llvm/lib/CodeGen/AsmPrinter/DebugHandlerBase.cpp

   1 //===-- llvm/lib/CodeGen/AsmPrinter/DebugHandlerBase.cpp -------*- C++ -*--===//
   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 // Common functionality for different debug information format backends.
  10 // LLVM currently supports DWARF and CodeView.
  11 //
  12 //===----------------------------------------------------------------------===//
  13
  14 #include "llvm/CodeGen/DebugHandlerBase.h"
  15 #include "llvm/ADT/Optional.h"
  16 #include "llvm/ADT/Twine.h"
  17 #include "llvm/CodeGen/AsmPrinter.h"
  18 #include "llvm/CodeGen/MachineFunction.h"
  19 #include "llvm/CodeGen/MachineInstr.h"
  20 #include "llvm/CodeGen/MachineModuleInfo.h"
  21 #include "llvm/CodeGen/TargetSubtargetInfo.h"
  22 #include "llvm/IR/DebugInfo.h"
  23 #include "llvm/MC/MCStreamer.h"
  24 #include "llvm/Support/CommandLine.h"
  25
  26 using namespace llvm;
  27
  28 #define DEBUG_TYPE "dwarfdebug"
  29
  30 /// If true, we drop variable location ranges which exist entirely outside the
  31 /// variable's lexical scope instruction ranges.
  32 static cl::opt<bool> TrimVarLocs("trim-var-locs", cl::Hidden, cl::init(true));
  33
  34 Optional<DbgVariableLocation>
  35 DbgVariableLocation::extractFromMachineInstruction(
  36     const MachineInstr &Instruction) {
  37   DbgVariableLocation Location;
  38   // Variables calculated from multiple locations can't be represented here.
  39   if (Instruction.getNumDebugOperands() != 1)
  40     return None;
  41   if (!Instruction.getDebugOperand(0).isReg())
  42     return None;
  43   Location.Register = Instruction.getDebugOperand(0).getReg();
  44   Location.FragmentInfo.reset();
  45   // We only handle expressions generated by DIExpression::appendOffset,
  46   // which doesn't require a full stack machine.
  47   int64_t Offset = 0;
  48   const DIExpression *DIExpr = Instruction.getDebugExpression();
  49   auto Op = DIExpr->expr_op_begin();
  50   // We can handle a DBG_VALUE_LIST iff it has exactly one location operand that
  51   // appears exactly once at the start of the expression.
  52   if (Instruction.isDebugValueList()) {
  53     if (Instruction.getNumDebugOperands() == 1 &&
  54         Op->getOp() == dwarf::DW_OP_LLVM_arg)
  55       ++Op;
  56     else
  57       return None;
  58   }
  59   while (Op != DIExpr->expr_op_end()) {
  60     switch (Op->getOp()) {
  61     case dwarf::DW_OP_constu: {
  62       int Value = Op->getArg(0);
  63       ++Op;
  64       if (Op != DIExpr->expr_op_end()) {
  65         switch (Op->getOp()) {
  66         case dwarf::DW_OP_minus:
  67           Offset -= Value;
  68           break;
  69         case dwarf::DW_OP_plus:
  70           Offset += Value;
  71           break;
  72         default:
  73           continue;
  74         }
  75       }
  76     } break;
  77     case dwarf::DW_OP_plus_uconst:
  78       Offset += Op->getArg(0);
  79       break;
  80     case dwarf::DW_OP_LLVM_fragment:
  81       Location.FragmentInfo = {Op->getArg(1), Op->getArg(0)};
  82       break;
  83     case dwarf::DW_OP_deref:
  84       Location.LoadChain.push_back(Offset);
  85       Offset = 0;
  86       break;
  87     default:
  88       return None;
  89     }
  90     ++Op;
  91   }
  92
  93   // Do one final implicit DW_OP_deref if this was an indirect DBG_VALUE
  94   // instruction.
  95   // FIXME: Replace these with DIExpression.
  96   if (Instruction.isIndirectDebugValue())
  97     Location.LoadChain.push_back(Offset);
  98
  99   return Location;
 100 }
 101
 102 DebugHandlerBase::DebugHandlerBase(AsmPrinter *A) : Asm(A), MMI(Asm->MMI) {}
 103
 104 void DebugHandlerBase::beginModule(Module *M) {
 105   if (M->debug_compile_units().empty())
 106     Asm = nullptr;
 107 }
 108
 109 // Each LexicalScope has first instruction and last instruction to mark
 110 // beginning and end of a scope respectively. Create an inverse map that list
 111 // scopes starts (and ends) with an instruction. One instruction may start (or
 112 // end) multiple scopes. Ignore scopes that are not reachable.
