTBB &&
- TailBB->pred_size() == 1 && PrevBB.succ_size() == 1 &&
- !TailBB->hasAddressTaken()) {
- DEBUG(errs() << "\nMerging into block: " << PrevBB
- << "From MBB: " << *TailBB);
- PrevBB.splice(PrevBB.end(), TailBB, TailBB->begin(), TailBB->end());
- PrevBB.removeSuccessor(PrevBB.succ_begin());;
- assert(PrevBB.succ_empty());
- PrevBB.transferSuccessors(TailBB);
- Changed = true;
- }
-
- return Changed;
-}
-
/// OptimizeBlock - Analyze and optimize control flow related to the specified
/// block. This is never called on the entry block.
bool BranchFolder::OptimizeBlock(MachineBasicBlock *MBB) {
@@ -1275,7 +1052,7 @@ ReoptimizeBlock:
// the assert condition out of the loop body.
if (MBB->succ_empty() && !PriorCond.empty() && PriorFBB == 0 &&
MachineFunction::iterator(PriorTBB) == FallThrough &&
- !CanFallThrough(MBB)) {
+ !MBB->canFallThrough()) {
bool DoTransform = true;
// We have to be careful that the succs of PredBB aren't both no-successor
@@ -1299,7 +1076,7 @@ ReoptimizeBlock:
// In this case, we could actually be moving the return block *into* a
// loop!
if (DoTransform && !MBB->succ_empty() &&
- (!CanFallThrough(PriorTBB) || PriorTBB->empty()))
+ (!PriorTBB->canFallThrough() || PriorTBB->empty()))
DoTransform = false;
@@ -1431,13 +1208,11 @@ ReoptimizeBlock:
// If the prior block doesn't fall through into this block, and if this
// block doesn't fall through into some other block, see if we can find a
// place to move this block where a fall-through will happen.
- if (!CanFallThrough(&PrevBB, PriorUnAnalyzable,
- PriorTBB, PriorFBB, PriorCond)) {
+ if (!PrevBB.canFallThrough()) {
// Now we know that there was no fall-through into this block, check to
// see if it has a fall-through into its successor.
- bool CurFallsThru = CanFallThrough(MBB, CurUnAnalyzable, CurTBB, CurFBB,
- CurCond);
+ bool CurFallsThru = MBB->canFallThrough();
if (!MBB->isLandingPad()) {
// Check all the predecessors of this block. If one of them has no fall
@@ -1449,7 +1224,7 @@ ReoptimizeBlock:
MachineFunction::iterator PredFallthrough = PredBB; ++PredFallthrough;
MachineBasicBlock *PredTBB, *PredFBB;
SmallVector<MachineOperand, 4> PredCond;
- if (PredBB != MBB && !CanFallThrough(PredBB) &&
+ if (PredBB != MBB && !PredBB->canFallThrough() &&
!TII->AnalyzeBranch(*PredBB, PredTBB, PredFBB, PredCond, true)
&& (!CurFallsThru || !CurTBB || !CurFBB)
&& (!CurFallsThru || MBB->getNumber() >= PredBB->getNumber())) {
@@ -1488,7 +1263,7 @@ ReoptimizeBlock:
// and if the successor isn't an EH destination, we can arrange for the
// fallthrough to happen.
