include/llvm/DebugInfo/DWARF/DWARFDebugLine.h

   1 //===- DWARFDebugLine.h -----------------------------------------*- 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 #ifndef LLVM_DEBUGINFO_DWARFDEBUGLINE_H
  10 #define LLVM_DEBUGINFO_DWARFDEBUGLINE_H
  11
  12 #include "llvm/ADT/Optional.h"
  13 #include "llvm/ADT/StringRef.h"
  14 #include "llvm/DebugInfo/DIContext.h"
  15 #include "llvm/DebugInfo/DWARF/DWARFCompileUnit.h"
  16 #include "llvm/DebugInfo/DWARF/DWARFDataExtractor.h"
  17 #include "llvm/DebugInfo/DWARF/DWARFFormValue.h"
  18 #include "llvm/DebugInfo/DWARF/DWARFRelocMap.h"
  19 #include "llvm/DebugInfo/DWARF/DWARFTypeUnit.h"
  20 #include "llvm/Support/MD5.h"
  21 #include <cstdint>
  22 #include <map>
  23 #include <string>
  24 #include <vector>
  25
  26 namespace llvm {
  27
  28 class DWARFUnit;
  29 class raw_ostream;
  30
  31 class DWARFDebugLine {
  32 public:
  33   struct FileNameEntry {
  34     FileNameEntry() = default;
  35
  36     DWARFFormValue Name;
  37     uint64_t DirIdx = 0;
  38     uint64_t ModTime = 0;
  39     uint64_t Length = 0;
  40     MD5::MD5Result Checksum;
  41     DWARFFormValue Source;
  42   };
  43
  44   /// Tracks which optional content types are present in a DWARF file name
  45   /// entry format.
  46   struct ContentTypeTracker {
  47     ContentTypeTracker() = default;
  48
  49     /// Whether filename entries provide a modification timestamp.
  50     bool HasModTime = false;
  51     /// Whether filename entries provide a file size.
  52     bool HasLength = false;
  53     /// For v5, whether filename entries provide an MD5 checksum.
  54     bool HasMD5 = false;
  55     /// For v5, whether filename entries provide source text.
  56     bool HasSource = false;
  57
  58     /// Update tracked content types with \p ContentType.
  59     void trackContentType(dwarf::LineNumberEntryFormat ContentType);
  60   };
  61
  62   struct Prologue {
  63     Prologue();
  64
  65     /// The size in bytes of the statement information for this compilation unit
  66     /// (not including the total_length field itself).
  67     uint64_t TotalLength;
  68     /// Version, address size (starting in v5), and DWARF32/64 format; these
  69     /// parameters affect interpretation of forms (used in the directory and
  70     /// file tables starting with v5).
  71     dwarf::FormParams FormParams;
  72     /// The number of bytes following the prologue_length field to the beginning
  73     /// of the first byte of the statement program itself.
  74     uint64_t PrologueLength;
  75     /// In v5, size in bytes of a segment selector.
  76     uint8_t SegSelectorSize;
  77     /// The size in bytes of the smallest target machine instruction. Statement
  78     /// program opcodes that alter the address register first multiply their
  79     /// operands by this value.
  80     uint8_t MinInstLength;
  81     /// The maximum number of individual operations that may be encoded in an
  82     /// instruction.
  83     uint8_t MaxOpsPerInst;
  84     /// The initial value of theis_stmtregister.
  85     uint8_t DefaultIsStmt;
  86     /// This parameter affects the meaning of the special opcodes. See below.
  87     int8_t LineBase;
  88     /// This parameter affects the meaning of the special opcodes. See below.
  89     uint8_t LineRange;
  90     /// The number assigned to the first special opcode.
  91     uint8_t OpcodeBase;
  92     /// This tracks which optional file format content types are present.
  93     ContentTypeTracker ContentTypes;
  94     std::vector<uint8_t> StandardOpcodeLengths;
  95     std::vector<DWARFFormValue> IncludeDirectories;
  96     std::vector<FileNameEntry> FileNames;
  97
  98     const dwarf::FormParams getFormParams() const { return FormParams; }
  99     uint16_t getVersion() const { return FormParams.Version; }
 100     uint8_t getAddressSize() const { return FormParams.AddrSize; }
 101     bool isDWARF64() const { return FormParams.Format == dwarf::DWARF64; }
 102
 103     uint32_t sizeofTotalLength() const { return isDWARF64() ? 12 : 4; }
 104
 105     uint32_t sizeofPrologueLength() const { return isDWARF64() ? 8 : 4; }
 106
 107     bool totalLengthIsValid() const;
 108
 109     /// Length of the prologue in bytes.
 110     uint32_t getLength() const {
 111       return PrologueLength + sizeofTotalLength() + sizeof(getVersion()) +
 112              sizeofPrologueLength();
 113     }
 114
 115     /// Length of the line table data in bytes (not including the prologue).
