1 //===- MergeICmps.cpp - Optimize chains of integer comparisons ------------===// 2 // 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 4 // See https://llvm.org/LICENSE.txt for license information. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 6 // 7 //===----------------------------------------------------------------------===// 8 // 9 // This pass turns chains of integer comparisons into memcmp (the memcmp is 10 // later typically inlined as a chain of efficient hardware comparisons). This 11 // typically benefits c++ member or nonmember operator==(). 12 // 13 // The basic idea is to replace a longer chain of integer comparisons loaded 14 // from contiguous memory locations into a shorter chain of larger integer 15 // comparisons. Benefits are double: 16 // - There are less jumps, and therefore less opportunities for mispredictions 17 // and I-cache misses. 18 // - Code size is smaller, both because jumps are removed and because the 19 // encoding of a 2*n byte compare is smaller than that of two n-byte 20 // compares. 21 // 22 // Example: 23 // 24 // struct S { 25 // int a; 26 // char b; 27 // char c; 28 // uint16_t d; 29 // bool operator==(const S& o) const { 30 // return a == o.a && b == o.b && c == o.c && d == o.d; 31 // } 32 // }; 33 // 34 // Is optimized as : 35 // 36 // bool S::operator==(const S& o) const { 37 // return memcmp(this, &o, 8) == 0; 38 // } 39 // 40 // Which will later be expanded (ExpandMemCmp) as a single 8-bytes icmp. 41 // 42 //===----------------------------------------------------------------------===// 43 44 #include "llvm/Transforms/Scalar/MergeICmps.h" 45 #include "llvm/Analysis/DomTreeUpdater.h" 46 #include "llvm/Analysis/GlobalsModRef.h" 47 #include "llvm/Analysis/Loads.h" 48 #include "llvm/Analysis/TargetLibraryInfo.h" 49 #include "llvm/Analysis/TargetTransformInfo.h" 50 #include "llvm/IR/Dominators.h" 51 #include "llvm/IR/Function.h" 52 #include "llvm/IR/IRBuilder.h" 53 #include "llvm/InitializePasses.h" 54 #include "llvm/Pass.h" 55 #include "llvm/Transforms/Scalar.h" 56 #include "llvm/Transforms/Utils/BasicBlockUtils.h" 57 #include "llvm/Transforms/Utils/BuildLibCalls.h" 58 #include <algorithm> 59 #include <numeric> 60 #include <utility> 61 #include <vector> 62 63 using namespace llvm; 64 65 namespace { 66 67 #define DEBUG_TYPE "mergeicmps" 68 69 // A BCE atom "Binary Compare Expression Atom" represents an integer load 70 // that is a constant offset from a base value, e.g. `a` or `o.c` in the example 71 // at the top. 72 struct BCEAtom { 73 BCEAtom() = default; 74 BCEAtom(GetElementPtrInst *GEP, LoadInst *LoadI, int BaseId, APInt Offset) 75 : GEP(GEP), LoadI(LoadI), BaseId(BaseId), Offset(Offset) {} 76 77 BCEAtom(const BCEAtom &) = delete; 78 BCEAtom &operator=(const BCEAtom &) = delete; 79 80 BCEAtom(BCEAtom &&that) = default; 81 BCEAtom &operator=(BCEAtom &&that) { 82 if (this == &that) 83 return *this; 84 GEP = that.GEP; 85 LoadI = that.LoadI; 86 BaseId = that.BaseId; 87 Offset = std::move(that.Offset); 88 return *this; 89 } 90 91 // We want to order BCEAtoms by (Base, Offset). However we cannot use 92 // the pointer values for Base because these are non-deterministic. 93 // To make sure that the sort order is stable, we first assign to each atom 94 // base value an index based on its order of appearance in the chain of 95 // comparisons. We call this index `BaseOrdering`. For example, for: 96 // b[3] == c[2] && a[1] == d[1] && b[4] == c[3] 97 // | block 1 | | block 2 | | block 3 | 98 // b gets assigned index 0 and a index 1, because b appears as LHS in block 1, 99 // which is before block 2. 100 // We then sort by (BaseOrdering[LHS.Base()], LHS.Offset), which is stable. 101 bool operator<(const BCEAtom &O) const { 102 return BaseId != O.BaseId ? BaseId < O.BaseId : Offset.slt(O.Offset); 103 } 104 105 GetElementPtrInst *GEP = nullptr; 106 LoadInst *LoadI = nullptr; 107 unsigned BaseId = 0; 108 APInt Offset; 109 }; 110 111 // A class that assigns increasing ids to values in the order in which they are 112 // seen. See comment in `BCEAtom::operator<()``. 113 class BaseIdentifier { 114 public: 115 // Returns the id for value `Base`, after assigning one if `Base` has not been 116 // seen before. 