1 //===- CodeGenPrepare.cpp - Prepare a function for code generation --------===//
3 // The LLVM Compiler Infrastructure
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
8 //===----------------------------------------------------------------------===//
10 // This pass munges the code in the input function to better prepare it for
11 // SelectionDAG-based code generation. This works around limitations in it's
12 // basic-block-at-a-time approach. It should eventually be removed.
14 //===----------------------------------------------------------------------===//
16 #define DEBUG_TYPE "codegenprepare"
17 #include "llvm/Transforms/Scalar.h"
18 #include "llvm/Constants.h"
19 #include "llvm/DerivedTypes.h"
20 #include "llvm/Function.h"
21 #include "llvm/InlineAsm.h"
22 #include "llvm/Instructions.h"
23 #include "llvm/IntrinsicInst.h"
24 #include "llvm/Pass.h"
25 #include "llvm/Analysis/ProfileInfo.h"
26 #include "llvm/Target/TargetData.h"
27 #include "llvm/Target/TargetLowering.h"
28 #include "llvm/Transforms/Utils/AddrModeMatcher.h"
29 #include "llvm/Transforms/Utils/BasicBlockUtils.h"
30 #include "llvm/Transforms/Utils/Local.h"
31 #include "llvm/ADT/DenseMap.h"
32 #include "llvm/ADT/SmallSet.h"
33 #include "llvm/Assembly/Writer.h"
34 #include "llvm/Support/CallSite.h"
35 #include "llvm/Support/CommandLine.h"
36 #include "llvm/Support/Debug.h"
37 #include "llvm/Support/GetElementPtrTypeIterator.h"
38 #include "llvm/Support/PatternMatch.h"
39 #include "llvm/Support/raw_ostream.h"
41 using namespace llvm::PatternMatch;
43 static cl::opt<bool> FactorCommonPreds("split-critical-paths-tweak",
44 cl::init(false), cl::Hidden);
47 class CodeGenPrepare : public FunctionPass {
48 /// TLI - Keep a pointer of a TargetLowering to consult for determining
49 /// transformation profitability.
50 const TargetLowering *TLI;
53 /// BackEdges - Keep a set of all the loop back edges.
55 SmallSet<std::pair<const BasicBlock*, const BasicBlock*>, 8> BackEdges;
57 static char ID; // Pass identification, replacement for typeid
58 explicit CodeGenPrepare(const TargetLowering *tli = 0)
59 : FunctionPass(&ID), TLI(tli) {}
60 bool runOnFunction(Function &F);
62 virtual void getAnalysisUsage(AnalysisUsage &AU) const {
63 AU.addPreserved<ProfileInfo>();
66 virtual void releaseMemory() {
71 bool EliminateMostlyEmptyBlocks(Function &F);
72 bool CanMergeBlocks(const BasicBlock *BB, const BasicBlock *DestBB) const;
73 void EliminateMostlyEmptyBlock(BasicBlock *BB);
74 bool OptimizeBlock(BasicBlock &BB);
75 bool OptimizeMemoryInst(Instruction *I, Value *Addr, const Type *AccessTy,
76 DenseMap<Value*,Value*> &SunkAddrs);
77 bool OptimizeInlineAsmInst(Instruction *I, CallSite CS,
78 DenseMap<Value*,Value*> &SunkAddrs);
79 bool MoveExtToFormExtLoad(Instruction *I);
80 bool OptimizeExtUses(Instruction *I);
81 void findLoopBackEdges(const Function &F);
85 char CodeGenPrepare::ID = 0;
86 static RegisterPass<CodeGenPrepare> X("codegenprepare",
87 "Optimize for code generation");
89 FunctionPass *llvm::createCodeGenPreparePass(const TargetLowering *TLI) {
90 return new CodeGenPrepare(TLI);
93 /// findLoopBackEdges - Do a DFS walk to find loop back edges.
95 void CodeGenPrepare::findLoopBackEdges(const Function &F) {
96 SmallVector<std::pair<const BasicBlock*,const BasicBlock*>, 32> Edges;
97 FindFunctionBackedges(F, Edges);
99 BackEdges.insert(Edges.begin(), Edges.end());
103 bool CodeGenPrepare::runOnFunction(Function &F) {
104 bool EverMadeChange = false;
106 PFI = getAnalysisIfAvailable<ProfileInfo>();
107 // First pass, eliminate blocks that contain only PHI nodes and an
108 // unconditional branch.
109 EverMadeChange |= EliminateMostlyEmptyBlocks(F);
111 // Now find loop back edges.
