lib/CodeGen/GCRootLowering.cpp

   1 //===-- GCRootLowering.cpp - Garbage collection infrastructure ------------===//
   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 file implements the lowering for the gc.root mechanism.
  10 //
  11 //===----------------------------------------------------------------------===//
  12
  13 #include "llvm/CodeGen/GCMetadata.h"
  14 #include "llvm/CodeGen/GCStrategy.h"
  15 #include "llvm/CodeGen/MachineFrameInfo.h"
  16 #include "llvm/CodeGen/MachineFunctionPass.h"
  17 #include "llvm/CodeGen/MachineInstrBuilder.h"
  18 #include "llvm/CodeGen/MachineModuleInfo.h"
  19 #include "llvm/CodeGen/Passes.h"
  20 #include "llvm/CodeGen/TargetFrameLowering.h"
  21 #include "llvm/CodeGen/TargetInstrInfo.h"
  22 #include "llvm/CodeGen/TargetRegisterInfo.h"
  23 #include "llvm/CodeGen/TargetSubtargetInfo.h"
  24 #include "llvm/IR/Dominators.h"
  25 #include "llvm/IR/IntrinsicInst.h"
  26 #include "llvm/IR/Module.h"
  27 #include "llvm/Support/Debug.h"
  28 #include "llvm/Support/ErrorHandling.h"
  29 #include "llvm/Support/raw_ostream.h"
  30
  31 using namespace llvm;
  32
  33 namespace {
  34
  35 /// LowerIntrinsics - This pass rewrites calls to the llvm.gcread or
  36 /// llvm.gcwrite intrinsics, replacing them with simple loads and stores as
  37 /// directed by the GCStrategy. It also performs automatic root initialization
  38 /// and custom intrinsic lowering.
  39 class LowerIntrinsics : public FunctionPass {
  40   bool DoLowering(Function &F, GCStrategy &S);
  41
  42 public:
  43   static char ID;
  44
  45   LowerIntrinsics();
  46   StringRef getPassName() const override;
  47   void getAnalysisUsage(AnalysisUsage &AU) const override;
  48
  49   bool doInitialization(Module &M) override;
  50   bool runOnFunction(Function &F) override;
  51 };
  52
  53 /// GCMachineCodeAnalysis - This is a target-independent pass over the machine
  54 /// function representation to identify safe points for the garbage collector
  55 /// in the machine code. It inserts labels at safe points and populates a
  56 /// GCMetadata record for each function.
  57 class GCMachineCodeAnalysis : public MachineFunctionPass {
  58   GCFunctionInfo *FI;
  59   MachineModuleInfo *MMI;
  60   const TargetInstrInfo *TII;
  61
  62   void FindSafePoints(MachineFunction &MF);
  63   void VisitCallPoint(MachineBasicBlock::iterator CI);
  64   MCSymbol *InsertLabel(MachineBasicBlock &MBB, MachineBasicBlock::iterator MI,
  65                         const DebugLoc &DL) const;
  66
  67   void FindStackOffsets(MachineFunction &MF);
  68
  69 public:
  70   static char ID;
  71
  72   GCMachineCodeAnalysis();
  73   void getAnalysisUsage(AnalysisUsage &AU) const override;
  74
  75   bool runOnMachineFunction(MachineFunction &MF) override;
  76 };
  77 }
  78
  79 // -----------------------------------------------------------------------------
  80
  81 INITIALIZE_PASS_BEGIN(LowerIntrinsics, "gc-lowering", "GC Lowering", false,
  82                       false)
  83 INITIALIZE_PASS_DEPENDENCY(GCModuleInfo)
  84 INITIALIZE_PASS_END(LowerIntrinsics, "gc-lowering", "GC Lowering", false, false)
  85
  86 FunctionPass *llvm::createGCLoweringPass() { return new LowerIntrinsics(); }
  87
  88 char LowerIntrinsics::ID = 0;
  89
  90 LowerIntrinsics::LowerIntrinsics() : FunctionPass(ID) {
  91   initializeLowerIntrinsicsPass(*PassRegistry::getPassRegistry());
  92 }
  93
  94 StringRef LowerIntrinsics::getPassName() const {
  95   return "Lower Garbage Collection Instructions";
  96 }
  97
  98 void LowerIntrinsics::getAnalysisUsage(AnalysisUsage &AU) const {
  99   FunctionPass::getAnalysisUsage(AU);
 100   AU.addRequired<GCModuleInfo>();
 101   AU.addPreserved<DominatorTreeWrapperPass>();
 102 }
 103
 104 /// doInitialization - If this module uses the GC intrinsics, find them now.
 105 bool LowerIntrinsics::doInitialization(Module &M) {
 106   GCModuleInfo *MI = getAnalysisIfAvailable<GCModuleInfo>();
 107   assert(MI && "LowerIntrinsics didn't require GCModuleInfo!?");
 108   for (Module::iterator I = M.begin(), E = M.end(); I != E; ++I)
 109     if (!I->isDeclaration() && I->hasGC())
 110       MI->getFunctionInfo(*I); // Instantiate the GC strategy.
 111
 112   return false;
 113 }
 114
 115 /// CouldBecomeSafePoint - Predicate to conservatively determine whether the
 116 /// instruction could introduce a safe point.