117 int getBaseId(const Value *Base) { 118 assert(Base && "invalid base"); 119 const auto Insertion = BaseToIndex.try_emplace(Base, Order); 120 if (Insertion.second) 121 ++Order; 122 return Insertion.first->second; 123 } 124 125 private: 126 unsigned Order = 1; 127 DenseMap<const Value*, int> BaseToIndex; 128 }; 129 130 // If this value is a load from a constant offset w.r.t. a base address, and 131 // there are no other users of the load or address, returns the base address and 132 // the offset. 133 BCEAtom visitICmpLoadOperand(Value *const Val, BaseIdentifier &BaseId) { 134 auto *const LoadI = dyn_cast<LoadInst>(Val); 135 if (!LoadI) 136 return {}; 137 LLVM_DEBUG(dbgs() << "load\n"); 138 if (LoadI->isUsedOutsideOfBlock(LoadI->getParent())) { 139 LLVM_DEBUG(dbgs() << "used outside of block\n"); 140 return {}; 141 } 142 // Do not optimize atomic loads to non-atomic memcmp 143 if (!LoadI->isSimple()) { 144 LLVM_DEBUG(dbgs() << "volatile or atomic\n"); 145 return {}; 146 } 147 Value *const Addr = LoadI->getOperand(0); 148 if (Addr->getType()->getPointerAddressSpace() != 0) { 149 LLVM_DEBUG(dbgs() << "from non-zero AddressSpace\n"); 150 return {}; 151 } 152 auto *const GEP = dyn_cast<GetElementPtrInst>(Addr); 153 if (!GEP) 154 return {}; 155 LLVM_DEBUG(dbgs() << "GEP\n"); 156 if (GEP->isUsedOutsideOfBlock(LoadI->getParent())) { 157 LLVM_DEBUG(dbgs() << "used outside of block\n"); 158 return {}; 159 } 160 const auto &DL = GEP->getModule()->getDataLayout(); 161 if (!isDereferenceablePointer(GEP, LoadI->getType(), DL)) { 162 LLVM_DEBUG(dbgs() << "not dereferenceable\n"); 163 // We need to make sure that we can do comparison in any order, so we 164 // require memory to be unconditionnally dereferencable. 165 return {}; 166 } 167 APInt Offset = APInt(DL.getPointerTypeSizeInBits(GEP->getType()), 0); 168 if (!GEP->accumulateConstantOffset(DL, Offset)) 169 return {}; 170 return BCEAtom(GEP, LoadI, BaseId.getBaseId(GEP->getPointerOperand()), 171 Offset); 172 } 173 174 // A comparison between two BCE atoms, e.g. `a == o.a` in the example at the 175 // top. 176 // Note: the terminology is misleading: the comparison is symmetric, so there 177 // is no real {l/r}hs. What we want though is to have the same base on the 178 // left (resp. right), so that we can detect consecutive loads. To ensure this 179 // we put the smallest atom on the left. 180 struct BCECmp { 181 BCEAtom Lhs; 182 BCEAtom Rhs; 183 int SizeBits; 184 const ICmpInst *CmpI; 185 186 BCECmp(BCEAtom L, BCEAtom R, int SizeBits, const ICmpInst *CmpI) 187 : Lhs(std::move(L)), Rhs(std::move(R)), SizeBits(SizeBits), CmpI(CmpI) { 188 if (Rhs < Lhs) std::swap(Rhs, Lhs); 189 } 190 }; 191 192 // A basic block with a comparison between two BCE atoms. 193 // The block might do extra work besides the atom comparison, in which case 194 // doesOtherWork() returns true. Under some conditions, the block can be 195 // split into the atom comparison part and the "other work" part 196 // (see canSplit()). 197 class BCECmpBlock { 198 public: 199 typedef SmallDenseSet<const Instruction *, 8> InstructionSet; 200 201 BCECmpBlock(BCECmp Cmp, BasicBlock *BB, InstructionSet BlockInsts) 202 : BB(BB), BlockInsts(std::move(BlockInsts)), Cmp(std::move(Cmp)) {} 203 204 const BCEAtom &Lhs() const { return Cmp.Lhs; } 205 const BCEAtom &Rhs() const { return Cmp.Rhs; } 206 int SizeBits() const { return Cmp.SizeBits; } 207 208 // Returns true if the block does other works besides comparison. 209 bool doesOtherWork() const; 210 211 // Returns true if the non-BCE-cmp instructions can be separated from BCE-cmp 212 // instructions in the block. 213 bool canSplit(AliasAnalysis &AA) const; 214 215 // Return true if this all the relevant instructions in the BCE-cmp-block can 216 // be sunk below this instruction. By doing this, we know we can separate the 217 // BCE-cmp-block instructions from the non-BCE-cmp-block instructions in the 218 // block. 219 bool canSinkBCECmpInst(const Instruction *, AliasAnalysis &AA) const; 220 221 // We can separate the BCE-cmp-block instructions and the non-BCE-cmp-block 222 // instructions. Split the old block and move all non-BCE-cmp-insts into the 223 // new parent block. 224 void split(BasicBlock *NewParent, AliasAnalysis &AA) const; 225 226 // The basic block where this comparison happens. 