1 //===- llvm/Bitcode/BitcodeReader.h - Bitcode reader ------------*- 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 // This header defines interfaces to read LLVM bitcode files/streams.
10 //
11 //===----------------------------------------------------------------------===//
12 
13 #ifndef LLVM_BITCODE_BITCODEREADER_H
14 #define LLVM_BITCODE_BITCODEREADER_H
15 
16 #include "llvm/ADT/ArrayRef.h"
17 #include "llvm/ADT/StringRef.h"
18 #include "llvm/Bitstream/BitCodeEnums.h"
19 #include "llvm/IR/GlobalValue.h"
20 #include "llvm/Support/Endian.h"
21 #include "llvm/Support/Error.h"
22 #include "llvm/Support/ErrorOr.h"
23 #include "llvm/Support/MemoryBufferRef.h"
24 #include <cstdint>
25 #include <memory>
26 #include <optional>
27 #include <string>
28 #include <system_error>
29 #include <vector>
30 namespace llvm {
31 
32 class LLVMContext;
33 class Module;
34 class MemoryBuffer;
35 class Metadata;
36 class ModuleSummaryIndex;
37 class Type;
38 class Value;
39 
40 // Callback to override the data layout string of an imported bitcode module.
41 // The first argument is the target triple, the second argument the data layout
42 // string from the input, or a default string. It will be used if the callback
43 // returns std::nullopt.
44 typedef std::function<std::optional<std::string>(StringRef, StringRef)>
45     DataLayoutCallbackFuncTy;
46 
47 typedef std::function<Type *(unsigned)> GetTypeByIDTy;
48 
49 typedef std::function<unsigned(unsigned, unsigned)> GetContainedTypeIDTy;
50 
51 typedef std::function<void(Value *, unsigned, GetTypeByIDTy,
52                            GetContainedTypeIDTy)>
53     ValueTypeCallbackTy;
54 
55 typedef std::function<void(Metadata **, unsigned, GetTypeByIDTy,
56                            GetContainedTypeIDTy)>
57     MDTypeCallbackTy;
58 
59 // These functions are for converting Expected/Error values to
60 // ErrorOr/std::error_code for compatibility with legacy clients. FIXME:
61 // Remove these functions once no longer needed by the C and libLTO APIs.
62 
63 std::error_code errorToErrorCodeAndEmitErrors(LLVMContext &Ctx, Error Err);
64 
65 template <typename T>
expectedToErrorOrAndEmitErrors(LLVMContext & Ctx,Expected<T> Val)66 ErrorOr<T> expectedToErrorOrAndEmitErrors(LLVMContext &Ctx, Expected<T> Val) {
67   if (!Val)
68     return errorToErrorCodeAndEmitErrors(Ctx, Val.takeError());
69   return std::move(*Val);
70 }
71 
72 struct ParserCallbacks {
73   std::optional<DataLayoutCallbackFuncTy> DataLayout;
74   /// The ValueType callback is called for every function definition or
75   /// declaration and allows accessing the type information, also behind
76   /// pointers. This can be useful, when the opaque pointer upgrade cleans all
77   /// type information behind pointers.
78   /// The second argument to ValueTypeCallback is the type ID of the
79   /// function, the two passed functions can be used to extract type
80   /// information.
81   std::optional<ValueTypeCallbackTy> ValueType;
82   /// The MDType callback is called for every value in metadata.
83   std::optional<MDTypeCallbackTy> MDType;
84 
85   ParserCallbacks() = default;
ParserCallbacksParserCallbacks86   explicit ParserCallbacks(DataLayoutCallbackFuncTy DataLayout)
87       : DataLayout(DataLayout) {}
88 };
89 
90   struct BitcodeFileContents;
91 
92   /// Basic information extracted from a bitcode module to be used for LTO.
93   struct BitcodeLTOInfo {
94     bool IsThinLTO;
95     bool HasSummary;
96     bool EnableSplitLTOUnit;
97   };
98 
99   /// Represents a module in a bitcode file.