 117 static bool CouldBecomeSafePoint(Instruction *I) {
 118   // The natural definition of instructions which could introduce safe points
 119   // are:
 120   //
 121   //   - call, invoke (AfterCall, BeforeCall)
 122   //   - phis (Loops)
 123   //   - invoke, ret, unwind (Exit)
 124   //
 125   // However, instructions as seemingly inoccuous as arithmetic can become
 126   // libcalls upon lowering (e.g., div i64 on a 32-bit platform), so instead
 127   // it is necessary to take a conservative approach.
 128
 129   if (isa<AllocaInst>(I) || isa<GetElementPtrInst>(I) || isa<StoreInst>(I) ||
 130       isa<LoadInst>(I))
 131     return false;
 132
 133   // llvm.gcroot is safe because it doesn't do anything at runtime.
 134   if (CallInst *CI = dyn_cast<CallInst>(I))
 135     if (Function *F = CI->getCalledFunction())
 136       if (Intrinsic::ID IID = F->getIntrinsicID())
 137         if (IID == Intrinsic::gcroot)
 138           return false;
 139
 140   return true;
 141 }
 142
 143 static bool InsertRootInitializers(Function &F, ArrayRef<AllocaInst *> Roots) {
 144   // Scroll past alloca instructions.
 145   BasicBlock::iterator IP = F.getEntryBlock().begin();
 146   while (isa<AllocaInst>(IP))
 147     ++IP;
 148
 149   // Search for initializers in the initial BB.
 150   SmallPtrSet<AllocaInst *, 16> InitedRoots;
 151   for (; !CouldBecomeSafePoint(&*IP); ++IP)
 152     if (StoreInst *SI = dyn_cast<StoreInst>(IP))
 153       if (AllocaInst *AI =
 154               dyn_cast<AllocaInst>(SI->getOperand(1)->stripPointerCasts()))
 155         InitedRoots.insert(AI);
 156
 157   // Add root initializers.
 158   bool MadeChange = false;
 159
 160   for (AllocaInst *Root : Roots)
 161     if (!InitedRoots.count(Root)) {
 162       StoreInst *SI = new StoreInst(
 163           ConstantPointerNull::get(cast<PointerType>(Root->getAllocatedType())),
 164           Root);
 165       SI->insertAfter(Root);
 166       MadeChange = true;
 167     }
 168
 169   return MadeChange;
 170 }
 171
 172 /// runOnFunction - Replace gcread/gcwrite intrinsics with loads and stores.
 173 /// Leave gcroot intrinsics; the code generator needs to see those.
 174 bool LowerIntrinsics::runOnFunction(Function &F) {
 175   // Quick exit for functions that do not use GC.
 176   if (!F.hasGC())
 177     return false;
 178
 179   GCFunctionInfo &FI = getAnalysis<GCModuleInfo>().getFunctionInfo(F);
 180   GCStrategy &S = FI.getStrategy();
 181
 182   return DoLowering(F, S);
 183 }
 184
 185 /// Lower barriers out of existance (if the associated GCStrategy hasn't
 186 /// already done so...), and insert initializing stores to roots as a defensive
 187 /// measure.  Given we're going to report all roots live at all safepoints, we
 188 /// need to be able to ensure each root has been initialized by the point the
 189 /// first safepoint is reached.  This really should have been done by the
 190 /// frontend, but the old API made this non-obvious, so we do a potentially
 191 /// redundant store just in case.
 192 bool LowerIntrinsics::DoLowering(Function &F, GCStrategy &S) {
 193   SmallVector<AllocaInst *, 32> Roots;
 194
 195   bool MadeChange = false;
 196   for (BasicBlock &BB : F)
 197     for (BasicBlock::iterator II = BB.begin(), E = BB.end(); II != E;) {
 198       IntrinsicInst *CI = dyn_cast<IntrinsicInst>(II++);
 199       if (!CI)
 200         continue;
 201
 202       Function *F = CI->getCalledFunction();
 203       switch (F->getIntrinsicID()) {
 204       default: break;
 205       case Intrinsic::gcwrite: {
 206         // Replace a write barrier with a simple store.
 207         Value *St = new StoreInst(CI->getArgOperand(0),
 208                                   CI->getArgOperand(2), CI);
 209         CI->replaceAllUsesWith(St);
 210         CI->eraseFromParent();
 211         MadeChange = true;
 212         break;
 213       }
 214       case Intrinsic::gcread: {
 215         // Replace a read barrier with a simple load.
 216         Value *Ld = new LoadInst(CI->getType(), CI->getArgOperand(1), "", CI);
 217         Ld->takeName(CI);
 218         CI->replaceAllUsesWith(Ld);
 219         CI->eraseFromParent();
 220         MadeChange = true;
 221         break;
 222       }
 223       case Intrinsic::gcroot: {
 224         // Initialize the GC root, but do not delete the intrinsic. The
 225         // backend needs the intrinsic to flag the stack slot.