227 BasicBlock *BB; 228 // Instructions relating to the BCECmp and branch. 229 InstructionSet BlockInsts; 230 // The block requires splitting. 231 bool RequireSplit = false; 232 // Original order of this block in the chain. 233 unsigned OrigOrder = 0; 234 235 private: 236 BCECmp Cmp; 237 }; 238 239 bool BCECmpBlock::canSinkBCECmpInst(const Instruction *Inst, 240 AliasAnalysis &AA) const { 241 // If this instruction may clobber the loads and is in middle of the BCE cmp 242 // block instructions, then bail for now. 243 if (Inst->mayWriteToMemory()) { 244 auto MayClobber = [&](LoadInst *LI) { 245 // If a potentially clobbering instruction comes before the load, 246 // we can still safely sink the load. 247 return !Inst->comesBefore(LI) && 248 isModSet(AA.getModRefInfo(Inst, MemoryLocation::get(LI))); 249 }; 250 if (MayClobber(Cmp.Lhs.LoadI) || MayClobber(Cmp.Rhs.LoadI)) 251 return false; 252 } 253 // Make sure this instruction does not use any of the BCE cmp block 254 // instructions as operand. 255 return llvm::none_of(Inst->operands(), [&](const Value *Op) { 256 const Instruction *OpI = dyn_cast<Instruction>(Op); 257 return OpI && BlockInsts.contains(OpI); 258 }); 259 } 260 261 void BCECmpBlock::split(BasicBlock *NewParent, AliasAnalysis &AA) const { 262 llvm::SmallVector<Instruction *, 4> OtherInsts; 263 for (Instruction &Inst : *BB) { 264 if (BlockInsts.count(&Inst)) 265 continue; 266 assert(canSinkBCECmpInst(&Inst, AA) && "Split unsplittable block"); 267 // This is a non-BCE-cmp-block instruction. And it can be separated 268 // from the BCE-cmp-block instruction. 269 OtherInsts.push_back(&Inst); 270 } 271 272 // Do the actual spliting. 273 for (Instruction *Inst : reverse(OtherInsts)) { 274 Inst->moveBefore(&*NewParent->begin()); 275 } 276 } 277 278 bool BCECmpBlock::canSplit(AliasAnalysis &AA) const { 279 for (Instruction &Inst : *BB) { 280 if (!BlockInsts.count(&Inst)) { 281 if (!canSinkBCECmpInst(&Inst, AA)) 282 return false; 283 } 284 } 285 return true; 286 } 287 288 bool BCECmpBlock::doesOtherWork() const { 289 // TODO(courbet): Can we allow some other things ? This is very conservative. 290 // We might be able to get away with anything does not have any side 291 // effects outside of the basic block. 292 // Note: The GEPs and/or loads are not necessarily in the same block. 293 for (const Instruction &Inst : *BB) { 294 if (!BlockInsts.count(&Inst)) 295 return true; 296 } 297 return false; 298 } 299 300 // Visit the given comparison. If this is a comparison between two valid 301 // BCE atoms, returns the comparison. 302 Optional<BCECmp> visitICmp(const ICmpInst *const CmpI, 303 const ICmpInst::Predicate ExpectedPredicate, 304 BaseIdentifier &BaseId) { 305 // The comparison can only be used once: 306 // - For intermediate blocks, as a branch condition. 307 // - For the final block, as an incoming value for the Phi. 308 // If there are any other uses of the comparison, we cannot merge it with 309 // other comparisons as we would create an orphan use of the value. 310 if (!CmpI->hasOneUse()) { 311 LLVM_DEBUG(dbgs() << "cmp has several uses\n"); 312 return None; 313 } 314 if (CmpI->getPredicate() != ExpectedPredicate) 315 return None; 316 LLVM_DEBUG(dbgs() << "cmp " 317 << (ExpectedPredicate == ICmpInst::ICMP_EQ ? "eq" : "ne") 318 << "\n"); 319 auto Lhs = visitICmpLoadOperand(CmpI->getOperand(0), BaseId); 320 if (!Lhs.BaseId) 321 return None; 322 auto Rhs = visitICmpLoadOperand(CmpI->getOperand(1), BaseId); 323 if (!Rhs.BaseId) 324 return None; 325 const auto &DL = CmpI->getModule()->getDataLayout(); 326 return BCECmp(std::move(Lhs), std::move(Rhs), 327 DL.getTypeSizeInBits(CmpI->getOperand(0)->getType()), CmpI); 328 } 329 330 // Visit the given comparison block. If this is a comparison between two valid 331 // BCE atoms, returns the comparison. 332 Optional<BCECmpBlock> visitCmpBlock(Value *const Val, BasicBlock *const Block, 333 const BasicBlock *const PhiBlock, 334 BaseIdentifier &BaseId) { 335 if (Block->empty()) return None; 336 auto *const BranchI = dyn_cast<BranchInst>(Block->getTerminator()); 337 if (!BranchI) return None; 338 LLVM_DEBUG(dbgs() << "branch\n"); 339 Value *Cond; 340 ICmpInst::Predicate ExpectedPredicate; 341 if (BranchI->isUnconditional()) { 342 // In this case, we expect an incoming value which is the result of the 343 // comparison. This is the last link in the chain of comparisons (note 344 // that this does not mean that this is the last incoming value, blocks 345 // can be reordered). 