if (SuccBB != MBB && &*SuccPrev != MBB &&
- !CanFallThrough(SuccPrev) && !CurUnAnalyzable &&
+ !SuccPrev->canFallThrough() && !CurUnAnalyzable &&
!SuccBB->isLandingPad()) {
MBB->moveBefore(SuccBB);
MadeChange = true;
diff --git a/lib/CodeGen/BranchFolding.h b/lib/CodeGen/BranchFolding.h
index 4920755c22..b087395640 100644
--- a/lib/CodeGen/BranchFolding.h
+++ b/lib/CodeGen/BranchFolding.h
@@ -105,18 +105,10 @@ namespace llvm {
unsigned CreateCommonTailOnlyBlock(MachineBasicBlock *&PredBB,
unsigned maxCommonTailLength);
- bool TailDuplicateBlocks(MachineFunction &MF);
- bool TailDuplicate(MachineBasicBlock *TailBB, MachineFunction &MF);
-
bool OptimizeBranches(MachineFunction &MF);
bool OptimizeBlock(MachineBasicBlock *MBB);
void RemoveDeadBlock(MachineBasicBlock *MBB);
bool OptimizeImpDefsBlock(MachineBasicBlock *MBB);
-
- bool CanFallThrough(MachineBasicBlock *CurBB);
- bool CanFallThrough(MachineBasicBlock *CurBB, bool BranchUnAnalyzable,
- MachineBasicBlock *TBB, MachineBasicBlock *FBB,
- const SmallVectorImpl<MachineOperand> &Cond);
};
}
diff --git a/lib/CodeGen/CMakeLists.txt b/lib/CodeGen/CMakeLists.txt
index 9583edcbe4..6f86614c90 100644
--- a/lib/CodeGen/CMakeLists.txt
+++ b/lib/CodeGen/CMakeLists.txt
@@ -63,6 +63,7 @@ add_llvm_library(LLVMCodeGen
StackProtector.cpp
StackSlotColoring.cpp
StrongPHIElimination.cpp
+ TailDuplication.cpp
TargetInstrInfoImpl.cpp
TwoAddressInstructionPass.cpp
UnreachableBlockElim.cpp
diff --git a/lib/CodeGen/LLVMTargetMachine.cpp b/lib/CodeGen/LLVMTargetMachine.cpp
index 740c950842..9164f8278a 100644
--- a/lib/CodeGen/LLVMTargetMachine.cpp
+++ b/lib/CodeGen/LLVMTargetMachine.cpp
@@ -35,6 +35,8 @@ static cl::opt<bool> DisablePostRA("disable-post-ra", cl::Hidden,
cl::desc("Disable Post Regalloc"));
static cl::opt<bool> DisableBranchFold("disable-branch-fold", cl::Hidden,
cl::desc("Disable branch folding"));
+static cl::opt<bool> DisableTailDuplicate("disable-tail-duplicate", cl::Hidden,
+ cl::desc("Disable tail duplication"));
static cl::opt<bool> DisableCodePlace("disable-code-place", cl::Hidden,
cl::desc("Disable code placement"));
static cl::opt<bool> DisableSSC("disable-ssc", cl::Hidden,
@@ -344,6 +346,12 @@ bool LLVMTargetMachine::addCommonCodeGenPasses(PassManagerBase &PM,
printAndVerify(PM, "After BranchFolding");
}
+ // Tail duplication.
+ if (OptLevel != CodeGenOpt::None && !DisableTailDuplicate) {
+ PM.add(createTailDuplicationPass());
+ printAndVerify(PM, "After TailDuplication");
+ }
+
PM.add(createGCMachineCodeAnalysisPass());
if (PrintGCInfo)
diff --git a/lib/CodeGen/MachineBasicBlock.cpp b/lib/CodeGen/MachineBasicBlock.cpp
index e6a187820c..e55e3694bc 100644
--- a/lib/CodeGen/MachineBasicBlock.cpp
+++ b/lib/CodeGen/MachineBasicBlock.cpp
@@ -362,6 +362,51 @@ bool MachineBasicBlock::isLayoutSuccessor(const MachineBasicBlock *MBB) const {
return next(I) == MachineFunction::const_iterator(MBB);
}
+bool MachineBasicBlock::canFallThrough() {
+ MachineBasicBlock *TBB = 0, *FBB = 0;
+ SmallVector<MachineOperand, 4> Cond;
+ const TargetInstrInfo *TII = getParent()->getTarget().getInstrInfo();
+ bool BranchUnAnalyzable = TII->AnalyzeBranch(*this, TBB, FBB, Cond, true);
+
+ MachineFunction::iterator Fallthrough = this;
+ ++Fallthrough;
+ // If FallthroughBlock is off the end of the function, it can't fall through.
+ if (Fallthrough == getParent()->end())
+ return false;
+
+ // If FallthroughBlock isn't a successor, no fallthrough is possible.
+ if (!isSuccessor(Fallthrough))
+ return false;
+
+ // If we couldn't analyze the branch, examine the last instruction.