 113 void DebugHandlerBase::identifyScopeMarkers() {
 114   SmallVector<LexicalScope *, 4> WorkList;
 115   WorkList.push_back(LScopes.getCurrentFunctionScope());
 116   while (!WorkList.empty()) {
 117     LexicalScope *S = WorkList.pop_back_val();
 118
 119     const SmallVectorImpl<LexicalScope *> &Children = S->getChildren();
 120     if (!Children.empty())
 121       WorkList.append(Children.begin(), Children.end());
 122
 123     if (S->isAbstractScope())
 124       continue;
 125
 126     for (const InsnRange &R : S->getRanges()) {
 127       assert(R.first && "InsnRange does not have first instruction!");
 128       assert(R.second && "InsnRange does not have second instruction!");
 129       requestLabelBeforeInsn(R.first);
 130       requestLabelAfterInsn(R.second);
 131     }
 132   }
 133 }
 134
 135 // Return Label preceding the instruction.
 136 MCSymbol *DebugHandlerBase::getLabelBeforeInsn(const MachineInstr *MI) {
 137   MCSymbol *Label = LabelsBeforeInsn.lookup(MI);
 138   assert(Label && "Didn't insert label before instruction");
 139   return Label;
 140 }
 141
 142 // Return Label immediately following the instruction.
 143 MCSymbol *DebugHandlerBase::getLabelAfterInsn(const MachineInstr *MI) {
 144   return LabelsAfterInsn.lookup(MI);
 145 }
 146
 147 /// If this type is derived from a base type then return base type size.
 148 uint64_t DebugHandlerBase::getBaseTypeSize(const DIType *Ty) {
 149   assert(Ty);
 150   const DIDerivedType *DDTy = dyn_cast<DIDerivedType>(Ty);
 151   if (!DDTy)
 152     return Ty->getSizeInBits();
 153
 154   unsigned Tag = DDTy->getTag();
 155
 156   if (Tag != dwarf::DW_TAG_member && Tag != dwarf::DW_TAG_typedef &&
 157       Tag != dwarf::DW_TAG_const_type && Tag != dwarf::DW_TAG_volatile_type &&
 158       Tag != dwarf::DW_TAG_restrict_type && Tag != dwarf::DW_TAG_atomic_type &&
 159       Tag != dwarf::DW_TAG_immutable_type)
 160     return DDTy->getSizeInBits();
 161
 162   DIType *BaseType = DDTy->getBaseType();
 163
 164   if (!BaseType)
 165     return 0;
 166
 167   // If this is a derived type, go ahead and get the base type, unless it's a
 168   // reference then it's just the size of the field. Pointer types have no need
 169   // of this since they're a different type of qualification on the type.
 170   if (BaseType->getTag() == dwarf::DW_TAG_reference_type ||
 171       BaseType->getTag() == dwarf::DW_TAG_rvalue_reference_type)
 172     return Ty->getSizeInBits();
 173
 174   return getBaseTypeSize(BaseType);
 175 }
 176
 177 bool DebugHandlerBase::isUnsignedDIType(const DIType *Ty) {
 178   if (isa<DIStringType>(Ty)) {
 179     // Some transformations (e.g. instcombine) may decide to turn a Fortran
 180     // character object into an integer, and later ones (e.g. SROA) may
 181     // further inject a constant integer in a llvm.dbg.value call to track
 182     // the object's value. Here we trust the transformations are doing the
 183     // right thing, and treat the constant as unsigned to preserve that value
 184     // (i.e. avoid sign extension).
 185     return true;
 186   }
 187
 188   if (auto *CTy = dyn_cast<DICompositeType>(Ty)) {
 189     if (CTy->getTag() == dwarf::DW_TAG_enumeration_type) {
 190       if (!(Ty = CTy->getBaseType()))
 191         // FIXME: Enums without a fixed underlying type have unknown signedness
 192         // here, leading to incorrectly emitted constants.
 193         return false;
 194     } else
 195       // (Pieces of) aggregate types that get hacked apart by SROA may be
 196       // represented by a constant. Encode them as unsigned bytes.
 197       return true;
 198   }
 199
 200   if (auto *DTy = dyn_cast<DIDerivedType>(Ty)) {
 201     dwarf::Tag T = (dwarf::Tag)Ty->getTag();
 202     // Encode pointer constants as unsigned bytes. This is used at least for
 203     // null pointer constant emission.
 204     // FIXME: reference and rvalue_reference /probably/ shouldn't be allowed
 205     // here, but accept them for now due to a bug in SROA producing bogus
 206     // dbg.values.