100   class BitcodeModule {
101     // This covers the identification (if present) and module blocks.
102     ArrayRef<uint8_t> Buffer;
103     StringRef ModuleIdentifier;
104 
105     // The string table used to interpret this module.
106     StringRef Strtab;
107 
108     // The bitstream location of the IDENTIFICATION_BLOCK.
109     uint64_t IdentificationBit;
110 
111     // The bitstream location of this module's MODULE_BLOCK.
112     uint64_t ModuleBit;
113 
BitcodeModule(ArrayRef<uint8_t> Buffer,StringRef ModuleIdentifier,uint64_t IdentificationBit,uint64_t ModuleBit)114     BitcodeModule(ArrayRef<uint8_t> Buffer, StringRef ModuleIdentifier,
115                   uint64_t IdentificationBit, uint64_t ModuleBit)
116         : Buffer(Buffer), ModuleIdentifier(ModuleIdentifier),
117           IdentificationBit(IdentificationBit), ModuleBit(ModuleBit) {}
118 
119     // Calls the ctor.
120     friend Expected<BitcodeFileContents>
121     getBitcodeFileContents(MemoryBufferRef Buffer);
122 
123     Expected<std::unique_ptr<Module>>
124     getModuleImpl(LLVMContext &Context, bool MaterializeAll,
125                   bool ShouldLazyLoadMetadata, bool IsImporting,
126                   ParserCallbacks Callbacks = {});
127 
128   public:
getBuffer()129     StringRef getBuffer() const {
130       return StringRef((const char *)Buffer.begin(), Buffer.size());
131     }
132 
getStrtab()133     StringRef getStrtab() const { return Strtab; }
134 
getModuleIdentifier()135     StringRef getModuleIdentifier() const { return ModuleIdentifier; }
136 
137     /// Read the bitcode module and prepare for lazy deserialization of function
138     /// bodies. If ShouldLazyLoadMetadata is true, lazily load metadata as well.
139     /// If IsImporting is true, this module is being parsed for ThinLTO
140     /// importing into another module.
141     Expected<std::unique_ptr<Module>>
142     getLazyModule(LLVMContext &Context, bool ShouldLazyLoadMetadata,
143                   bool IsImporting, ParserCallbacks Callbacks = {});
144 
145     /// Read the entire bitcode module and return it.
146     Expected<std::unique_ptr<Module>>
147     parseModule(LLVMContext &Context, ParserCallbacks Callbacks = {});
148 
149     /// Returns information about the module to be used for LTO: whether to
150     /// compile with ThinLTO, and whether it has a summary.
151     Expected<BitcodeLTOInfo> getLTOInfo();
152 
153     /// Parse the specified bitcode buffer, returning the module summary index.
154     Expected<std::unique_ptr<ModuleSummaryIndex>> getSummary();
155 
156     /// Parse the specified bitcode buffer and merge its module summary index
157     /// into CombinedIndex.
158     Error
159     readSummary(ModuleSummaryIndex &CombinedIndex, StringRef ModulePath,
160                 uint64_t ModuleId,
161                 std::function<bool(GlobalValue::GUID)> IsPrevailing = nullptr);
162   };
163 
164   struct BitcodeFileContents {
165     std::vector<BitcodeModule> Mods;
166     StringRef Symtab, StrtabForSymtab;
167   };
168 
169   /// Returns the contents of a bitcode file. This includes the raw contents of
170   /// the symbol table embedded in the bitcode file. Clients which require a
171   /// symbol table should prefer to use irsymtab::read instead of this function
172   /// because it creates a reader for the irsymtab and handles upgrading bitcode
173   /// files without a symbol table or with an old symbol table.
174   Expected<BitcodeFileContents> getBitcodeFileContents(MemoryBufferRef Buffer);
175 
176   /// Returns a list of modules in the specified bitcode buffer.