 116     uint32_t getStatementTableLength() const {
 117       return TotalLength + sizeofTotalLength() - getLength();
 118     }
 119
 120     int32_t getMaxLineIncrementForSpecialOpcode() const {
 121       return LineBase + (int8_t)LineRange - 1;
 122     }
 123
 124     void clear();
 125     void dump(raw_ostream &OS, DIDumpOptions DumpOptions) const;
 126     Error parse(const DWARFDataExtractor &DebugLineData, uint32_t *OffsetPtr,
 127                 const DWARFContext &Ctx, const DWARFUnit *U = nullptr);
 128   };
 129
 130   /// Standard .debug_line state machine structure.
 131   struct Row {
 132     explicit Row(bool DefaultIsStmt = false);
 133
 134     /// Called after a row is appended to the matrix.
 135     void postAppend();
 136     void reset(bool DefaultIsStmt);
 137     void dump(raw_ostream &OS) const;
 138
 139     static void dumpTableHeader(raw_ostream &OS);
 140
 141     static bool orderByAddress(const Row &LHS, const Row &RHS) {
 142       return LHS.Address < RHS.Address;
 143     }
 144
 145     /// The program-counter value corresponding to a machine instruction
 146     /// generated by the compiler.
 147     uint64_t Address;
 148     /// An unsigned integer indicating a source line number. Lines are numbered
 149     /// beginning at 1. The compiler may emit the value 0 in cases where an
 150     /// instruction cannot be attributed to any source line.
 151     uint32_t Line;
 152     /// An unsigned integer indicating a column number within a source line.
 153     /// Columns are numbered beginning at 1. The value 0 is reserved to indicate
 154     /// that a statement begins at the 'left edge' of the line.
 155     uint16_t Column;
 156     /// An unsigned integer indicating the identity of the source file
 157     /// corresponding to a machine instruction.
 158     uint16_t File;
 159     /// An unsigned integer representing the DWARF path discriminator value
 160     /// for this location.
 161     uint32_t Discriminator;
 162     /// An unsigned integer whose value encodes the applicable instruction set
 163     /// architecture for the current instruction.
 164     uint8_t Isa;
 165     /// A boolean indicating that the current instruction is the beginning of a
 166     /// statement.
 167     uint8_t IsStmt : 1,
 168         /// A boolean indicating that the current instruction is the
 169         /// beginning of a basic block.
 170         BasicBlock : 1,
 171         /// A boolean indicating that the current address is that of the
 172         /// first byte after the end of a sequence of target machine
 173         /// instructions.
 174         EndSequence : 1,
 175         /// A boolean indicating that the current address is one (of possibly
 176         /// many) where execution should be suspended for an entry breakpoint
 177         /// of a function.
 178         PrologueEnd : 1,
 179         /// A boolean indicating that the current address is one (of possibly
 180         /// many) where execution should be suspended for an exit breakpoint
 181         /// of a function.
 182         EpilogueBegin : 1;
 183   };
 184
 185   /// Represents a series of contiguous machine instructions. Line table for
 186   /// each compilation unit may consist of multiple sequences, which are not
 187   /// guaranteed to be in the order of ascending instruction address.
 188   struct Sequence {
 189     Sequence();
 190
 191     /// Sequence describes instructions at address range [LowPC, HighPC)
 192     /// and is described by line table rows [FirstRowIndex, LastRowIndex).
 193     uint64_t LowPC;
 194     uint64_t HighPC;
 195     unsigned FirstRowIndex;
 196     unsigned LastRowIndex;
 197     bool Empty;
 198
 199     void reset();
 200
 201     static bool orderByLowPC(const Sequence &LHS, const Sequence &RHS) {
 202       return LHS.LowPC < RHS.LowPC;
 203     }
 204
 205     bool isValid() const {
 206       return !Empty && (LowPC < HighPC) && (FirstRowIndex < LastRowIndex);
 207     }
 208
 209     bool containsPC(uint64_t PC) const { return (LowPC <= PC && PC < HighPC); }
 210   };
 211
 212   struct LineTable {
 213     LineTable();
 214
 215     /// Represents an invalid row
 216     const uint32_t UnknownRowIndex = UINT32_MAX;
 217
 218     void appendRow(const DWARFDebugLine::Row &R) { Rows.push_back(R); }
 219
 220     void appendSequence(const DWARFDebugLine::Sequence &S) {
 221       Sequences.push_back(S);
 222     }
 223
 224     /// Returns the index of the row with file/line info for a given address,
 225     /// or UnknownRowIndex if there is no such row.
 226     uint32_t lookupAddress(uint64_t Address) const;
 227
 228     bool lookupAddressRange(uint64_t Address, uint64_t Size,
 229                             std::vector<uint32_t> &Result) const;
 230
 231     bool hasFileAtIndex(uint64_t FileIndex) const;
 232
 233     /// Extracts filename by its index in filename table in prologue.
 234     /// Returns true on success.
 235     bool getFileNameByIndex(uint64_t FileIndex, const char *CompDir,
 236                             DILineInfoSpecifier::FileLineInfoKind Kind,
 237                             std::string &Result) const;
 238
 239     /// Fills the Result argument with the file and line information
 240     /// corresponding to Address. Returns true on success.
