1 //===-- DWARFUnit.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 LLDB_SOURCE_PLUGINS_SYMBOLFILE_DWARF_DWARFUNIT_H
10 #define LLDB_SOURCE_PLUGINS_SYMBOLFILE_DWARF_DWARFUNIT_H
11 
12 #include "DWARFDIE.h"
13 #include "DWARFDebugInfoEntry.h"
14 #include "lldb/Utility/XcodeSDK.h"
15 #include "lldb/lldb-enumerations.h"
16 #include "llvm/DebugInfo/DWARF/DWARFDebugRnglists.h"
17 #include "llvm/Support/RWMutex.h"
18 #include <atomic>
19 #include <optional>
20 
21 class DWARFUnit;
22 class DWARFCompileUnit;
23 class NameToDIE;
24 class SymbolFileDWARF;
25 class SymbolFileDWARFDwo;
26 
27 typedef std::shared_ptr<DWARFUnit> DWARFUnitSP;
28 
29 enum DWARFProducer {
30   eProducerInvalid = 0,
31   eProducerClang,
32   eProducerGCC,
33   eProducerLLVMGCC,
34   eProducerSwift,
35   eProducerOther
36 };
37 
38 /// Base class describing the header of any kind of "unit."  Some information
39 /// is specific to certain unit types.  We separate this class out so we can
40 /// parse the header before deciding what specific kind of unit to construct.
41 class DWARFUnitHeader {
42   dw_offset_t m_offset = 0;
43   dw_offset_t m_length = 0;
44   uint16_t m_version = 0;
45   dw_offset_t m_abbr_offset = 0;
46 
47   const llvm::DWARFUnitIndex::Entry *m_index_entry = nullptr;
48 
49   uint8_t m_unit_type = 0;
50   uint8_t m_addr_size = 0;
51 
52   uint64_t m_type_hash = 0;
53   uint32_t m_type_offset = 0;
54 
55   std::optional<uint64_t> m_dwo_id;
56 
57   DWARFUnitHeader() = default;
58 
59 public:
60   dw_offset_t GetOffset() const { return m_offset; }
61   uint16_t GetVersion() const { return m_version; }
62   uint16_t GetAddressByteSize() const { return m_addr_size; }
63   dw_offset_t GetLength() const { return m_length; }
64   dw_offset_t GetAbbrOffset() const { return m_abbr_offset; }
65   uint8_t GetUnitType() const { return m_unit_type; }
66   const llvm::DWARFUnitIndex::Entry *GetIndexEntry() const {
67     return m_index_entry;
68   }
69   uint64_t GetTypeHash() const { return m_type_hash; }
70   dw_offset_t GetTypeOffset() const { return m_type_offset; }
71   std::optional<uint64_t> GetDWOId() const { return m_dwo_id; }
72   bool IsTypeUnit() const {
73     return m_unit_type == llvm::dwarf::DW_UT_type ||
74            m_unit_type == llvm::dwarf::DW_UT_split_type;
75   }
76   uint32_t GetNextUnitOffset() const { return m_offset + m_length + 4; }
77 
78   static llvm::Expected<DWARFUnitHeader>
79   extract(const lldb_private::DWARFDataExtractor &data, DIERef::Section section,
80           lldb_private::DWARFContext &dwarf_context,
81           lldb::offset_t *offset_ptr);
82 };
83 
84 class DWARFUnit : public lldb_private::UserID {
85   using die_iterator_range =
86       llvm::iterator_range<DWARFDebugInfoEntry::collection::iterator>;
87 
88 public:
89   static llvm::Expected<DWARFUnitSP>
90   extract(SymbolFileDWARF &dwarf2Data, lldb::user_id_t uid,
91           const lldb_private::DWARFDataExtractor &debug_info,
92           DIERef::Section section, lldb::offset_t *offset_ptr);
93   virtual ~DWARFUnit();
94 
95   bool IsDWOUnit() { return m_is_dwo; }
96   std::optional<uint64_t> GetDWOId();
97 
98   void ExtractUnitDIEIfNeeded();
99   void ExtractUnitDIENoDwoIfNeeded();
100   void ExtractDIEsIfNeeded();
101 
102   class ScopedExtractDIEs {
103     DWARFUnit *m_cu;
104   public:
105     bool m_clear_dies = false;
106     ScopedExtractDIEs(DWARFUnit &cu);
107     ~ScopedExtractDIEs();
108     ScopedExtractDIEs(const ScopedExtractDIEs &) = delete;
109     const ScopedExtractDIEs &operator=(const ScopedExtractDIEs &) = delete;
110     ScopedExtractDIEs(ScopedExtractDIEs &&rhs);
111     ScopedExtractDIEs &operator=(ScopedExtractDIEs &&rhs);
112   };
113   ScopedExtractDIEs ExtractDIEsScoped();
114 
115   bool Verify(lldb_private::Stream *s) const;
116   virtual void Dump(lldb_private::Stream *s) const = 0;
117   /// Get the data that contains the DIE information for this unit.