346 Cond = Val; 347 ExpectedPredicate = ICmpInst::ICMP_EQ; 348 } else { 349 // In this case, we expect a constant incoming value (the comparison is 350 // chained). 351 const auto *const Const = cast<ConstantInt>(Val); 352 LLVM_DEBUG(dbgs() << "const\n"); 353 if (!Const->isZero()) return None; 354 LLVM_DEBUG(dbgs() << "false\n"); 355 assert(BranchI->getNumSuccessors() == 2 && "expecting a cond branch"); 356 BasicBlock *const FalseBlock = BranchI->getSuccessor(1); 357 Cond = BranchI->getCondition(); 358 ExpectedPredicate = 359 FalseBlock == PhiBlock ? ICmpInst::ICMP_EQ : ICmpInst::ICMP_NE; 360 } 361 362 auto *CmpI = dyn_cast<ICmpInst>(Cond); 363 if (!CmpI) return None; 364 LLVM_DEBUG(dbgs() << "icmp\n"); 365 366 Optional<BCECmp> Result = visitICmp(CmpI, ExpectedPredicate, BaseId); 367 if (!Result) 368 return None; 369 370 BCECmpBlock::InstructionSet BlockInsts( 371 {Result->Lhs.GEP, Result->Rhs.GEP, Result->Lhs.LoadI, Result->Rhs.LoadI, 372 Result->CmpI, BranchI}); 373 return BCECmpBlock(std::move(*Result), Block, BlockInsts); 374 } 375 376 static inline void enqueueBlock(std::vector<BCECmpBlock> &Comparisons, 377 BCECmpBlock &&Comparison) { 378 LLVM_DEBUG(dbgs() << "Block '" << Comparison.BB->getName() 379 << "': Found cmp of " << Comparison.SizeBits() 380 << " bits between " << Comparison.Lhs().BaseId << " + " 381 << Comparison.Lhs().Offset << " and " 382 << Comparison.Rhs().BaseId << " + " 383 << Comparison.Rhs().Offset << "\n"); 384 LLVM_DEBUG(dbgs() << "\n"); 385 Comparison.OrigOrder = Comparisons.size(); 386 Comparisons.push_back(std::move(Comparison)); 387 } 388 389 // A chain of comparisons. 390 class BCECmpChain { 391 public: 392 using ContiguousBlocks = std::vector<BCECmpBlock>; 393 394 BCECmpChain(const std::vector<BasicBlock *> &Blocks, PHINode &Phi, 395 AliasAnalysis &AA); 396 397 bool simplify(const TargetLibraryInfo &TLI, AliasAnalysis &AA, 398 DomTreeUpdater &DTU); 399 400 bool atLeastOneMerged() const { 401 return any_of(MergedBlocks_, 402 [](const auto &Blocks) { return Blocks.size() > 1; }); 403 } 404 405 private: 406 PHINode &Phi_; 407 // The list of all blocks in the chain, grouped by contiguity. 408 std::vector<ContiguousBlocks> MergedBlocks_; 409 // The original entry block (before sorting); 410 BasicBlock *EntryBlock_; 411 }; 412 413 static bool areContiguous(const BCECmpBlock &First, const BCECmpBlock &Second) { 414 return First.Lhs().BaseId == Second.Lhs().BaseId && 415 First.Rhs().BaseId == Second.Rhs().BaseId && 416 First.Lhs().Offset + First.SizeBits() / 8 == Second.Lhs().Offset && 417 First.Rhs().Offset + First.SizeBits() / 8 == Second.Rhs().Offset; 418 } 419 420 static unsigned getMinOrigOrder(const BCECmpChain::ContiguousBlocks &Blocks) { 421 unsigned MinOrigOrder = std::numeric_limits<unsigned>::max(); 422 for (const BCECmpBlock &Block : Blocks) 423 MinOrigOrder = std::min(MinOrigOrder, Block.OrigOrder); 424 return MinOrigOrder; 425 } 426 427 /// Given a chain of comparison blocks, groups the blocks into contiguous 428 /// ranges that can be merged together into a single comparison. 429 static std::vector<BCECmpChain::ContiguousBlocks> 430 mergeBlocks(std::vector<BCECmpBlock> &&Blocks) { 431 std::vector<BCECmpChain::ContiguousBlocks> MergedBlocks; 432 433 // Sort to detect continuous offsets. 434 llvm::sort(Blocks, 435 [](const BCECmpBlock &LhsBlock, const BCECmpBlock &RhsBlock) { 436 return std::tie(LhsBlock.Lhs(), LhsBlock.Rhs()) < 437 std::tie(RhsBlock.Lhs(), RhsBlock.Rhs()); 438 }); 439 440 BCECmpChain::ContiguousBlocks *LastMergedBlock = nullptr; 441 for (BCECmpBlock &Block : Blocks) { 442 if (!LastMergedBlock || !areContiguous(LastMergedBlock->back(), Block)) { 443 MergedBlocks.emplace_back(); 444 LastMergedBlock = &MergedBlocks.back(); 445 } else { 446 LLVM_DEBUG(dbgs() << "Merging block " << Block.BB->getName() << " into " 447 << LastMergedBlock->back().BB->getName() << "\n"); 448 } 449 LastMergedBlock->push_back(std::move(Block)); 450 } 451 452 // While we allow reordering for merging, do not reorder unmerged comparisons. 453 // Doing so may introduce branch on poison. 