112 findLoopBackEdges(F);
114 bool MadeChange = true;
117 for (Function::iterator BB = F.begin(), E = F.end(); BB != E; ++BB)
118 MadeChange |= OptimizeBlock(*BB);
119 EverMadeChange |= MadeChange;
121 return EverMadeChange;
124 /// EliminateMostlyEmptyBlocks - eliminate blocks that contain only PHI nodes,
125 /// debug info directives, and an unconditional branch. Passes before isel
126 /// (e.g. LSR/loopsimplify) often split edges in ways that are non-optimal for
127 /// isel. Start by eliminating these blocks so we can split them the way we
129 bool CodeGenPrepare::EliminateMostlyEmptyBlocks(Function &F) {
130 bool MadeChange = false;
131 // Note that this intentionally skips the entry block.
132 for (Function::iterator I = ++F.begin(), E = F.end(); I != E; ) {
133 BasicBlock *BB = I++;
135 // If this block doesn't end with an uncond branch, ignore it.
136 BranchInst *BI = dyn_cast<BranchInst>(BB->getTerminator());
137 if (!BI || !BI->isUnconditional())
140 // If the instruction before the branch (skipping debug info) isn't a phi
141 // node, then other stuff is happening here.
142 BasicBlock::iterator BBI = BI;
143 if (BBI != BB->begin()) {
145 while (isa<DbgInfoIntrinsic>(BBI)) {
146 if (BBI == BB->begin())
150 if (!isa<DbgInfoIntrinsic>(BBI) && !isa<PHINode>(BBI))
154 // Do not break infinite loops.
155 BasicBlock *DestBB = BI->getSuccessor(0);
159 if (!CanMergeBlocks(BB, DestBB))
162 EliminateMostlyEmptyBlock(BB);
168 /// CanMergeBlocks - Return true if we can merge BB into DestBB if there is a
169 /// single uncond branch between them, and BB contains no other non-phi
171 bool CodeGenPrepare::CanMergeBlocks(const BasicBlock *BB,
172 const BasicBlock *DestBB) const {
173 // We only want to eliminate blocks whose phi nodes are used by phi nodes in
174 // the successor. If there are more complex condition (e.g. preheaders),
175 // don't mess around with them.
176 BasicBlock::const_iterator BBI = BB->begin();
177 while (const PHINode *PN = dyn_cast<PHINode>(BBI++)) {
178 for (Value::use_const_iterator UI = PN->use_begin(), E = PN->use_end();
180 const Instruction *User = cast<Instruction>(*UI);
181 if (User->getParent() != DestBB || !isa<PHINode>(User))
183 // If User is inside DestBB block and it is a PHINode then check
184 // incoming value. If incoming value is not from BB then this is
185 // a complex condition (e.g. preheaders) we want to avoid here.
186 if (User->getParent() == DestBB) {
187 if (const PHINode *UPN = dyn_cast<PHINode>(User))
188 for (unsigned I = 0, E = UPN->getNumIncomingValues(); I != E; ++I) {
189 Instruction *Insn = dyn_cast<Instruction>(UPN->getIncomingValue(I));
190 if (Insn && Insn->getParent() == BB &&
191 Insn->getParent() != UPN->getIncomingBlock(I))
198 // If BB and DestBB contain any common predecessors, then the phi nodes in BB
199 // and DestBB may have conflicting incoming values for the block. If so, we
200 // can't merge the block.
201 const PHINode *DestBBPN = dyn_cast<PHINode>(DestBB->begin());
202 if (!DestBBPN) return true; // no conflict.
204 // Collect the preds of BB.
205 SmallPtrSet<const BasicBlock*, 16> BBPreds;
206 if (const PHINode *BBPN = dyn_cast<PHINode>(BB->begin())) {
207 // It is faster to get preds from a PHI than with pred_iterator.
208 for (unsigned i = 0, e = BBPN->getNumIncomingValues(); i != e; ++i)
209 BBPreds.insert(BBPN->getIncomingBlock(i));
211 BBPreds.insert(pred_begin(BB), pred_end(BB));
214 // Walk the preds of DestBB.
215 for (unsigned i = 0, e = DestBBPN->getNumIncomingValues(); i != e; ++i) {
216 BasicBlock *Pred = DestBBPN->getIncomingBlock(i);
217 if (BBPreds.count(Pred)) { // Common predecessor?
218 BBI = DestBB->begin();
219 while (const PHINode *PN = dyn_cast<PHINode>(BBI++)) {
220 const Value *V1 = PN->getIncomingValueForBlock(Pred);
221 const Value *V2 = PN->getIncomingValueForBlock(BB);
223 // If V2 is a phi node in BB, look up what the mapped value will be.