118   ///
119   /// This will return the correct bytes that contain the data for
120   /// this DWARFUnit. It could be .debug_info or .debug_types
121   /// depending on where the data for this unit originates.
122   ///
123   /// \return
124   ///   The correct data for the DIE information in this unit.
125   const lldb_private::DWARFDataExtractor &GetData() const;
126 
127   /// Get the size in bytes of the unit header.
128   ///
129   /// \return
130   ///     Byte size of the unit header
131   uint32_t GetHeaderByteSize() const;
132 
133   // Offset of the initial length field.
134   dw_offset_t GetOffset() const { return m_header.GetOffset(); }
135   /// Get the size in bytes of the length field in the header.
136   ///
137   /// In DWARF32 this is just 4 bytes
138   ///
139   /// \return
140   ///     Byte size of the compile unit header length field
141   size_t GetLengthByteSize() const { return 4; }
142 
143   bool ContainsDIEOffset(dw_offset_t die_offset) const {
144     return die_offset >= GetFirstDIEOffset() &&
145            die_offset < GetNextUnitOffset();
146   }
147   dw_offset_t GetFirstDIEOffset() const {
148     return GetOffset() + GetHeaderByteSize();
149   }
150   dw_offset_t GetNextUnitOffset() const { return m_header.GetNextUnitOffset(); }
151   // Size of the CU data (without initial length and without header).
152   size_t GetDebugInfoSize() const;
153   // Size of the CU data incl. header but without initial length.
154   uint32_t GetLength() const { return m_header.GetLength(); }
155   uint16_t GetVersion() const { return m_header.GetVersion(); }
156   const DWARFAbbreviationDeclarationSet *GetAbbreviations() const;
157   dw_offset_t GetAbbrevOffset() const;
158   uint8_t GetAddressByteSize() const { return m_header.GetAddressByteSize(); }
159   dw_addr_t GetAddrBase() const { return m_addr_base.value_or(0); }
160   dw_addr_t GetBaseAddress() const { return m_base_addr; }
161   dw_offset_t GetLineTableOffset();
162   dw_addr_t GetRangesBase() const { return m_ranges_base; }
163   dw_addr_t GetStrOffsetsBase() const { return m_str_offsets_base; }
164   void SetAddrBase(dw_addr_t addr_base);
165   void SetLoclistsBase(dw_addr_t loclists_base);
166   void SetRangesBase(dw_addr_t ranges_base);
167   void SetStrOffsetsBase(dw_offset_t str_offsets_base);
168   virtual void BuildAddressRangeTable(DWARFDebugAranges *debug_aranges) = 0;
169 
170   dw_addr_t ReadAddressFromDebugAddrSection(uint32_t index) const;
171 
172   lldb::ByteOrder GetByteOrder() const;
173 
174   const DWARFDebugAranges &GetFunctionAranges();
175 
176   void SetBaseAddress(dw_addr_t base_addr);
177 
178   DWARFBaseDIE GetUnitDIEOnly() { return {this, GetUnitDIEPtrOnly()}; }
179 
180   DWARFDIE DIE() { return DWARFDIE(this, DIEPtr()); }
181 
182   DWARFDIE GetDIE(dw_offset_t die_offset);
183 
184   DWARFUnit &GetNonSkeletonUnit();
185 
186   static uint8_t GetAddressByteSize(const DWARFUnit *cu);
187 
188   static uint8_t GetDefaultAddressSize();
189 
190   void *GetUserData() const;
191 
192   void SetUserData(void *d);
193 
194   bool Supports_DW_AT_APPLE_objc_complete_type();
195 
196   bool DW_AT_decl_file_attributes_are_invalid();
197 
198   bool Supports_unnamed_objc_bitfields();
199 
200   SymbolFileDWARF &GetSymbolFileDWARF() const { return m_dwarf; }
201 
202   DWARFProducer GetProducer();
203 
204   llvm::VersionTuple GetProducerVersion();
205 
206   uint64_t GetDWARFLanguageType();
207 
208   bool GetIsOptimized();
209 
210   const lldb_private::FileSpec &GetCompilationDirectory();
211   const lldb_private::FileSpec &GetAbsolutePath();
212   lldb_private::FileSpec GetFile(size_t file_idx);
213   lldb_private::FileSpec::Style GetPathStyle();
214 
215   SymbolFileDWARFDwo *GetDwoSymbolFile();
216 
217   die_iterator_range dies() {
218     ExtractDIEsIfNeeded();
219     return die_iterator_range(m_die_array.begin(), m_die_array.end());
220   }
221 
222   DIERef::Section GetDebugSection() const { return m_section; }
223 
224   uint8_t GetUnitType() const { return m_header.GetUnitType(); }
225   bool IsTypeUnit() const { return m_header.IsTypeUnit(); }
226   /// Note that this check only works for DWARF5+.