454 llvm::sort(MergedBlocks, [](const BCECmpChain::ContiguousBlocks &LhsBlocks, 455 const BCECmpChain::ContiguousBlocks &RhsBlocks) { 456 return getMinOrigOrder(LhsBlocks) < getMinOrigOrder(RhsBlocks); 457 }); 458 459 return MergedBlocks; 460 } 461 462 BCECmpChain::BCECmpChain(const std::vector<BasicBlock *> &Blocks, PHINode &Phi, 463 AliasAnalysis &AA) 464 : Phi_(Phi) { 465 assert(!Blocks.empty() && "a chain should have at least one block"); 466 // Now look inside blocks to check for BCE comparisons. 467 std::vector<BCECmpBlock> Comparisons; 468 BaseIdentifier BaseId; 469 for (BasicBlock *const Block : Blocks) { 470 assert(Block && "invalid block"); 471 Optional<BCECmpBlock> Comparison = visitCmpBlock( 472 Phi.getIncomingValueForBlock(Block), Block, Phi.getParent(), BaseId); 473 if (!Comparison) { 474 LLVM_DEBUG(dbgs() << "chain with invalid BCECmpBlock, no merge.\n"); 475 return; 476 } 477 if (Comparison->doesOtherWork()) { 478 LLVM_DEBUG(dbgs() << "block '" << Comparison->BB->getName() 479 << "' does extra work besides compare\n"); 480 if (Comparisons.empty()) { 481 // This is the initial block in the chain, in case this block does other 482 // work, we can try to split the block and move the irrelevant 483 // instructions to the predecessor. 484 // 485 // If this is not the initial block in the chain, splitting it wont 486 // work. 487 // 488 // As once split, there will still be instructions before the BCE cmp 489 // instructions that do other work in program order, i.e. within the 490 // chain before sorting. Unless we can abort the chain at this point 491 // and start anew. 492 // 493 // NOTE: we only handle blocks a with single predecessor for now. 494 if (Comparison->canSplit(AA)) { 495 LLVM_DEBUG(dbgs() 496 << "Split initial block '" << Comparison->BB->getName() 497 << "' that does extra work besides compare\n"); 498 Comparison->RequireSplit = true; 499 enqueueBlock(Comparisons, std::move(*Comparison)); 500 } else { 501 LLVM_DEBUG(dbgs() 502 << "ignoring initial block '" << Comparison->BB->getName() 503 << "' that does extra work besides compare\n"); 504 } 505 continue; 506 } 507 // TODO(courbet): Right now we abort the whole chain. We could be 508 // merging only the blocks that don't do other work and resume the 509 // chain from there. For example: 510 // if (a[0] == b[0]) { // bb1 511 // if (a[1] == b[1]) { // bb2 512 // some_value = 3; //bb3 513 // if (a[2] == b[2]) { //bb3 514 // do a ton of stuff //bb4 515 // } 516 // } 517 // } 518 // 519 // This is: 520 // 521 // bb1 --eq--> bb2 --eq--> bb3* -eq--> bb4 --+ 522 // \ \ \ \ 523 // ne ne ne \ 524 // \ \ \ v 525 // +------------+-----------+----------> bb_phi 526 // 527 // We can only merge the first two comparisons, because bb3* does 528 // "other work" (setting some_value to 3). 529 // We could still merge bb1 and bb2 though. 530 return; 531 } 532 enqueueBlock(Comparisons, std::move(*Comparison)); 533 } 534 535 // It is possible we have no suitable comparison to merge. 536 if (Comparisons.empty()) { 537 LLVM_DEBUG(dbgs() << "chain with no BCE basic blocks, no merge\n"); 538 return; 539 } 540 EntryBlock_ = Comparisons[0].BB; 541 MergedBlocks_ = mergeBlocks(std::move(Comparisons)); 542 } 543 544 namespace { 545 546 // A class to compute the name of a set of merged basic blocks. 547 // This is optimized for the common case of no block names. 548 class MergedBlockName { 549 // Storage for the uncommon case of several named blocks. 550 SmallString<16> Scratch; 551 552 public: 553 explicit MergedBlockName(ArrayRef<BCECmpBlock> Comparisons) 554 : Name(makeName(Comparisons)) {} 555 const StringRef Name; 556 557 private: 558 StringRef makeName(ArrayRef<BCECmpBlock> Comparisons) { 559 assert(!Comparisons.empty() && "no basic block"); 560 // Fast path: only one block, or no names at all. 561 if (Comparisons.size() == 1) 562 return Comparisons[0].BB->getName(); 563 const int size = std::accumulate(Comparisons.begin(), Comparisons.end(), 0, 564 [](int i, const BCECmpBlock &Cmp) { 565 return i + Cmp.BB->getName().size(); 566 }); 567 if (size == 0) 568 return StringRef("", 0); 569 570 // Slow path: at least two blocks, at least one block with a name. 571 Scratch.clear(); 572 // We'll have `size` bytes for name and `Comparisons.size() - 1` bytes for 573 // separators. 