224 if (const PHINode *V2PN = dyn_cast<PHINode>(V2))
225 if (V2PN->getParent() == BB)
226 V2 = V2PN->getIncomingValueForBlock(Pred);
228 // If there is a conflict, bail out.
229 if (V1 != V2) return false;
238 /// EliminateMostlyEmptyBlock - Eliminate a basic block that have only phi's and
239 /// an unconditional branch in it.
240 void CodeGenPrepare::EliminateMostlyEmptyBlock(BasicBlock *BB) {
241 BranchInst *BI = cast<BranchInst>(BB->getTerminator());
242 BasicBlock *DestBB = BI->getSuccessor(0);
244 DEBUG(dbgs() << "MERGING MOSTLY EMPTY BLOCKS - BEFORE:\n" << *BB << *DestBB);
246 // If the destination block has a single pred, then this is a trivial edge,
248 if (BasicBlock *SinglePred = DestBB->getSinglePredecessor()) {
249 if (SinglePred != DestBB) {
250 // Remember if SinglePred was the entry block of the function. If so, we
251 // will need to move BB back to the entry position.
252 bool isEntry = SinglePred == &SinglePred->getParent()->getEntryBlock();
253 MergeBasicBlockIntoOnlyPred(DestBB, this);
255 if (isEntry && BB != &BB->getParent()->getEntryBlock())
256 BB->moveBefore(&BB->getParent()->getEntryBlock());
258 DEBUG(dbgs() << "AFTER:\n" << *DestBB << "\n\n\n");
263 // Otherwise, we have multiple predecessors of BB. Update the PHIs in DestBB
264 // to handle the new incoming edges it is about to have.
266 for (BasicBlock::iterator BBI = DestBB->begin();
267 (PN = dyn_cast<PHINode>(BBI)); ++BBI) {
268 // Remove the incoming value for BB, and remember it.
269 Value *InVal = PN->removeIncomingValue(BB, false);
271 // Two options: either the InVal is a phi node defined in BB or it is some
272 // value that dominates BB.
273 PHINode *InValPhi = dyn_cast<PHINode>(InVal);
274 if (InValPhi && InValPhi->getParent() == BB) {
275 // Add all of the input values of the input PHI as inputs of this phi.
276 for (unsigned i = 0, e = InValPhi->getNumIncomingValues(); i != e; ++i)
277 PN->addIncoming(InValPhi->getIncomingValue(i),
278 InValPhi->getIncomingBlock(i));
280 // Otherwise, add one instance of the dominating value for each edge that
281 // we will be adding.
282 if (PHINode *BBPN = dyn_cast<PHINode>(BB->begin())) {
283 for (unsigned i = 0, e = BBPN->getNumIncomingValues(); i != e; ++i)
284 PN->addIncoming(InVal, BBPN->getIncomingBlock(i));
286 for (pred_iterator PI = pred_begin(BB), E = pred_end(BB); PI != E; ++PI)
287 PN->addIncoming(InVal, *PI);
292 // The PHIs are now updated, change everything that refers to BB to use
293 // DestBB and remove BB.
294 BB->replaceAllUsesWith(DestBB);
296 PFI->replaceAllUses(BB, DestBB);
297 PFI->removeEdge(ProfileInfo::getEdge(BB, DestBB));
299 BB->eraseFromParent();
301 DEBUG(dbgs() << "AFTER:\n" << *DestBB << "\n\n\n");
305 /// SplitEdgeNicely - Split the critical edge from TI to its specified
306 /// successor if it will improve codegen. We only do this if the successor has
307 /// phi nodes (otherwise critical edges are ok). If there is already another
308 /// predecessor of the succ that is empty (and thus has no phi nodes), use it
309 /// instead of introducing a new block.
310 static void SplitEdgeNicely(TerminatorInst *TI, unsigned SuccNum,
311 SmallSet<std::pair<const BasicBlock*,
312 const BasicBlock*>, 8> &BackEdges,
314 BasicBlock *TIBB = TI->getParent();
315 BasicBlock *Dest = TI->getSuccessor(SuccNum);
316 assert(isa<PHINode>(Dest->begin()) &&
317 "This should only be called if Dest has a PHI!");
319 // Do not split edges to EH landing pads.
320 if (InvokeInst *Invoke = dyn_cast<InvokeInst>(TI)) {
321 if (Invoke->getSuccessor(1) == Dest)
326 // As a hack, never split backedges of loops. Even though the copy for any
327 // PHIs inserted on the backedge would be dead for exits from the loop, we
328 // assume that the cost of *splitting* the backedge would be too high.