 207     if (T == dwarf::DW_TAG_pointer_type ||
 208         T == dwarf::DW_TAG_ptr_to_member_type ||
 209         T == dwarf::DW_TAG_reference_type ||
 210         T == dwarf::DW_TAG_rvalue_reference_type)
 211       return true;
 212     assert(T == dwarf::DW_TAG_typedef || T == dwarf::DW_TAG_const_type ||
 213            T == dwarf::DW_TAG_volatile_type ||
 214            T == dwarf::DW_TAG_restrict_type || T == dwarf::DW_TAG_atomic_type ||
 215            T == dwarf::DW_TAG_immutable_type);
 216     assert(DTy->getBaseType() && "Expected valid base type");
 217     return isUnsignedDIType(DTy->getBaseType());
 218   }
 219
 220   auto *BTy = cast<DIBasicType>(Ty);
 221   unsigned Encoding = BTy->getEncoding();
 222   assert((Encoding == dwarf::DW_ATE_unsigned ||
 223           Encoding == dwarf::DW_ATE_unsigned_char ||
 224           Encoding == dwarf::DW_ATE_signed ||
 225           Encoding == dwarf::DW_ATE_signed_char ||
 226           Encoding == dwarf::DW_ATE_float || Encoding == dwarf::DW_ATE_UTF ||
 227           Encoding == dwarf::DW_ATE_boolean ||
 228           (Ty->getTag() == dwarf::DW_TAG_unspecified_type &&
 229            Ty->getName() == "decltype(nullptr)")) &&
 230          "Unsupported encoding");
 231   return Encoding == dwarf::DW_ATE_unsigned ||
 232          Encoding == dwarf::DW_ATE_unsigned_char ||
 233          Encoding == dwarf::DW_ATE_UTF || Encoding == dwarf::DW_ATE_boolean ||
 234          Ty->getTag() == dwarf::DW_TAG_unspecified_type;
 235 }
 236
 237 static bool hasDebugInfo(const MachineModuleInfo *MMI,
 238                          const MachineFunction *MF) {
 239   if (!MMI->hasDebugInfo())
 240     return false;
 241   auto *SP = MF->getFunction().getSubprogram();
 242   if (!SP)
 243     return false;
 244   assert(SP->getUnit());
 245   auto EK = SP->getUnit()->getEmissionKind();
 246   if (EK == DICompileUnit::NoDebug)
 247     return false;
 248   return true;
 249 }
 250
 251 void DebugHandlerBase::beginFunction(const MachineFunction *MF) {
 252   PrevInstBB = nullptr;
 253
 254   if (!Asm || !hasDebugInfo(MMI, MF)) {
 255     skippedNonDebugFunction();
 256     return;
 257   }
 258
 259   // Grab the lexical scopes for the function, if we don't have any of those
 260   // then we're not going to be able to do anything.
 261   LScopes.initialize(*MF);
 262   if (LScopes.empty()) {
 263     beginFunctionImpl(MF);
 264     return;
 265   }
 266
 267   // Make sure that each lexical scope will have a begin/end label.
 268   identifyScopeMarkers();
 269
 270   // Calculate history for local variables.
 271   assert(DbgValues.empty() && "DbgValues map wasn't cleaned!");
 272   assert(DbgLabels.empty() && "DbgLabels map wasn't cleaned!");
 273   calculateDbgEntityHistory(MF, Asm->MF->getSubtarget().getRegisterInfo(),
 274                             DbgValues, DbgLabels);
 275   InstOrdering.initialize(*MF);
 276   if (TrimVarLocs)
 277     DbgValues.trimLocationRanges(*MF, LScopes, InstOrdering);
 278   LLVM_DEBUG(DbgValues.dump());
 279
 280   // Request labels for the full history.
 281   for (const auto &I : DbgValues) {
 282     const auto &Entries = I.second;
 283     if (Entries.empty())
 284       continue;
 285
 286     auto IsDescribedByReg = [](const MachineInstr *MI) {
 287       return any_of(MI->debug_operands(),
 288                     [](auto &MO) { return MO.isReg() && MO.getReg(); });
 289     };
 290
 291     // The first mention of a function argument gets the CurrentFnBegin label,
 292     // so arguments are visible when breaking at function entry.
 293     //
 294     // We do not change the label for values that are described by registers,
 295     // as that could place them above their defining instructions. We should
 296     // ideally not change the labels for constant debug values either, since
 297     // doing that violates the ranges that are calculated in the history map.
 298     // However, we currently do not emit debug values for constant arguments
 299     // directly at the start of the function, so this code is still useful.
 300     const DILocalVariable *DIVar =
 301         Entries.front().getInstr()->getDebugVariable();
 302     if (DIVar->isParameter() &&
 303         getDISubprogram(DIVar->getScope())->describes(&MF->getFunction())) {
 304       if (!IsDescribedByReg(Entries.front().getInstr()))
 305         LabelsBeforeInsn[Entries.front().getInstr()] = Asm->getFunctionBegin();
 306       if (Entries.front().getInstr()->getDebugExpression()->isFragment()) {
 307         // Mark all non-overlapping initial fragments.