574 Scratch.reserve(size + Comparisons.size() - 1); 575 const auto append = [this](StringRef str) { 576 Scratch.append(str.begin(), str.end()); 577 }; 578 append(Comparisons[0].BB->getName()); 579 for (int I = 1, E = Comparisons.size(); I < E; ++I) { 580 const BasicBlock *const BB = Comparisons[I].BB; 581 if (!BB->getName().empty()) { 582 append("+"); 583 append(BB->getName()); 584 } 585 } 586 return Scratch.str(); 587 } 588 }; 589 } // namespace 590 591 // Merges the given contiguous comparison blocks into one memcmp block. 592 static BasicBlock *mergeComparisons(ArrayRef<BCECmpBlock> Comparisons, 593 BasicBlock *const InsertBefore, 594 BasicBlock *const NextCmpBlock, 595 PHINode &Phi, const TargetLibraryInfo &TLI, 596 AliasAnalysis &AA, DomTreeUpdater &DTU) { 597 assert(!Comparisons.empty() && "merging zero comparisons"); 598 LLVMContext &Context = NextCmpBlock->getContext(); 599 const BCECmpBlock &FirstCmp = Comparisons[0]; 600 601 // Create a new cmp block before next cmp block. 602 BasicBlock *const BB = 603 BasicBlock::Create(Context, MergedBlockName(Comparisons).Name, 604 NextCmpBlock->getParent(), InsertBefore); 605 IRBuilder<> Builder(BB); 606 // Add the GEPs from the first BCECmpBlock. 607 Value *const Lhs = Builder.Insert(FirstCmp.Lhs().GEP->clone()); 608 Value *const Rhs = Builder.Insert(FirstCmp.Rhs().GEP->clone()); 609 610 Value *IsEqual = nullptr; 611 LLVM_DEBUG(dbgs() << "Merging " << Comparisons.size() << " comparisons -> " 612 << BB->getName() << "\n"); 613 614 // If there is one block that requires splitting, we do it now, i.e. 615 // just before we know we will collapse the chain. The instructions 616 // can be executed before any of the instructions in the chain. 617 const auto ToSplit = llvm::find_if( 618 Comparisons, [](const BCECmpBlock &B) { return B.RequireSplit; }); 619 if (ToSplit != Comparisons.end()) { 620 LLVM_DEBUG(dbgs() << "Splitting non_BCE work to header\n"); 621 ToSplit->split(BB, AA); 622 } 623 624 if (Comparisons.size() == 1) { 625 LLVM_DEBUG(dbgs() << "Only one comparison, updating branches\n"); 626 Value *const LhsLoad = 627 Builder.CreateLoad(FirstCmp.Lhs().LoadI->getType(), Lhs); 628 Value *const RhsLoad = 629 Builder.CreateLoad(FirstCmp.Rhs().LoadI->getType(), Rhs); 630 // There are no blocks to merge, just do the comparison. 631 IsEqual = Builder.CreateICmpEQ(LhsLoad, RhsLoad); 632 } else { 633 const unsigned TotalSizeBits = std::accumulate( 634 Comparisons.begin(), Comparisons.end(), 0u, 635 [](int Size, const BCECmpBlock &C) { return Size + C.SizeBits(); }); 636 637 // Create memcmp() == 0. 638 const auto &DL = Phi.getModule()->getDataLayout(); 639 Value *const MemCmpCall = emitMemCmp( 640 Lhs, Rhs, 641 ConstantInt::get(DL.getIntPtrType(Context), TotalSizeBits / 8), Builder, 642 DL, &TLI); 643 IsEqual = Builder.CreateICmpEQ( 644 MemCmpCall, ConstantInt::get(Type::getInt32Ty(Context), 0)); 645 } 646 647 BasicBlock *const PhiBB = Phi.getParent(); 648 // Add a branch to the next basic block in the chain. 649 if (NextCmpBlock == PhiBB) { 650 // Continue to phi, passing it the comparison result. 651 Builder.CreateBr(PhiBB); 652 Phi.addIncoming(IsEqual, BB); 653 DTU.applyUpdates({{DominatorTree::Insert, BB, PhiBB}}); 654 } else { 655 // Continue to next block if equal, exit to phi else. 656 Builder.CreateCondBr(IsEqual, NextCmpBlock, PhiBB); 657 Phi.addIncoming(ConstantInt::getFalse(Context), BB); 658 DTU.applyUpdates({{DominatorTree::Insert, BB, NextCmpBlock}, 659 {DominatorTree::Insert, BB, PhiBB}}); 660 } 661 return BB; 662 } 663 664 bool BCECmpChain::simplify(const TargetLibraryInfo &TLI, AliasAnalysis &AA, 665 DomTreeUpdater &DTU) { 666 assert(atLeastOneMerged() && "simplifying trivial BCECmpChain"); 667 LLVM_DEBUG(dbgs() << "Simplifying comparison chain starting at block " 668 << EntryBlock_->getName() << "\n"); 669 670 // Effectively merge blocks. We go in the reverse direction from the phi block 671 // so that the next block is always available to branch to. 672 BasicBlock *InsertBefore = EntryBlock_; 673 BasicBlock *NextCmpBlock = Phi_.getParent(); 674 for (const auto &Blocks : reverse(MergedBlocks_)) { 675 InsertBefore = NextCmpBlock = mergeComparisons( 676 Blocks, InsertBefore, NextCmpBlock, Phi_, TLI, AA, DTU); 677 } 678 679 // Replace the original cmp chain with the new cmp chain by pointing all 680 // predecessors of EntryBlock_ to NextCmpBlock instead. This makes all cmp 681 // blocks in the old chain unreachable. 