329 if (BackEdges.count(std::make_pair(TIBB, Dest)))
332 if (!FactorCommonPreds) {
333 /// TIPHIValues - This array is lazily computed to determine the values of
334 /// PHIs in Dest that TI would provide.
335 SmallVector<Value*, 32> TIPHIValues;
337 // Check to see if Dest has any blocks that can be used as a split edge for
339 for (pred_iterator PI = pred_begin(Dest), E = pred_end(Dest); PI != E; ++PI) {
340 BasicBlock *Pred = *PI;
341 // To be usable, the pred has to end with an uncond branch to the dest.
342 BranchInst *PredBr = dyn_cast<BranchInst>(Pred->getTerminator());
343 if (!PredBr || !PredBr->isUnconditional())
345 // Must be empty other than the branch and debug info.
346 BasicBlock::iterator I = Pred->begin();
347 while (isa<DbgInfoIntrinsic>(I))
349 if (dyn_cast<Instruction>(I) != PredBr)
351 // Cannot be the entry block; its label does not get emitted.
352 if (Pred == &(Dest->getParent()->getEntryBlock()))
355 // Finally, since we know that Dest has phi nodes in it, we have to make
356 // sure that jumping to Pred will have the same effect as going to Dest in
357 // terms of PHI values.
360 bool FoundMatch = true;
361 for (BasicBlock::iterator I = Dest->begin();
362 (PN = dyn_cast<PHINode>(I)); ++I, ++PHINo) {
363 if (PHINo == TIPHIValues.size())
364 TIPHIValues.push_back(PN->getIncomingValueForBlock(TIBB));
366 // If the PHI entry doesn't work, we can't use this pred.
367 if (TIPHIValues[PHINo] != PN->getIncomingValueForBlock(Pred)) {
373 // If we found a workable predecessor, change TI to branch to Succ.
375 ProfileInfo *PFI = P->getAnalysisIfAvailable<ProfileInfo>();
377 PFI->splitEdge(TIBB, Dest, Pred);
378 Dest->removePredecessor(TIBB);
379 TI->setSuccessor(SuccNum, Pred);
384 SplitCriticalEdge(TI, SuccNum, P, true);
389 SmallVector<Value*, 8> TIPHIValues;
390 for (BasicBlock::iterator I = Dest->begin();
391 (PN = dyn_cast<PHINode>(I)); ++I)
392 TIPHIValues.push_back(PN->getIncomingValueForBlock(TIBB));
394 SmallVector<BasicBlock*, 8> IdenticalPreds;
395 for (pred_iterator PI = pred_begin(Dest), E = pred_end(Dest); PI != E; ++PI) {
396 BasicBlock *Pred = *PI;
397 if (BackEdges.count(std::make_pair(Pred, Dest)))
400 IdenticalPreds.push_back(Pred);
402 bool Identical = true;
404 for (BasicBlock::iterator I = Dest->begin();
405 (PN = dyn_cast<PHINode>(I)); ++I, ++PHINo)
406 if (TIPHIValues[PHINo] != PN->getIncomingValueForBlock(Pred)) {
411 IdenticalPreds.push_back(Pred);
415 assert(!IdenticalPreds.empty());
416 SplitBlockPredecessors(Dest, &IdenticalPreds[0], IdenticalPreds.size(),
421 /// OptimizeNoopCopyExpression - If the specified cast instruction is a noop
422 /// copy (e.g. it's casting from one pointer type to another, i32->i8 on PPC),
423 /// sink it into user blocks to reduce the number of virtual
424 /// registers that must be created and coalesced.
426 /// Return true if any changes are made.
428 static bool OptimizeNoopCopyExpression(CastInst *CI, const TargetLowering &TLI){
429 // If this is a noop copy,
430 EVT SrcVT = TLI.getValueType(CI->getOperand(0)->getType());
431 EVT DstVT = TLI.getValueType(CI->getType());
433 // This is an fp<->int conversion?
434 if (SrcVT.isInteger() != DstVT.isInteger())
437 // If this is an extension, it will be a zero or sign extension, which
439 if (SrcVT.bitsLT(DstVT)) return false;
441 // If these values will be promoted, find out what they will be promoted
442 // to. This helps us consider truncates on PPC as noop copies when they
444 if (TLI.getTypeAction(CI->getContext(), SrcVT) == TargetLowering::Promote)
445 SrcVT = TLI.getTypeToTransformTo(CI->getContext(), SrcVT);
446 if (TLI.getTypeAction(CI->getContext(), DstVT) == TargetLowering::Promote)
447 DstVT = TLI.getTypeToTransformTo(CI->getContext(), DstVT);
449 // If, after promotion, these are the same types, this is a noop copy.