 241     bool getFileLineInfoForAddress(uint64_t Address, const char *CompDir,
 242                                    DILineInfoSpecifier::FileLineInfoKind Kind,
 243                                    DILineInfo &Result) const;
 244
 245     void dump(raw_ostream &OS, DIDumpOptions DumpOptions) const;
 246     void clear();
 247
 248     /// Parse prologue and all rows.
 249     Error parse(
 250         DWARFDataExtractor &DebugLineData, uint32_t *OffsetPtr,
 251         const DWARFContext &Ctx, const DWARFUnit *U,
 252         std::function<void(Error)> RecoverableErrorCallback,
 253         raw_ostream *OS = nullptr);
 254
 255     using RowVector = std::vector<Row>;
 256     using RowIter = RowVector::const_iterator;
 257     using SequenceVector = std::vector<Sequence>;
 258     using SequenceIter = SequenceVector::const_iterator;
 259
 260     struct Prologue Prologue;
 261     RowVector Rows;
 262     SequenceVector Sequences;
 263
 264   private:
 265     uint32_t findRowInSeq(const DWARFDebugLine::Sequence &Seq,
 266                           uint64_t Address) const;
 267     Optional<StringRef>
 268     getSourceByIndex(uint64_t FileIndex,
 269                      DILineInfoSpecifier::FileLineInfoKind Kind) const;
 270   };
 271
 272   const LineTable *getLineTable(uint32_t Offset) const;
 273   Expected<const LineTable *> getOrParseLineTable(
 274       DWARFDataExtractor &DebugLineData, uint32_t Offset,
 275       const DWARFContext &Ctx, const DWARFUnit *U,
 276       std::function<void(Error)> RecoverableErrorCallback);
 277
 278   /// Helper to allow for parsing of an entire .debug_line section in sequence.
 279   class SectionParser {
 280   public:
 281     using cu_range = DWARFUnitVector::iterator_range;
 282     using tu_range = DWARFUnitVector::iterator_range;
 283     using LineToUnitMap = std::map<uint64_t, DWARFUnit *>;
 284
 285     SectionParser(DWARFDataExtractor &Data, const DWARFContext &C, cu_range CUs,
 286                   tu_range TUs);
 287
 288     /// Get the next line table from the section. Report any issues via the
 289     /// callbacks.
 290     ///
 291     /// \param RecoverableErrorCallback - any issues that don't prevent further
 292     /// parsing of the table will be reported through this callback.
 293     /// \param UnrecoverableErrorCallback - any issues that prevent further
 294     /// parsing of the table will be reported through this callback.
 295     /// \param OS - if not null, the parser will print information about the
 296     /// table as it parses it.
 297     LineTable
 298     parseNext(
 299         function_ref<void(Error)> RecoverableErrorCallback,
 300         function_ref<void(Error)> UnrecoverableErrorCallback,
 301         raw_ostream *OS = nullptr);
 302
 303     /// Skip the current line table and go to the following line table (if
 304     /// present) immediately.
 305     ///
 306     /// \param ErrorCallback - report any prologue parsing issues via this
 307     /// callback.
 308     void skip(function_ref<void(Error)> ErrorCallback);
 309
 310     /// Indicates if the parser has parsed as much as possible.
 311     ///
 312     /// \note Certain problems with the line table structure might mean that
 313     /// parsing stops before the end of the section is reached.
 314     bool done() const { return Done; }
 315
 316     /// Get the offset the parser has reached.
 317     uint32_t getOffset() const { return Offset; }
 318
 319   private:
 320     DWARFUnit *prepareToParse(uint32_t Offset);
 321     void moveToNextTable(uint32_t OldOffset, const Prologue &P);
 322
 323     LineToUnitMap LineToUnit;
 324
 325     DWARFDataExtractor &DebugLineData;
 326     const DWARFContext &Context;
 327     uint32_t Offset = 0;
 328     bool Done = false;
 329   };
 330
 331 private:
 332   struct ParsingState {
 333     ParsingState(struct LineTable *LT);
 334
 335     void resetRowAndSequence();
 336     void appendRowToMatrix(uint32_t Offset);
 337
 338     /// Line table we're currently parsing.
 339     struct LineTable *LineTable;
 340     /// The row number that starts at zero for the prologue, and increases for
 341     /// each row added to the matrix.
 342     unsigned RowNumber = 0;
 343     struct Row Row;
 344     struct Sequence Sequence;
 345   };
 346
 347   using LineTableMapTy = std::map<uint32_t, LineTable>;
 348   using LineTableIter = LineTableMapTy::iterator;
 349   using LineTableConstIter = LineTableMapTy::const_iterator;
 350
 351   LineTableMapTy LineTableMap;
 352 };
 353
 354 } // end namespace llvm
 355
 356 #endif // LLVM_DEBUGINFO_DWARFDEBUGLINE_H