+ // If the block doesn't end in a known control barrier, assume fallthrough
+ // is possible. The isPredicable check is needed because this code can be
+ // called during IfConversion, where an instruction which is normally a
+ // Barrier is predicated and thus no longer an actual control barrier. This
+ // is over-conservative though, because if an instruction isn't actually
+ // predicated we could still treat it like a barrier.
+ if (BranchUnAnalyzable)
+ return empty() || !back().getDesc().isBarrier() ||
+ back().getDesc().isPredicable();
+
+ // If there is no branch, control always falls through.
+ if (TBB == 0) return true;
+
+ // If there is some explicit branch to the fallthrough block, it can obviously
+ // reach, even though the branch should get folded to fall through implicitly.
+ if (MachineFunction::iterator(TBB) == Fallthrough ||
+ MachineFunction::iterator(FBB) == Fallthrough)
+ return true;
+
+ // If it's an unconditional branch to some block not the fall through, it
+ // doesn't fall through.
+ if (Cond.empty()) return false;
+
+ // Otherwise, if it is conditional and has no explicit false block, it falls
+ // through.
+ return FBB == 0;
+}
+
/// removeFromParent - This method unlinks 'this' from the containing function,
/// and returns it, but does not delete it.
MachineBasicBlock *MachineBasicBlock::removeFromParent() {
diff --git a/lib/CodeGen/TailDuplication.cpp b/lib/CodeGen/TailDuplication.cpp
new file mode 100644
index 0000000000..9fa05876b0
--- /dev/null
+++ b/lib/CodeGen/TailDuplication.cpp
@@ -0,0 +1,250 @@
+//===-- TailDuplication.cpp - Duplicate blocks into predecessors' tails ---===//
+//
+// The LLVM Compiler Infrastructure
+//
+// This file is distributed under the University of Illinois Open Source
+// License. See LICENSE.TXT for details.
+//
+//===----------------------------------------------------------------------===//
+//
+// This pass duplicates basic blocks ending in unconditional branches into
+// the tails of their predecessors.
+//
+//===----------------------------------------------------------------------===//
+
+#define DEBUG_TYPE "tailduplication"
+#include "llvm/Function.h"
+#include "llvm/CodeGen/Passes.h"
+#include "llvm/CodeGen/MachineModuleInfo.h"
+#include "llvm/CodeGen/MachineFunctionPass.h"
+#include "llvm/Target/TargetInstrInfo.h"
+#include "llvm/Support/CommandLine.h"
+#include "llvm/Support/Debug.h"
+#include "llvm/Support/raw_ostream.h"
+#include "llvm/ADT/SmallSet.h"
+#include "llvm/ADT/SetVector.h"
+#include "llvm/ADT/Statistic.h"
+using namespace llvm;
+
+STATISTIC(NumTailDups , "Number of tail duplicated blocks");
+STATISTIC(NumInstrDups , "Additional instructions due to tail duplication");
+STATISTIC(NumDeadBlocks, "Number of dead blocks removed");
+
+// Heuristic for tail duplication.
+static cl::opt<unsigned>
+TailDuplicateSize("tail-dup-size",
+ cl::desc("Maximum instructions to consider tail duplicating"),
+ cl::init(2), cl::Hidden);
+
+namespace {
+ /// TailDuplicationPass - Perform tail duplication.
+ class TailDuplicationPass : public MachineFunctionPass {
+ const TargetInstrInfo *TII;
+ MachineModuleInfo *MMI;
+
+ public:
+ static char ID;
+ explicit TailDuplicationPass() : MachineFunctionPass(&ID) {}
+
+ virtual bool runOnMachineFunction(MachineFunction &MF);
+ virtual const char *getPassName() const { return "Tail Duplication"; }
+
+ private:
+ bool TailDuplicateBlocks(MachineFunction &MF);
+ bool TailDuplicate(MachineBasicBlock *TailBB, MachineFunction &MF);
+ void RemoveDeadBlock(MachineBasicBlock *MBB);
+ };
+
+ char TailDuplicationPass::ID = 0;
+}
+
+FunctionPass *llvm::createTailDuplicationPass() {
+ return new TailDuplicationPass();
+}
+
+bool TailDuplicationPass::runOnMachineFunction(MachineFunction &MF) {
+ TII = MF.getTarget().getInstrInfo();
+ MMI = getAnalysisIfAvailable<MachineModuleInfo>();
+
+ bool MadeChange = false;
+ bool MadeChangeThisIteration = true;
+ while (MadeChangeThisIteration) {
+ MadeChangeThisIteration = false;
+ MadeChangeThisIteration |= TailDuplicateBlocks(MF);
+ MadeChange |= MadeChangeThisIteration;
+ }
+
+ return MadeChange;
+}
+
+/// TailDuplicateBlocks - Look for small blocks that are unconditionally
+/// branched to and do not fall through. Tail-duplicate their instructions
+/// into their predecessors to eliminate (dynamic) branches.