453 BasicBlock *DefBB = CI->getParent();
455 /// InsertedCasts - Only insert a cast in each block once.
456 DenseMap<BasicBlock*, CastInst*> InsertedCasts;
458 bool MadeChange = false;
459 for (Value::use_iterator UI = CI->use_begin(), E = CI->use_end();
461 Use &TheUse = UI.getUse();
462 Instruction *User = cast<Instruction>(*UI);
464 // Figure out which BB this cast is used in. For PHI's this is the
465 // appropriate predecessor block.
466 BasicBlock *UserBB = User->getParent();
467 if (PHINode *PN = dyn_cast<PHINode>(User)) {
468 UserBB = PN->getIncomingBlock(UI);
471 // Preincrement use iterator so we don't invalidate it.
474 // If this user is in the same block as the cast, don't change the cast.
475 if (UserBB == DefBB) continue;
477 // If we have already inserted a cast into this block, use it.
478 CastInst *&InsertedCast = InsertedCasts[UserBB];
481 BasicBlock::iterator InsertPt = UserBB->getFirstNonPHI();
484 CastInst::Create(CI->getOpcode(), CI->getOperand(0), CI->getType(), "",
489 // Replace a use of the cast with a use of the new cast.
490 TheUse = InsertedCast;
493 // If we removed all uses, nuke the cast.
494 if (CI->use_empty()) {
495 CI->eraseFromParent();
502 /// OptimizeCmpExpression - sink the given CmpInst into user blocks to reduce
503 /// the number of virtual registers that must be created and coalesced. This is
504 /// a clear win except on targets with multiple condition code registers
505 /// (PowerPC), where it might lose; some adjustment may be wanted there.
507 /// Return true if any changes are made.
508 static bool OptimizeCmpExpression(CmpInst *CI) {
509 BasicBlock *DefBB = CI->getParent();
511 /// InsertedCmp - Only insert a cmp in each block once.
512 DenseMap<BasicBlock*, CmpInst*> InsertedCmps;
514 bool MadeChange = false;
515 for (Value::use_iterator UI = CI->use_begin(), E = CI->use_end();
517 Use &TheUse = UI.getUse();
518 Instruction *User = cast<Instruction>(*UI);
520 // Preincrement use iterator so we don't invalidate it.
523 // Don't bother for PHI nodes.
524 if (isa<PHINode>(User))
527 // Figure out which BB this cmp is used in.
528 BasicBlock *UserBB = User->getParent();
530 // If this user is in the same block as the cmp, don't change the cmp.
531 if (UserBB == DefBB) continue;
533 // If we have already inserted a cmp into this block, use it.
534 CmpInst *&InsertedCmp = InsertedCmps[UserBB];
537 BasicBlock::iterator InsertPt = UserBB->getFirstNonPHI();
540 CmpInst::Create(CI->getOpcode(),
541 CI->getPredicate(), CI->getOperand(0),
542 CI->getOperand(1), "", InsertPt);
546 // Replace a use of the cmp with a use of the new cmp.
547 TheUse = InsertedCmp;
550 // If we removed all uses, nuke the cmp.
552 CI->eraseFromParent();
557 //===----------------------------------------------------------------------===//
558 // Memory Optimization
559 //===----------------------------------------------------------------------===//
561 /// IsNonLocalValue - Return true if the specified values are defined in a
562 /// different basic block than BB.
563 static bool IsNonLocalValue(Value *V, BasicBlock *BB) {
564 if (Instruction *I = dyn_cast<Instruction>(V))
565 return I->getParent() != BB;
569 /// OptimizeMemoryInst - Load and Store Instructions often have
570 /// addressing modes that can do significant amounts of computation. As such,
571 /// instruction selection will try to get the load or store to do as much
572 /// computation as possible for the program. The problem is that isel can only
573 /// see within a single block. As such, we sink as much legal addressing mode
574 /// stuff into the block as possible.
576 /// This method is used to optimize both load/store and inline asms with memory
578 bool CodeGenPrepare::OptimizeMemoryInst(Instruction *MemoryInst, Value *Addr,
579 const Type *AccessTy,
580 DenseMap<Value*,Value*> &SunkAddrs) {
581 // Figure out what addressing mode will be built up for this operation.
