Index: include/llvm/Transforms/Utils/UnrollLoop.h =================================================================== --- include/llvm/Transforms/Utils/UnrollLoop.h +++ include/llvm/Transforms/Utils/UnrollLoop.h @@ -119,9 +119,6 @@ TargetTransformInfo::UnrollingPreferences &UP, bool &UseUpperBound); -BasicBlock *foldBlockIntoPredecessor(BasicBlock *BB, LoopInfo *LI, - ScalarEvolution *SE, DominatorTree *DT); - void remapInstruction(Instruction *I, ValueToValueMapTy &VMap); void simplifyLoopAfterUnroll(Loop *L, bool SimplifyIVs, LoopInfo *LI, Index: lib/Transforms/Utils/BasicBlockUtils.cpp =================================================================== --- lib/Transforms/Utils/BasicBlockUtils.cpp +++ lib/Transforms/Utils/BasicBlockUtils.cpp @@ -38,6 +38,8 @@ #include "llvm/IR/Value.h" #include "llvm/IR/ValueHandle.h" #include "llvm/Support/Casting.h" +#include "llvm/Support/Debug.h" +#include "llvm/Support/raw_ostream.h" #include "llvm/Transforms/Utils/Local.h" #include #include @@ -47,6 +49,8 @@ using namespace llvm; +#define DEBUG_TYPE "basicblock-utils" + void llvm::DetatchDeadBlocks( ArrayRef BBs, SmallVectorImpl *Updates, @@ -190,6 +194,9 @@ if (IncValue == &PN) return false; + LLVM_DEBUG(dbgs() << "Merging: " << BB->getName() << " into " + << PredBB->getName() << "\n"); + // Begin by getting rid of unneeded PHIs. SmallVector, 4> IncomingValues; if (isa(BB->front())) { Index: lib/Transforms/Utils/LoopUnroll.cpp =================================================================== --- lib/Transforms/Utils/LoopUnroll.cpp +++ lib/Transforms/Utils/LoopUnroll.cpp @@ -93,66 +93,6 @@ } } -/// Folds a basic block into its predecessor if it only has one predecessor, and -/// that predecessor only has one successor. -/// The LoopInfo Analysis that is passed will be kept consistent. -BasicBlock *llvm::foldBlockIntoPredecessor(BasicBlock *BB, LoopInfo *LI, - ScalarEvolution *SE, - DominatorTree *DT) { - // Merge basic blocks into their predecessor if there is only one distinct - // pred, and if there is only one distinct successor of the predecessor, and - // if there are no PHI nodes. - BasicBlock *OnlyPred = BB->getSinglePredecessor(); - if (!OnlyPred) return nullptr; - - if (OnlyPred->getTerminator()->getNumSuccessors() != 1) - return nullptr; - - LLVM_DEBUG(dbgs() << "Merging: " << BB->getName() << " into " - << OnlyPred->getName() << "\n"); - - // Resolve any PHI nodes at the start of the block. They are all - // guaranteed to have exactly one entry if they exist, unless there are - // multiple duplicate (but guaranteed to be equal) entries for the - // incoming edges. This occurs when there are multiple edges from - // OnlyPred to OnlySucc. - FoldSingleEntryPHINodes(BB); - - // Delete the unconditional branch from the predecessor... - OnlyPred->getInstList().pop_back(); - - // Make all PHI nodes that referred to BB now refer to Pred as their - // source... - BB->replaceAllUsesWith(OnlyPred); - - // Move all definitions in the successor to the predecessor... - OnlyPred->getInstList().splice(OnlyPred->end(), BB->getInstList()); - - // OldName will be valid until erased. - StringRef OldName = BB->getName(); - - // Erase the old block and update dominator info. - if (DT) - if (DomTreeNode *DTN = DT->getNode(BB)) { - DomTreeNode *PredDTN = DT->getNode(OnlyPred); - SmallVector Children(DTN->begin(), DTN->end()); - for (auto *DI : Children) - DT->changeImmediateDominator(DI, PredDTN); - - DT->eraseNode(BB); - } - - LI->removeBlock(BB); - - // Inherit predecessor's name if it exists... - if (!OldName.empty() && !OnlyPred->hasName()) - OnlyPred->setName(OldName); - - BB->eraseFromParent(); - - return OnlyPred; -} - /// Check if unrolling created a situation where we need to insert phi nodes to /// preserve LCSSA form. /// \param Blocks is a vector of basic blocks representing unrolled loop. @@ -818,12 +758,14 @@ assert(!DT || !UnrollVerifyDomtree || DT->verify(DominatorTree::VerificationLevel::Fast)); + DomTreeUpdater DTU(DT, DomTreeUpdater::UpdateStrategy::Eager); // Merge adjacent basic blocks, if possible. for (BasicBlock *Latch : Latches) { BranchInst *Term = cast(Latch->getTerminator()); if (Term->isUnconditional()) { BasicBlock *Dest = Term->getSuccessor(0); - if (BasicBlock *Fold = foldBlockIntoPredecessor(Dest, LI, SE, DT)) { + BasicBlock *Fold = Dest->getUniquePredecessor(); + if (MergeBlockIntoPredecessor(Dest, &DTU, LI)) { // Dest has been folded into Fold. Update our worklists accordingly. std::replace(Latches.begin(), Latches.end(), Dest, Fold); UnrolledLoopBlocks.erase(std::remove(UnrolledLoopBlocks.begin(), Index: lib/Transforms/Utils/LoopUnrollAndJam.cpp =================================================================== --- lib/Transforms/Utils/LoopUnrollAndJam.cpp +++ lib/Transforms/Utils/LoopUnrollAndJam.cpp @@ -538,12 +538,14 @@ MergeBlocks.insert(ForeBlocksLast.begin(), ForeBlocksLast.end()); MergeBlocks.insert(SubLoopBlocksLast.begin(), SubLoopBlocksLast.end()); MergeBlocks.insert(AftBlocksLast.begin(), AftBlocksLast.end()); + DomTreeUpdater DTU(DT, DomTreeUpdater::UpdateStrategy::Eager); while (!MergeBlocks.empty()) { BasicBlock *BB = *MergeBlocks.begin(); BranchInst *Term = dyn_cast(BB->getTerminator()); if (Term && Term->isUnconditional() && L->contains(Term->getSuccessor(0))) { BasicBlock *Dest = Term->getSuccessor(0); - if (BasicBlock *Fold = foldBlockIntoPredecessor(Dest, LI, SE, DT)) { + BasicBlock *Fold = Dest->getUniquePredecessor(); + if (MergeBlockIntoPredecessor(Dest, &DTU, LI)) { // Don't remove BB and add Fold as they are the same BB assert(Fold == BB); (void)Fold; Index: test/CodeGen/AArch64/loop-micro-op-buffer-size-t99.ll =================================================================== --- test/CodeGen/AArch64/loop-micro-op-buffer-size-t99.ll +++ test/CodeGen/AArch64/loop-micro-op-buffer-size-t99.ll @@ -1,5 +1,5 @@ ; REQUIRES: asserts -; RUN: opt -mcpu=thunderx2t99 -loop-unroll --debug-only=loop-unroll -S -unroll-allow-partial < %s 2>&1 | FileCheck %s +; RUN: opt -mcpu=thunderx2t99 -loop-unroll --debug-only=loop-unroll --debug-only=basicblock-utils -S -unroll-allow-partial < %s 2>&1 | FileCheck %s target triple = "aarch64-unknown-linux-gnu"