+bool TailDuplicationPass::TailDuplicateBlocks(MachineFunction &MF) {
+ bool MadeChange = false;
+
+ for (MachineFunction::iterator I = ++MF.begin(), E = MF.end(); I != E; ) {
+ MachineBasicBlock *MBB = I++;
+
+ // Only duplicate blocks that end with unconditional branches.
+ if (MBB->canFallThrough())
+ continue;
+
+ MadeChange |= TailDuplicate(MBB, MF);
+
+ // If it is dead, remove it.
+ if (MBB->pred_empty()) {
+ NumInstrDups -= MBB->size();
+ RemoveDeadBlock(MBB);
+ MadeChange = true;
+ ++NumDeadBlocks;
+ }
+ }
+ return MadeChange;
+}
+
+/// TailDuplicate - If it is profitable, duplicate TailBB's contents in each
+/// of its predecessors.
+bool TailDuplicationPass::TailDuplicate(MachineBasicBlock *TailBB,
+ MachineFunction &MF) {
+ // Don't try to tail-duplicate single-block loops.
+ if (TailBB->isSuccessor(TailBB))
+ return false;
+
+ // Set the limit on the number of instructions to duplicate, with a default
+ // of one less than the tail-merge threshold. When optimizing for size,
+ // duplicate only one, because one branch instruction can be eliminated to
+ // compensate for the duplication.
+ unsigned MaxDuplicateCount;
+ if (MF.getFunction()->hasFnAttr(Attribute::OptimizeForSize))
+ MaxDuplicateCount = 1;
+ else if (TII->isProfitableToDuplicateIndirectBranch() &&
+ !TailBB->empty() && TailBB->back().getDesc().isIndirectBranch())
+ // If the target has hardware branch prediction that can handle indirect
+ // branches, duplicating them can often make them predictable when there
+ // are common paths through the code. The limit needs to be high enough
+ // to allow undoing the effects of tail merging.
+ MaxDuplicateCount = 20;
+ else
+ MaxDuplicateCount = TailDuplicateSize;
+
+ // Check the instructions in the block to determine whether tail-duplication
+ // is invalid or unlikely to be profitable.
+ unsigned i = 0;
+ bool HasCall = false;
+ for (MachineBasicBlock::iterator I = TailBB->begin();
+ I != TailBB->end(); ++I, ++i) {
+ // Non-duplicable things shouldn't be tail-duplicated.
+ if (I->getDesc().isNotDuplicable()) return false;
+ // Don't duplicate more than the threshold.
+ if (i == MaxDuplicateCount) return false;
+ // Remember if we saw a call.
+ if (I->getDesc().isCall()) HasCall = true;
+ }
+ // Heuristically, don't tail-duplicate calls if it would expand code size,
+ // as it's less likely to be worth the extra cost.
+ if (i > 1 && HasCall)
+ return false;
+
+ // Iterate through all the unique predecessors and tail-duplicate this
+ // block into them, if possible. Copying the list ahead of time also
+ // avoids trouble with the predecessor list reallocating.
+ bool Changed = false;
+ SmallSetVector<MachineBasicBlock *, 8> Preds(TailBB->pred_begin(),
+ TailBB->pred_end());
+ for (SmallSetVector<MachineBasicBlock *, 8>::iterator PI = Preds.begin(),
+ PE = Preds.end(); PI != PE; ++PI) {
+ MachineBasicBlock *PredBB = *PI;
+
+ assert(TailBB != PredBB &&
+ "Single-block loop should have been rejected earlier!");
+ if (PredBB->succ_size() > 1) continue;
+
+ MachineBasicBlock *PredTBB, *PredFBB;
+ SmallVector<MachineOperand, 4> PredCond;
+ if (TII->AnalyzeBranch(*PredBB, PredTBB, PredFBB, PredCond, true))
+ continue;
+ if (!PredCond.empty())
+ continue;
+ // EH edges are ignored by AnalyzeBranch.