582 SmallVector<Instruction*, 16> AddrModeInsts;
583 ExtAddrMode AddrMode = AddressingModeMatcher::Match(Addr, AccessTy,MemoryInst,
584 AddrModeInsts, *TLI);
586 // Check to see if any of the instructions supersumed by this addr mode are
587 // non-local to I's BB.
588 bool AnyNonLocal = false;
589 for (unsigned i = 0, e = AddrModeInsts.size(); i != e; ++i) {
590 if (IsNonLocalValue(AddrModeInsts[i], MemoryInst->getParent())) {
596 // If all the instructions matched are already in this BB, don't do anything.
598 DEBUG(dbgs() << "CGP: Found local addrmode: " << AddrMode << "\n");
602 // Insert this computation right after this user. Since our caller is
603 // scanning from the top of the BB to the bottom, reuse of the expr are
604 // guaranteed to happen later.
605 BasicBlock::iterator InsertPt = MemoryInst;
607 // Now that we determined the addressing expression we want to use and know
608 // that we have to sink it into this block. Check to see if we have already
609 // done this for some other load/store instr in this block. If so, reuse the
611 Value *&SunkAddr = SunkAddrs[Addr];
613 DEBUG(dbgs() << "CGP: Reusing nonlocal addrmode: " << AddrMode << " for "
615 if (SunkAddr->getType() != Addr->getType())
616 SunkAddr = new BitCastInst(SunkAddr, Addr->getType(), "tmp", InsertPt);
618 DEBUG(dbgs() << "CGP: SINKING nonlocal addrmode: " << AddrMode << " for "
620 const Type *IntPtrTy =
621 TLI->getTargetData()->getIntPtrType(AccessTy->getContext());
625 // Start with the base register. Do this first so that subsequent address
626 // matching finds it last, which will prevent it from trying to match it
627 // as the scaled value in case it happens to be a mul. That would be
628 // problematic if we've sunk a different mul for the scale, because then
629 // we'd end up sinking both muls.
630 if (AddrMode.BaseReg) {
631 Value *V = AddrMode.BaseReg;
632 if (isa<PointerType>(V->getType()))
633 V = new PtrToIntInst(V, IntPtrTy, "sunkaddr", InsertPt);
634 if (V->getType() != IntPtrTy)
635 V = CastInst::CreateIntegerCast(V, IntPtrTy, /*isSigned=*/true,
636 "sunkaddr", InsertPt);
640 // Add the scale value.
641 if (AddrMode.Scale) {
642 Value *V = AddrMode.ScaledReg;
643 if (V->getType() == IntPtrTy) {
645 } else if (isa<PointerType>(V->getType())) {
646 V = new PtrToIntInst(V, IntPtrTy, "sunkaddr", InsertPt);
647 } else if (cast<IntegerType>(IntPtrTy)->getBitWidth() <
648 cast<IntegerType>(V->getType())->getBitWidth()) {
649 V = new TruncInst(V, IntPtrTy, "sunkaddr", InsertPt);
651 V = new SExtInst(V, IntPtrTy, "sunkaddr", InsertPt);
653 if (AddrMode.Scale != 1)
654 V = BinaryOperator::CreateMul(V, ConstantInt::get(IntPtrTy,
656 "sunkaddr", InsertPt);
658 Result = BinaryOperator::CreateAdd(Result, V, "sunkaddr", InsertPt);
663 // Add in the BaseGV if present.
664 if (AddrMode.BaseGV) {
665 Value *V = new PtrToIntInst(AddrMode.BaseGV, IntPtrTy, "sunkaddr",
668 Result = BinaryOperator::CreateAdd(Result, V, "sunkaddr", InsertPt);
673 // Add in the Base Offset if present.
674 if (AddrMode.BaseOffs) {
675 Value *V = ConstantInt::get(IntPtrTy, AddrMode.BaseOffs);
677 Result = BinaryOperator::CreateAdd(Result, V, "sunkaddr", InsertPt);
683 SunkAddr = Constant::getNullValue(Addr->getType());
685 SunkAddr = new IntToPtrInst(Result, Addr->getType(), "sunkaddr",InsertPt);
688 MemoryInst->replaceUsesOfWith(Addr, SunkAddr);
690 if (Addr->use_empty())
691 RecursivelyDeleteTriviallyDeadInstructions(Addr);
695 /// OptimizeInlineAsmInst - If there are any memory operands, use
696 /// OptimizeMemoryInst to sink their address computing into the block when
697 /// possible / profitable.