 308         for (auto I = Entries.begin(); I != Entries.end(); ++I) {
 309           if (!I->isDbgValue())
 310             continue;
 311           const DIExpression *Fragment = I->getInstr()->getDebugExpression();
 312           if (std::any_of(Entries.begin(), I,
 313                           [&](DbgValueHistoryMap::Entry Pred) {
 314                             return Pred.isDbgValue() &&
 315                                    Fragment->fragmentsOverlap(
 316                                        Pred.getInstr()->getDebugExpression());
 317                           }))
 318             break;
 319           // The code that generates location lists for DWARF assumes that the
 320           // entries' start labels are monotonically increasing, and since we
 321           // don't change the label for fragments that are described by
 322           // registers, we must bail out when encountering such a fragment.
 323           if (IsDescribedByReg(I->getInstr()))
 324             break;
 325           LabelsBeforeInsn[I->getInstr()] = Asm->getFunctionBegin();
 326         }
 327       }
 328     }
 329
 330     for (const auto &Entry : Entries) {
 331       if (Entry.isDbgValue())
 332         requestLabelBeforeInsn(Entry.getInstr());
 333       else
 334         requestLabelAfterInsn(Entry.getInstr());
 335     }
 336   }
 337
 338   // Ensure there is a symbol before DBG_LABEL.
 339   for (const auto &I : DbgLabels) {
 340     const MachineInstr *MI = I.second;
 341     requestLabelBeforeInsn(MI);
 342   }
 343
 344   PrevInstLoc = DebugLoc();
 345   PrevLabel = Asm->getFunctionBegin();
 346   beginFunctionImpl(MF);
 347 }
 348
 349 void DebugHandlerBase::beginInstruction(const MachineInstr *MI) {
 350   if (!Asm || !MMI->hasDebugInfo())
 351     return;
 352
 353   assert(CurMI == nullptr);
 354   CurMI = MI;
 355
 356   // Insert labels where requested.
 357   DenseMap<const MachineInstr *, MCSymbol *>::iterator I =
 358       LabelsBeforeInsn.find(MI);
 359
 360   // No label needed.
 361   if (I == LabelsBeforeInsn.end())
 362     return;
 363
 364   // Label already assigned.
 365   if (I->second)
 366     return;
 367
 368   if (!PrevLabel) {
 369     PrevLabel = MMI->getContext().createTempSymbol();
 370     Asm->OutStreamer->emitLabel(PrevLabel);
 371   }
 372   I->second = PrevLabel;
 373 }
 374
 375 void DebugHandlerBase::endInstruction() {
 376   if (!Asm || !MMI->hasDebugInfo())
 377     return;
 378
 379   assert(CurMI != nullptr);
 380   // Don't create a new label after DBG_VALUE and other instructions that don't
 381   // generate code.
 382   if (!CurMI->isMetaInstruction()) {
 383     PrevLabel = nullptr;
 384     PrevInstBB = CurMI->getParent();
 385   }
 386
 387   DenseMap<const MachineInstr *, MCSymbol *>::iterator I =
 388       LabelsAfterInsn.find(CurMI);
 389
 390   // No label needed or label already assigned.
 391   if (I == LabelsAfterInsn.end() || I->second) {
 392     CurMI = nullptr;
 393     return;
 394   }
 395
 396   // We need a label after this instruction.  With basic block sections, just
 397   // use the end symbol of the section if this is the last instruction of the
 398   // section.  This reduces the need for an additional label and also helps
 399   // merging ranges.
 400   if (CurMI->getParent()->isEndSection() && CurMI->getNextNode() == nullptr) {
 401     PrevLabel = CurMI->getParent()->getEndSymbol();
 402   } else if (!PrevLabel) {
 403     PrevLabel = MMI->getContext().createTempSymbol();
 404     Asm->OutStreamer->emitLabel(PrevLabel);
 405   }
 406   I->second = PrevLabel;
 407   CurMI = nullptr;
 408 }
 409
 410 void DebugHandlerBase::endFunction(const MachineFunction *MF) {
 411   if (Asm && hasDebugInfo(MMI, MF))
 412     endFunctionImpl(MF);
 413   DbgValues.clear();
 414   DbgLabels.clear();
 415   LabelsBeforeInsn.clear();
 416   LabelsAfterInsn.clear();
 417   InstOrdering.clear();
 418 }
 419
 420 void DebugHandlerBase::beginBasicBlock(const MachineBasicBlock &MBB) {
 421   if (!MBB.isBeginSection())
 422     return;
 423
 424   PrevLabel = MBB.getSymbol();
 425 }
 426
 427 void DebugHandlerBase::endBasicBlock(const MachineBasicBlock &MBB) {
 428   if (!MBB.isEndSection())
 429     return;
 430
 431   PrevLabel = nullptr;
 432 }