+ if (PredBB->succ_size() != 1)
+ continue;
+ // Don't duplicate into a fall-through predecessor (at least for now).
+ if (PredBB->isLayoutSuccessor(TailBB) && PredBB->canFallThrough())
+ continue;
+
+ DEBUG(errs() << "\nTail-duplicating into PredBB: " << *PredBB
+ << "From Succ: " << *TailBB);
+
+ // Remove PredBB's unconditional branch.
+ TII->RemoveBranch(*PredBB);
+ // Clone the contents of TailBB into PredBB.
+ for (MachineBasicBlock::iterator I = TailBB->begin(), E = TailBB->end();
+ I != E; ++I) {
+ MachineInstr *NewMI = MF.CloneMachineInstr(I);
+ PredBB->insert(PredBB->end(), NewMI);
+ }
+ NumInstrDups += TailBB->size() - 1; // subtract one for removed branch
+
+ // Update the CFG.
+ PredBB->removeSuccessor(PredBB->succ_begin());
+ assert(PredBB->succ_empty() &&
+ "TailDuplicate called on block with multiple successors!");
+ for (MachineBasicBlock::succ_iterator I = TailBB->succ_begin(),
+ E = TailBB->succ_end(); I != E; ++I)
+ PredBB->addSuccessor(*I);
+
+ Changed = true;
+ ++NumTailDups;
+ }
+
+ // If TailBB was duplicated into all its predecessors except for the prior
+ // block, which falls through unconditionally, move the contents of this
+ // block into the prior block.
+ MachineBasicBlock &PrevBB = *prior(MachineFunction::iterator(TailBB));
+ MachineBasicBlock *PriorTBB = 0, *PriorFBB = 0;
+ SmallVector<MachineOperand, 4> PriorCond;
+ bool PriorUnAnalyzable =
+ TII->AnalyzeBranch(PrevBB, PriorTBB, PriorFBB, PriorCond, true);
+ // This has to check PrevBB->succ_size() because EH edges are ignored by
+ // AnalyzeBranch.
+ if (!PriorUnAnalyzable && PriorCond.empty() && !PriorTBB &&
+ TailBB->pred_size() == 1 && PrevBB.succ_size() == 1 &&
+ !TailBB->hasAddressTaken()) {
+ DEBUG(errs() << "\nMerging into block: " << PrevBB
+ << "From MBB: " << *TailBB);
+ PrevBB.splice(PrevBB.end(), TailBB, TailBB->begin(), TailBB->end());
+ PrevBB.removeSuccessor(PrevBB.succ_begin());;
+ assert(PrevBB.succ_empty());
+ PrevBB.transferSuccessors(TailBB);
+ Changed = true;
+ }
+
+ return Changed;
+}
+
+/// RemoveDeadBlock - Remove the specified dead machine basic block from the
+/// function, updating the CFG.
+void TailDuplicationPass::RemoveDeadBlock(MachineBasicBlock *MBB) {
+ assert(MBB->pred_empty() && "MBB must be dead!");
+ DEBUG(errs() << "\nRemoving MBB: " << *MBB);
+
+ // Remove all successors.
+ while (!MBB->succ_empty())
+ MBB->removeSuccessor(MBB->succ_end()-1);
+
+ // If there are any labels in the basic block, unregister them from
+ // MachineModuleInfo.
+ if (MMI && !MBB->empty()) {
+ for (MachineBasicBlock::iterator I = MBB->begin(), E = MBB->end();
+ I != E; ++I) {
+ if (I->isLabel())
+ // The label ID # is always operand #0, an immediate.
+ MMI->InvalidateLabel(I->getOperand(0).getImm());
+ }
+ }
+
+ // Remove the block.
+ MBB->eraseFromParent();
+}
+
generated by cgit v1.2.3-18-g5258 (git 2.32.0) at 2025-06-23 11:50:53 +0000