177   Expected<std::vector<BitcodeModule>>
178   getBitcodeModuleList(MemoryBufferRef Buffer);
179 
180   /// Read the header of the specified bitcode buffer and prepare for lazy
181   /// deserialization of function bodies. If ShouldLazyLoadMetadata is true,
182   /// lazily load metadata as well. If IsImporting is true, this module is
183   /// being parsed for ThinLTO importing into another module.
184   Expected<std::unique_ptr<Module>>
185   getLazyBitcodeModule(MemoryBufferRef Buffer, LLVMContext &Context,
186                        bool ShouldLazyLoadMetadata = false,
187                        bool IsImporting = false,
188                        ParserCallbacks Callbacks = {});
189 
190   /// Like getLazyBitcodeModule, except that the module takes ownership of
191   /// the memory buffer if successful. If successful, this moves Buffer. On
192   /// error, this *does not* move Buffer. If IsImporting is true, this module is
193   /// being parsed for ThinLTO importing into another module.
194   Expected<std::unique_ptr<Module>> getOwningLazyBitcodeModule(
195       std::unique_ptr<MemoryBuffer> &&Buffer, LLVMContext &Context,
196       bool ShouldLazyLoadMetadata = false, bool IsImporting = false,
197       ParserCallbacks Callbacks = {});
198 
199   /// Read the header of the specified bitcode buffer and extract just the
200   /// triple information. If successful, this returns a string. On error, this
201   /// returns "".
202   Expected<std::string> getBitcodeTargetTriple(MemoryBufferRef Buffer);
203 
204   /// Return true if \p Buffer contains a bitcode file with ObjC code (category
205   /// or class) in it.
206   Expected<bool> isBitcodeContainingObjCCategory(MemoryBufferRef Buffer);
207 
208   /// Read the header of the specified bitcode buffer and extract just the
209   /// producer string information. If successful, this returns a string. On
210   /// error, this returns "".
211   Expected<std::string> getBitcodeProducerString(MemoryBufferRef Buffer);
212 
213   /// Read the specified bitcode file, returning the module.
214   Expected<std::unique_ptr<Module>>
215   parseBitcodeFile(MemoryBufferRef Buffer, LLVMContext &Context,
216                    ParserCallbacks Callbacks = {});
217 
218   /// Returns LTO information for the specified bitcode file.
219   Expected<BitcodeLTOInfo> getBitcodeLTOInfo(MemoryBufferRef Buffer);
220 
221   /// Parse the specified bitcode buffer, returning the module summary index.
222   Expected<std::unique_ptr<ModuleSummaryIndex>>
223   getModuleSummaryIndex(MemoryBufferRef Buffer);
224 
225   /// Parse the specified bitcode buffer and merge the index into CombinedIndex.
226   Error readModuleSummaryIndex(MemoryBufferRef Buffer,
227                                ModuleSummaryIndex &CombinedIndex,
228                                uint64_t ModuleId);
229 
230   /// Parse the module summary index out of an IR file and return the module
231   /// summary index object if found, or an empty summary if not. If Path refers
232   /// to an empty file and IgnoreEmptyThinLTOIndexFile is true, then
233   /// this function will return nullptr.
234   Expected<std::unique_ptr<ModuleSummaryIndex>>
235   getModuleSummaryIndexForFile(StringRef Path,
236                                bool IgnoreEmptyThinLTOIndexFile = false);
237 
238   /// isBitcodeWrapper - Return true if the given bytes are the magic bytes
239   /// for an LLVM IR bitcode wrapper.
isBitcodeWrapper(const unsigned char * BufPtr,const unsigned char * BufEnd)240   inline bool isBitcodeWrapper(const unsigned char *BufPtr,
241                                const unsigned char *BufEnd) {
242     // See if you can find the hidden message in the magic bytes :-).
243     // (Hint: it's a little-endian encoding.)