682 while (!pred_empty(EntryBlock_)) { 683 BasicBlock* const Pred = *pred_begin(EntryBlock_); 684 LLVM_DEBUG(dbgs() << "Updating jump into old chain from " << Pred->getName() 685 << "\n"); 686 Pred->getTerminator()->replaceUsesOfWith(EntryBlock_, NextCmpBlock); 687 DTU.applyUpdates({{DominatorTree::Delete, Pred, EntryBlock_}, 688 {DominatorTree::Insert, Pred, NextCmpBlock}}); 689 } 690 691 // If the old cmp chain was the function entry, we need to update the function 692 // entry. 693 const bool ChainEntryIsFnEntry = EntryBlock_->isEntryBlock(); 694 if (ChainEntryIsFnEntry && DTU.hasDomTree()) { 695 LLVM_DEBUG(dbgs() << "Changing function entry from " 696 << EntryBlock_->getName() << " to " 697 << NextCmpBlock->getName() << "\n"); 698 DTU.getDomTree().setNewRoot(NextCmpBlock); 699 DTU.applyUpdates({{DominatorTree::Delete, NextCmpBlock, EntryBlock_}}); 700 } 701 EntryBlock_ = nullptr; 702 703 // Delete merged blocks. This also removes incoming values in phi. 704 SmallVector<BasicBlock *, 16> DeadBlocks; 705 for (const auto &Blocks : MergedBlocks_) { 706 for (const BCECmpBlock &Block : Blocks) { 707 LLVM_DEBUG(dbgs() << "Deleting merged block " << Block.BB->getName() 708 << "\n"); 709 DeadBlocks.push_back(Block.BB); 710 } 711 } 712 DeleteDeadBlocks(DeadBlocks, &DTU); 713 714 MergedBlocks_.clear(); 715 return true; 716 } 717 718 std::vector<BasicBlock *> getOrderedBlocks(PHINode &Phi, 719 BasicBlock *const LastBlock, 720 int NumBlocks) { 721 // Walk up from the last block to find other blocks. 722 std::vector<BasicBlock *> Blocks(NumBlocks); 723 assert(LastBlock && "invalid last block"); 724 BasicBlock *CurBlock = LastBlock; 725 for (int BlockIndex = NumBlocks - 1; BlockIndex > 0; --BlockIndex) { 726 if (CurBlock->hasAddressTaken()) { 727 // Somebody is jumping to the block through an address, all bets are 728 // off. 729 LLVM_DEBUG(dbgs() << "skip: block " << BlockIndex 730 << " has its address taken\n"); 731 return {}; 732 } 733 Blocks[BlockIndex] = CurBlock; 734 auto *SinglePredecessor = CurBlock->getSinglePredecessor(); 735 if (!SinglePredecessor) { 736 // The block has two or more predecessors. 737 LLVM_DEBUG(dbgs() << "skip: block " << BlockIndex 738 << " has two or more predecessors\n"); 739 return {}; 740 } 741 if (Phi.getBasicBlockIndex(SinglePredecessor) < 0) { 742 // The block does not link back to the phi. 743 LLVM_DEBUG(dbgs() << "skip: block " << BlockIndex 744 << " does not link back to the phi\n"); 745 return {}; 746 } 747 CurBlock = SinglePredecessor; 748 } 749 Blocks[0] = CurBlock; 750 return Blocks; 751 } 752 753 bool processPhi(PHINode &Phi, const TargetLibraryInfo &TLI, AliasAnalysis &AA, 754 DomTreeUpdater &DTU) { 755 LLVM_DEBUG(dbgs() << "processPhi()\n"); 756 if (Phi.getNumIncomingValues() <= 1) { 757 LLVM_DEBUG(dbgs() << "skip: only one incoming value in phi\n"); 758 return false; 759 } 760 // We are looking for something that has the following structure: 761 // bb1 --eq--> bb2 --eq--> bb3 --eq--> bb4 --+ 762 // \ \ \ \ 763 // ne ne ne \ 764 // \ \ \ v 765 // +------------+-----------+----------> bb_phi 766 // 767 // - The last basic block (bb4 here) must branch unconditionally to bb_phi. 768 // It's the only block that contributes a non-constant value to the Phi. 769 // - All other blocks (b1, b2, b3) must have exactly two successors, one of 770 // them being the phi block. 771 // - All intermediate blocks (bb2, bb3) must have only one predecessor. 772 // - Blocks cannot do other work besides the comparison, see doesOtherWork() 773 774 // The blocks are not necessarily ordered in the phi, so we start from the 775 // last block and reconstruct the order. 776 BasicBlock *LastBlock = nullptr; 777 for (unsigned I = 0; I < Phi.getNumIncomingValues(); ++I) { 778 if (isa<ConstantInt>(Phi.getIncomingValue(I))) continue; 779 if (LastBlock) { 780 // There are several non-constant values. 