 226         Roots.push_back(
 227             cast<AllocaInst>(CI->getArgOperand(0)->stripPointerCasts()));
 228         break;
 229       }
 230       }
 231     }
 232
 233   if (Roots.size())
 234     MadeChange |= InsertRootInitializers(F, Roots);
 235
 236   return MadeChange;
 237 }
 238
 239 // -----------------------------------------------------------------------------
 240
 241 char GCMachineCodeAnalysis::ID = 0;
 242 char &llvm::GCMachineCodeAnalysisID = GCMachineCodeAnalysis::ID;
 243
 244 INITIALIZE_PASS(GCMachineCodeAnalysis, "gc-analysis",
 245                 "Analyze Machine Code For Garbage Collection", false, false)
 246
 247 GCMachineCodeAnalysis::GCMachineCodeAnalysis() : MachineFunctionPass(ID) {}
 248
 249 void GCMachineCodeAnalysis::getAnalysisUsage(AnalysisUsage &AU) const {
 250   MachineFunctionPass::getAnalysisUsage(AU);
 251   AU.setPreservesAll();
 252   AU.addRequired<MachineModuleInfo>();
 253   AU.addRequired<GCModuleInfo>();
 254 }
 255
 256 MCSymbol *GCMachineCodeAnalysis::InsertLabel(MachineBasicBlock &MBB,
 257                                              MachineBasicBlock::iterator MI,
 258                                              const DebugLoc &DL) const {
 259   MCSymbol *Label = MBB.getParent()->getContext().createTempSymbol();
 260   BuildMI(MBB, MI, DL, TII->get(TargetOpcode::GC_LABEL)).addSym(Label);
 261   return Label;
 262 }
 263
 264 void GCMachineCodeAnalysis::VisitCallPoint(MachineBasicBlock::iterator CI) {
 265   // Find the return address (next instruction), since that's what will be on
 266   // the stack when the call is suspended and we need to inspect the stack.
 267   MachineBasicBlock::iterator RAI = CI;
 268   ++RAI;
 269
 270   MCSymbol *Label = InsertLabel(*CI->getParent(), RAI, CI->getDebugLoc());
 271   FI->addSafePoint(Label, CI->getDebugLoc());
 272 }
 273
 274 void GCMachineCodeAnalysis::FindSafePoints(MachineFunction &MF) {
 275   for (MachineBasicBlock &MBB : MF)
 276     for (MachineBasicBlock::iterator MI = MBB.begin(), ME = MBB.end();
 277          MI != ME; ++MI)
 278       if (MI->isCall()) {
 279         // Do not treat tail or sibling call sites as safe points.  This is
 280         // legal since any arguments passed to the callee which live in the
 281         // remnants of the callers frame will be owned and updated by the
 282         // callee if required.
 283         if (MI->isTerminator())
 284           continue;
 285         VisitCallPoint(MI);
 286       }
 287 }
 288
 289 void GCMachineCodeAnalysis::FindStackOffsets(MachineFunction &MF) {
 290   const TargetFrameLowering *TFI = MF.getSubtarget().getFrameLowering();
 291   assert(TFI && "TargetRegisterInfo not available!");
 292
 293   for (GCFunctionInfo::roots_iterator RI = FI->roots_begin();
 294        RI != FI->roots_end();) {
 295     // If the root references a dead object, no need to keep it.
 296     if (MF.getFrameInfo().isDeadObjectIndex(RI->Num)) {
 297       RI = FI->removeStackRoot(RI);
 298     } else {
 299       unsigned FrameReg; // FIXME: surely GCRoot ought to store the
 300                          // register that the offset is from?
 301       RI->StackOffset = TFI->getFrameIndexReference(MF, RI->Num, FrameReg);
 302       ++RI;
 303     }
 304   }
 305 }
 306
 307 bool GCMachineCodeAnalysis::runOnMachineFunction(MachineFunction &MF) {
 308   // Quick exit for functions that do not use GC.
 309   if (!MF.getFunction().hasGC())
 310     return false;
 311
 312   FI = &getAnalysis<GCModuleInfo>().getFunctionInfo(MF.getFunction());
 313   MMI = &getAnalysis<MachineModuleInfo>();
 314   TII = MF.getSubtarget().getInstrInfo();
 315
 316   // Find the size of the stack frame.  There may be no correct static frame
 317   // size, we use UINT64_MAX to represent this.
 318   const MachineFrameInfo &MFI = MF.getFrameInfo();
 319   const TargetRegisterInfo *RegInfo = MF.getSubtarget().getRegisterInfo();
 320   const bool DynamicFrameSize = MFI.hasVarSizedObjects() ||
 321     RegInfo->needsStackRealignment(MF);
 322   FI->setFrameSize(DynamicFrameSize ? UINT64_MAX : MFI.getStackSize());
 323
 324   // Find all safe points.
 325   if (FI->getStrategy().needsSafePoints())
 326     FindSafePoints(MF);
 327
 328   // Find the concrete stack offsets for all roots (stack slots)
 329   FindStackOffsets(MF);
 330
 331   return false;
 332 }