227   bool IsSkeletonUnit() const { return GetUnitType() == llvm::dwarf::DW_UT_skeleton; }
228 
229   std::optional<uint64_t> GetStringOffsetSectionItem(uint32_t index) const;
230 
231   /// Return a list of address ranges resulting from a (possibly encoded)
232   /// range list starting at a given offset in the appropriate ranges section.
233   llvm::Expected<DWARFRangeList> FindRnglistFromOffset(dw_offset_t offset);
234 
235   /// Return a list of address ranges retrieved from an encoded range
236   /// list whose offset is found via a table lookup given an index (DWARF v5
237   /// and later).
238   llvm::Expected<DWARFRangeList> FindRnglistFromIndex(uint32_t index);
239 
240   /// Return a rangelist's offset based on an index. The index designates
241   /// an entry in the rangelist table's offset array and is supplied by
242   /// DW_FORM_rnglistx.
243   llvm::Expected<uint64_t> GetRnglistOffset(uint32_t Index);
244 
245   std::optional<uint64_t> GetLoclistOffset(uint32_t Index) {
246     if (!m_loclist_table_header)
247       return std::nullopt;
248 
249     std::optional<uint64_t> Offset = m_loclist_table_header->getOffsetEntry(
250         m_dwarf.GetDWARFContext().getOrLoadLocListsData().GetAsLLVM(), Index);
251     if (!Offset)
252       return std::nullopt;
253     return *Offset + m_loclists_base;
254   }
255 
256   /// Return the location table for parsing the given location list data. The
257   /// format is chosen according to the unit type. Never returns null.
258   std::unique_ptr<llvm::DWARFLocationTable>
259   GetLocationTable(const lldb_private::DataExtractor &data) const;
260 
261   lldb_private::DWARFDataExtractor GetLocationData() const;
262 
263   /// Returns true if any DIEs in the unit match any DW_TAG values in \a tags.
264   ///
265   /// \param[in] tags
266   ///   An array of dw_tag_t values to check all abbrevitions for.
267   ///
268   /// \returns
269   ///   True if any DIEs match any tag in \a tags, false otherwise.
270   bool HasAny(llvm::ArrayRef<dw_tag_t> tags);
271 
272 
273   /// Get the fission .dwo file specific error for this compile unit.
274   ///
275   /// The skeleton compile unit only can have a DWO error. Any other type
276   /// of DWARFUnit will not have a valid DWO error.
277   ///
278   /// \returns
279   ///   A valid DWO error if there is a problem with anything in the
280   ///   locating or parsing inforamtion in the .dwo file
281   const lldb_private::Status &GetDwoError() const { return m_dwo_error; }
282 
283   /// Set the fission .dwo file specific error for this compile unit.
284   ///
285   /// This helps tracks issues that arise when trying to locate or parse a
286   /// .dwo file. Things like a missing .dwo file, DWO ID mismatch, and other
287   /// .dwo errors can be stored in each compile unit so the issues can be
288   /// communicated to the user.
289   void SetDwoError(const lldb_private::Status &error) { m_dwo_error = error; }
290 
291 protected:
292   DWARFUnit(SymbolFileDWARF &dwarf, lldb::user_id_t uid,
293             const DWARFUnitHeader &header,
294             const DWARFAbbreviationDeclarationSet &abbrevs,
295             DIERef::Section section, bool is_dwo);
296 
297   llvm::Error ExtractHeader(SymbolFileDWARF &dwarf,
298                             const lldb_private::DWARFDataExtractor &data,
299                             lldb::offset_t *offset_ptr);
300 
301   // Get the DWARF unit DWARF debug information entry. Parse the single DIE
302   // if needed.
303   const DWARFDebugInfoEntry *GetUnitDIEPtrOnly() {
304     ExtractUnitDIENoDwoIfNeeded();
305     // m_first_die_mutex is not required as m_first_die is never cleared.