244     return BufPtr != BufEnd &&
245            BufPtr[0] == 0xDE &&
246            BufPtr[1] == 0xC0 &&
247            BufPtr[2] == 0x17 &&
248            BufPtr[3] == 0x0B;
249   }
250 
251   /// isRawBitcode - Return true if the given bytes are the magic bytes for
252   /// raw LLVM IR bitcode (without a wrapper).
isRawBitcode(const unsigned char * BufPtr,const unsigned char * BufEnd)253   inline bool isRawBitcode(const unsigned char *BufPtr,
254                            const unsigned char *BufEnd) {
255     // These bytes sort of have a hidden message, but it's not in
256     // little-endian this time, and it's a little redundant.
257     return BufPtr != BufEnd &&
258            BufPtr[0] == 'B' &&
259            BufPtr[1] == 'C' &&
260            BufPtr[2] == 0xc0 &&
261            BufPtr[3] == 0xde;
262   }
263 
264   /// isBitcode - Return true if the given bytes are the magic bytes for
265   /// LLVM IR bitcode, either with or without a wrapper.
isBitcode(const unsigned char * BufPtr,const unsigned char * BufEnd)266   inline bool isBitcode(const unsigned char *BufPtr,
267                         const unsigned char *BufEnd) {
268     return isBitcodeWrapper(BufPtr, BufEnd) ||
269            isRawBitcode(BufPtr, BufEnd);
270   }
271 
272   /// SkipBitcodeWrapperHeader - Some systems wrap bc files with a special
273   /// header for padding or other reasons.  The format of this header is:
274   ///
275   /// struct bc_header {
276   ///   uint32_t Magic;         // 0x0B17C0DE
277   ///   uint32_t Version;       // Version, currently always 0.
278   ///   uint32_t BitcodeOffset; // Offset to traditional bitcode file.
279   ///   uint32_t BitcodeSize;   // Size of traditional bitcode file.
280   ///   ... potentially other gunk ...
281   /// };
282   ///
283   /// This function is called when we find a file with a matching magic number.
284   /// In this case, skip down to the subsection of the file that is actually a
285   /// BC file.
286   /// If 'VerifyBufferSize' is true, check that the buffer is large enough to
287   /// contain the whole bitcode file.
SkipBitcodeWrapperHeader(const unsigned char * & BufPtr,const unsigned char * & BufEnd,bool VerifyBufferSize)288   inline bool SkipBitcodeWrapperHeader(const unsigned char *&BufPtr,
289                                        const unsigned char *&BufEnd,
290                                        bool VerifyBufferSize) {
291     // Must contain the offset and size field!
292     if (unsigned(BufEnd - BufPtr) < BWH_SizeField + 4)
293       return true;
294 
295     unsigned Offset = support::endian::read32le(&BufPtr[BWH_OffsetField]);
296     unsigned Size = support::endian::read32le(&BufPtr[BWH_SizeField]);
297     uint64_t BitcodeOffsetEnd = (uint64_t)Offset + (uint64_t)Size;
298 
299     // Verify that Offset+Size fits in the file.
300     if (VerifyBufferSize && BitcodeOffsetEnd > uint64_t(BufEnd-BufPtr))
301       return true;
302     BufPtr += Offset;
303     BufEnd = BufPtr+Size;
304     return false;
305   }
306 
307   APInt readWideAPInt(ArrayRef<uint64_t> Vals, unsigned TypeBits);
308 
309   const std::error_category &BitcodeErrorCategory();
310   enum class BitcodeError { CorruptedBitcode = 1 };
make_error_code(BitcodeError E)311   inline std::error_code make_error_code(BitcodeError E) {
312     return std::error_code(static_cast<int>(E), BitcodeErrorCategory());
313   }
314 
315 } // end namespace llvm
316 
317 namespace std {
318 
319 template <> struct is_error_code_enum<llvm::BitcodeError> : std::true_type {};
320 
321 } // end namespace std
322 
323 #endif // LLVM_BITCODE_BITCODEREADER_H
324