698 bool CodeGenPrepare::OptimizeInlineAsmInst(Instruction *I, CallSite CS,
699 DenseMap<Value*,Value*> &SunkAddrs) {
700 bool MadeChange = false;
701 InlineAsm *IA = cast<InlineAsm>(CS.getCalledValue());
703 // Do a prepass over the constraints, canonicalizing them, and building up the
704 // ConstraintOperands list.
705 std::vector<InlineAsm::ConstraintInfo>
706 ConstraintInfos = IA->ParseConstraints();
708 /// ConstraintOperands - Information about all of the constraints.
709 std::vector<TargetLowering::AsmOperandInfo> ConstraintOperands;
710 unsigned ArgNo = 0; // ArgNo - The argument of the CallInst.
711 for (unsigned i = 0, e = ConstraintInfos.size(); i != e; ++i) {
713 push_back(TargetLowering::AsmOperandInfo(ConstraintInfos[i]));
714 TargetLowering::AsmOperandInfo &OpInfo = ConstraintOperands.back();
716 // Compute the value type for each operand.
717 switch (OpInfo.Type) {
718 case InlineAsm::isOutput:
719 if (OpInfo.isIndirect)
720 OpInfo.CallOperandVal = CS.getArgument(ArgNo++);
722 case InlineAsm::isInput:
723 OpInfo.CallOperandVal = CS.getArgument(ArgNo++);
725 case InlineAsm::isClobber:
730 // Compute the constraint code and ConstraintType to use.
731 TLI->ComputeConstraintToUse(OpInfo, SDValue(),
732 OpInfo.ConstraintType == TargetLowering::C_Memory);
734 if (OpInfo.ConstraintType == TargetLowering::C_Memory &&
736 Value *OpVal = OpInfo.CallOperandVal;
737 MadeChange |= OptimizeMemoryInst(I, OpVal, OpVal->getType(), SunkAddrs);
744 /// MoveExtToFormExtLoad - Move a zext or sext fed by a load into the same
745 /// basic block as the load, unless conditions are unfavorable. This allows
746 /// SelectionDAG to fold the extend into the load.
748 bool CodeGenPrepare::MoveExtToFormExtLoad(Instruction *I) {
749 // Look for a load being extended.
750 LoadInst *LI = dyn_cast<LoadInst>(I->getOperand(0));
751 if (!LI) return false;
753 // If they're already in the same block, there's nothing to do.
754 if (LI->getParent() == I->getParent())
757 // If the load has other users and the truncate is not free, this probably
759 if (!LI->hasOneUse() &&
760 TLI && !TLI->isTruncateFree(I->getType(), LI->getType()))
763 // Check whether the target supports casts folded into loads.
765 if (isa<ZExtInst>(I))
766 LType = ISD::ZEXTLOAD;
768 assert(isa<SExtInst>(I) && "Unexpected ext type!");
769 LType = ISD::SEXTLOAD;
771 if (TLI && !TLI->isLoadExtLegal(LType, TLI->getValueType(LI->getType())))
774 // Move the extend into the same block as the load, so that SelectionDAG
776 I->removeFromParent();
781 bool CodeGenPrepare::OptimizeExtUses(Instruction *I) {
782 BasicBlock *DefBB = I->getParent();
784 // If both result of the {s|z}xt and its source are live out, rewrite all
785 // other uses of the source with result of extension.
786 Value *Src = I->getOperand(0);
787 if (Src->hasOneUse())
790 // Only do this xform if truncating is free.
791 if (TLI && !TLI->isTruncateFree(I->getType(), Src->getType()))
794 // Only safe to perform the optimization if the source is also defined in
796 if (!isa<Instruction>(Src) || DefBB != cast<Instruction>(Src)->getParent())
799 bool DefIsLiveOut = false;
800 for (Value::use_iterator UI = I->use_begin(), E = I->use_end();
802 Instruction *User = cast<Instruction>(*UI);
804 // Figure out which BB this ext is used in.
805 BasicBlock *UserBB = User->getParent();
806 if (UserBB == DefBB) continue;
813 // Make sure non of the uses are PHI nodes.
814 for (Value::use_iterator UI = Src->use_begin(), E = Src->use_end();
816 Instruction *User = cast<Instruction>(*UI);
817 BasicBlock *UserBB = User->getParent();
818 if (UserBB == DefBB) continue;
819 // Be conservative. We don't want this xform to end up introducing
820 // reloads just before load / store instructions.
821 if (isa<PHINode>(User) || isa<LoadInst>(User) || isa<StoreInst>(User))
825 // InsertedTruncs - Only insert one trunc in each block once.