306     if (!m_first_die)
307       return nullptr;
308     return &m_first_die;
309   }
310 
311   // Get all DWARF debug informration entries. Parse all DIEs if needed.
312   const DWARFDebugInfoEntry *DIEPtr() {
313     ExtractDIEsIfNeeded();
314     if (m_die_array.empty())
315       return nullptr;
316     return &m_die_array[0];
317   }
318 
319   const std::optional<llvm::DWARFDebugRnglistTable> &GetRnglistTable();
320 
321   lldb_private::DWARFDataExtractor GetRnglistData() const;
322 
323   SymbolFileDWARF &m_dwarf;
324   std::shared_ptr<DWARFUnit> m_dwo;
325   DWARFUnitHeader m_header;
326   const DWARFAbbreviationDeclarationSet *m_abbrevs = nullptr;
327   void *m_user_data = nullptr;
328   // The compile unit debug information entry item
329   DWARFDebugInfoEntry::collection m_die_array;
330   mutable llvm::sys::RWMutex m_die_array_mutex;
331   // It is used for tracking of ScopedExtractDIEs instances.
332   mutable llvm::sys::RWMutex m_die_array_scoped_mutex;
333   // ScopedExtractDIEs instances should not call ClearDIEsRWLocked()
334   // as someone called ExtractDIEsIfNeeded().
335   std::atomic<bool> m_cancel_scopes;
336   // GetUnitDIEPtrOnly() needs to return pointer to the first DIE.
337   // But the first element of m_die_array after ExtractUnitDIEIfNeeded()
338   // would possibly move in memory after later ExtractDIEsIfNeeded().
339   DWARFDebugInfoEntry m_first_die;
340   llvm::sys::RWMutex m_first_die_mutex;
341   // A table similar to the .debug_aranges table, but this one points to the
342   // exact DW_TAG_subprogram DIEs
343   std::unique_ptr<DWARFDebugAranges> m_func_aranges_up;
344   dw_addr_t m_base_addr = 0;
345   DWARFProducer m_producer = eProducerInvalid;
346   llvm::VersionTuple m_producer_version;
347   std::optional<uint64_t> m_language_type;
348   lldb_private::LazyBool m_is_optimized = lldb_private::eLazyBoolCalculate;
349   std::optional<lldb_private::FileSpec> m_comp_dir;
350   std::optional<lldb_private::FileSpec> m_file_spec;
351   std::optional<dw_addr_t> m_addr_base;  ///< Value of DW_AT_addr_base.
352   dw_addr_t m_loclists_base = 0;         ///< Value of DW_AT_loclists_base.
353   dw_addr_t m_ranges_base = 0;           ///< Value of DW_AT_rnglists_base.
354   std::optional<uint64_t> m_gnu_addr_base;
355   std::optional<uint64_t> m_gnu_ranges_base;
356 
357   /// Value of DW_AT_stmt_list.
358   dw_offset_t m_line_table_offset = DW_INVALID_OFFSET;
359 
360   dw_offset_t m_str_offsets_base = 0; // Value of DW_AT_str_offsets_base.
361 
362   std::optional<llvm::DWARFDebugRnglistTable> m_rnglist_table;
363   bool m_rnglist_table_done = false;
364   std::optional<llvm::DWARFListTableHeader> m_loclist_table_header;
365 
366   const DIERef::Section m_section;
367   bool m_is_dwo;
368   bool m_has_parsed_non_skeleton_unit;
369   /// Value of DW_AT_GNU_dwo_id (v4) or dwo_id from CU header (v5).
370   std::optional<uint64_t> m_dwo_id;
371   /// If we get an error when trying to load a .dwo file, save that error here.
372   /// Errors include .dwo/.dwp file not found, or the .dwp/.dwp file was found
373   /// but DWO ID doesn't match, etc.
374   lldb_private::Status m_dwo_error;
375 
376 private:
377   void ParseProducerInfo();
378   void ExtractDIEsRWLocked();
379   void ClearDIEsRWLocked();
380 
381   void AddUnitDIE(const DWARFDebugInfoEntry &cu_die);
382   void SetDwoStrOffsetsBase();
383 
384   void ComputeCompDirAndGuessPathStyle();
385   void ComputeAbsolutePath();
386 
387   DWARFUnit(const DWARFUnit &) = delete;
388   const DWARFUnit &operator=(const DWARFUnit &) = delete;
389 };
390 
391 #endif // LLDB_SOURCE_PLUGINS_SYMBOLFILE_DWARF_DWARFUNIT_H
392