781 LLVM_DEBUG(dbgs() << "skip: several non-constant values\n"); 782 return false; 783 } 784 if (!isa<ICmpInst>(Phi.getIncomingValue(I)) || 785 cast<ICmpInst>(Phi.getIncomingValue(I))->getParent() != 786 Phi.getIncomingBlock(I)) { 787 // Non-constant incoming value is not from a cmp instruction or not 788 // produced by the last block. We could end up processing the value 789 // producing block more than once. 790 // 791 // This is an uncommon case, so we bail. 792 LLVM_DEBUG( 793 dbgs() 794 << "skip: non-constant value not from cmp or not from last block.\n"); 795 return false; 796 } 797 LastBlock = Phi.getIncomingBlock(I); 798 } 799 if (!LastBlock) { 800 // There is no non-constant block. 801 LLVM_DEBUG(dbgs() << "skip: no non-constant block\n"); 802 return false; 803 } 804 if (LastBlock->getSingleSuccessor() != Phi.getParent()) { 805 LLVM_DEBUG(dbgs() << "skip: last block non-phi successor\n"); 806 return false; 807 } 808 809 const auto Blocks = 810 getOrderedBlocks(Phi, LastBlock, Phi.getNumIncomingValues()); 811 if (Blocks.empty()) return false; 812 BCECmpChain CmpChain(Blocks, Phi, AA); 813 814 if (!CmpChain.atLeastOneMerged()) { 815 LLVM_DEBUG(dbgs() << "skip: nothing merged\n"); 816 return false; 817 } 818 819 return CmpChain.simplify(TLI, AA, DTU); 820 } 821 822 static bool runImpl(Function &F, const TargetLibraryInfo &TLI, 823 const TargetTransformInfo &TTI, AliasAnalysis &AA, 824 DominatorTree *DT) { 825 LLVM_DEBUG(dbgs() << "MergeICmpsLegacyPass: " << F.getName() << "\n"); 826 827 // We only try merging comparisons if the target wants to expand memcmp later. 828 // The rationale is to avoid turning small chains into memcmp calls. 829 if (!TTI.enableMemCmpExpansion(F.hasOptSize(), true)) 830 return false; 831 832 // If we don't have memcmp avaiable we can't emit calls to it. 833 if (!TLI.has(LibFunc_memcmp)) 834 return false; 835 836 DomTreeUpdater DTU(DT, /*PostDominatorTree*/ nullptr, 837 DomTreeUpdater::UpdateStrategy::Eager); 838 839 bool MadeChange = false; 840 841 for (BasicBlock &BB : llvm::drop_begin(F)) { 842 // A Phi operation is always first in a basic block. 843 if (auto *const Phi = dyn_cast<PHINode>(&*BB.begin())) 844 MadeChange |= processPhi(*Phi, TLI, AA, DTU); 845 } 846 847 return MadeChange; 848 } 849 850 class MergeICmpsLegacyPass : public FunctionPass { 851 public: 852 static char ID; 853 854 MergeICmpsLegacyPass() : FunctionPass(ID) { 855 initializeMergeICmpsLegacyPassPass(*PassRegistry::getPassRegistry()); 856 } 857 858 bool runOnFunction(Function &F) override { 859 if (skipFunction(F)) return false; 860 const auto &TLI = getAnalysis<TargetLibraryInfoWrapperPass>().getTLI(F); 861 const auto &TTI = getAnalysis<TargetTransformInfoWrapperPass>().getTTI(F); 862 // MergeICmps does not need the DominatorTree, but we update it if it's 863 // already available. 864 auto *DTWP = getAnalysisIfAvailable<DominatorTreeWrapperPass>(); 865 auto &AA = getAnalysis<AAResultsWrapperPass>().getAAResults(); 866 return runImpl(F, TLI, TTI, AA, DTWP ? &DTWP->getDomTree() : nullptr); 867 } 868 869 private: 870 void getAnalysisUsage(AnalysisUsage &AU) const override { 871 AU.addRequired<TargetLibraryInfoWrapperPass>(); 872 AU.addRequired<TargetTransformInfoWrapperPass>(); 873 AU.addRequired<AAResultsWrapperPass>(); 874 AU.addPreserved<GlobalsAAWrapperPass>(); 875 AU.addPreserved<DominatorTreeWrapperPass>(); 876 } 877 }; 878 879 } // namespace 880 881 char MergeICmpsLegacyPass::ID = 0; 882 INITIALIZE_PASS_BEGIN(MergeICmpsLegacyPass, "mergeicmps", 883 "Merge contiguous icmps into a memcmp", false, false) 884 INITIALIZE_PASS_DEPENDENCY(TargetLibraryInfoWrapperPass) 885 INITIALIZE_PASS_DEPENDENCY(TargetTransformInfoWrapperPass) 886 INITIALIZE_PASS_DEPENDENCY(AAResultsWrapperPass) 887 INITIALIZE_PASS_END(MergeICmpsLegacyPass, "mergeicmps", 888 "Merge contiguous icmps into a memcmp", false, false) 889 890 Pass *llvm::createMergeICmpsLegacyPass() { return new MergeICmpsLegacyPass(); } 891 892 PreservedAnalyses MergeICmpsPass::run(Function &F, 893 FunctionAnalysisManager &AM) { 894 auto &TLI = AM.getResult<TargetLibraryAnalysis>(F); 895 auto &TTI = AM.getResult<TargetIRAnalysis>(F); 896 auto &AA = AM.getResult<AAManager>(F); 897 auto *DT = AM.getCachedResult<DominatorTreeAnalysis>(F); 898 const bool MadeChanges = runImpl(F, TLI, TTI, AA, DT); 899 if (!MadeChanges) 900 return PreservedAnalyses::all(); 901 PreservedAnalyses PA; 902 PA.preserve<DominatorTreeAnalysis>(); 903 return PA; 904 } 905