826 DenseMap<BasicBlock*, Instruction*> InsertedTruncs;
828 bool MadeChange = false;
829 for (Value::use_iterator UI = Src->use_begin(), E = Src->use_end();
831 Use &TheUse = UI.getUse();
832 Instruction *User = cast<Instruction>(*UI);
834 // Figure out which BB this ext is used in.
835 BasicBlock *UserBB = User->getParent();
836 if (UserBB == DefBB) continue;
838 // Both src and def are live in this block. Rewrite the use.
839 Instruction *&InsertedTrunc = InsertedTruncs[UserBB];
841 if (!InsertedTrunc) {
842 BasicBlock::iterator InsertPt = UserBB->getFirstNonPHI();
844 InsertedTrunc = new TruncInst(I, Src->getType(), "", InsertPt);
847 // Replace a use of the {s|z}ext source with a use of the result.
848 TheUse = InsertedTrunc;
856 // In this pass we look for GEP and cast instructions that are used
857 // across basic blocks and rewrite them to improve basic-block-at-a-time
859 bool CodeGenPrepare::OptimizeBlock(BasicBlock &BB) {
860 bool MadeChange = false;
862 // Split all critical edges where the dest block has a PHI.
863 TerminatorInst *BBTI = BB.getTerminator();
864 if (BBTI->getNumSuccessors() > 1 && !isa<IndirectBrInst>(BBTI)) {
865 for (unsigned i = 0, e = BBTI->getNumSuccessors(); i != e; ++i) {
866 BasicBlock *SuccBB = BBTI->getSuccessor(i);
867 if (isa<PHINode>(SuccBB->begin()) && isCriticalEdge(BBTI, i, true))
868 SplitEdgeNicely(BBTI, i, BackEdges, this);
872 // Keep track of non-local addresses that have been sunk into this block.
873 // This allows us to avoid inserting duplicate code for blocks with multiple
874 // load/stores of the same address.
875 DenseMap<Value*, Value*> SunkAddrs;
877 for (BasicBlock::iterator BBI = BB.begin(), E = BB.end(); BBI != E; ) {
878 Instruction *I = BBI++;
880 if (CastInst *CI = dyn_cast<CastInst>(I)) {
881 // If the source of the cast is a constant, then this should have
882 // already been constant folded. The only reason NOT to constant fold
883 // it is if something (e.g. LSR) was careful to place the constant
884 // evaluation in a block other than then one that uses it (e.g. to hoist
885 // the address of globals out of a loop). If this is the case, we don't
886 // want to forward-subst the cast.
887 if (isa<Constant>(CI->getOperand(0)))
892 Change = OptimizeNoopCopyExpression(CI, *TLI);
893 MadeChange |= Change;
896 if (!Change && (isa<ZExtInst>(I) || isa<SExtInst>(I))) {
897 MadeChange |= MoveExtToFormExtLoad(I);
898 MadeChange |= OptimizeExtUses(I);
900 } else if (CmpInst *CI = dyn_cast<CmpInst>(I)) {
901 MadeChange |= OptimizeCmpExpression(CI);
902 } else if (LoadInst *LI = dyn_cast<LoadInst>(I)) {
904 MadeChange |= OptimizeMemoryInst(I, I->getOperand(0), LI->getType(),
906 } else if (StoreInst *SI = dyn_cast<StoreInst>(I)) {
908 MadeChange |= OptimizeMemoryInst(I, SI->getOperand(1),
909 SI->getOperand(0)->getType(),
911 } else if (GetElementPtrInst *GEPI = dyn_cast<GetElementPtrInst>(I)) {
912 if (GEPI->hasAllZeroIndices()) {
913 /// The GEP operand must be a pointer, so must its result -> BitCast
914 Instruction *NC = new BitCastInst(GEPI->getOperand(0), GEPI->getType(),
915 GEPI->getName(), GEPI);
916 GEPI->replaceAllUsesWith(NC);
917 GEPI->eraseFromParent();
921 } else if (CallInst *CI = dyn_cast<CallInst>(I)) {
922 // If we found an inline asm expession, and if the target knows how to
923 // lower it to normal LLVM code, do so now.
924 if (TLI && isa<InlineAsm>(CI->getCalledValue())) {
925 if (TLI->ExpandInlineAsm(CI)) {
927 // Avoid processing instructions out of order, which could cause
928 // reuse before a value is defined.
931 // Sink address computing for memory operands into the block.
932 MadeChange |= OptimizeInlineAsmInst(I, &(*CI), SunkAddrs);