1 //===- llvm/Bitcode/BitcodeReader.h - Bitcode reader ------------*- C++ -*-===//
2 //
3 //                     The LLVM Compiler Infrastructure
4 //
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
7 //
8 //===----------------------------------------------------------------------===//
9 //
10 // This header defines interfaces to read LLVM bitcode files/streams.
11 //
12 //===----------------------------------------------------------------------===//
13 
14 #ifndef LLVM_BITCODE_BITCODEREADER_H
15 #define LLVM_BITCODE_BITCODEREADER_H
16 
17 #include "llvm/ADT/ArrayRef.h"
18 #include "llvm/ADT/StringRef.h"
19 #include "llvm/Bitcode/BitCodes.h"
20 #include "llvm/IR/ModuleSummaryIndex.h"
21 #include "llvm/Support/Endian.h"
22 #include "llvm/Support/Error.h"
23 #include "llvm/Support/ErrorOr.h"
24 #include "llvm/Support/MemoryBuffer.h"
25 #include <cstdint>
26 #include <memory>
27 #include <string>
28 #include <system_error>
29 #include <vector>
30 namespace llvm {
31 
32 class LLVMContext;
33 class Module;
34 
35   // These functions are for converting Expected/Error values to
36   // ErrorOr/std::error_code for compatibility with legacy clients. FIXME:
37   // Remove these functions once no longer needed by the C and libLTO APIs.
38 
39   std::error_code errorToErrorCodeAndEmitErrors(LLVMContext &Ctx, Error Err);
40 
41   template <typename T>
expectedToErrorOrAndEmitErrors(LLVMContext & Ctx,Expected<T> Val)42   ErrorOr<T> expectedToErrorOrAndEmitErrors(LLVMContext &Ctx, Expected<T> Val) {
43     if (!Val)
44       return errorToErrorCodeAndEmitErrors(Ctx, Val.takeError());
45     return std::move(*Val);
46   }
47 
48   struct BitcodeFileContents;
49 
50   /// Basic information extracted from a bitcode module to be used for LTO.
51   struct BitcodeLTOInfo {
52     bool IsThinLTO;
53     bool HasSummary;
54     bool EnableSplitLTOUnit;
55   };
56 
57   /// Represents a module in a bitcode file.
58   class BitcodeModule {
59     // This covers the identification (if present) and module blocks.
60     ArrayRef<uint8_t> Buffer;
61     StringRef ModuleIdentifier;
62 
63     // The string table used to interpret this module.
64     StringRef Strtab;
65 
66     // The bitstream location of the IDENTIFICATION_BLOCK.
67     uint64_t IdentificationBit;
68 
69     // The bitstream location of this module's MODULE_BLOCK.
70     uint64_t ModuleBit;
71 
BitcodeModule(ArrayRef<uint8_t> Buffer,StringRef ModuleIdentifier,uint64_t IdentificationBit,uint64_t ModuleBit)72     BitcodeModule(ArrayRef<uint8_t> Buffer, StringRef ModuleIdentifier,
73                   uint64_t IdentificationBit, uint64_t ModuleBit)
74         : Buffer(Buffer), ModuleIdentifier(ModuleIdentifier),
75           IdentificationBit(IdentificationBit), ModuleBit(ModuleBit) {}
76 
77     // Calls the ctor.
78     friend Expected<BitcodeFileContents>
79     getBitcodeFileContents(MemoryBufferRef Buffer);
80 
81     Expected<std::unique_ptr<Module>> getModuleImpl(LLVMContext &Context,
82                                                     bool MaterializeAll,
83                                                     bool ShouldLazyLoadMetadata,
84                                                     bool IsImporting);
85 
86   public:
getBuffer()87     StringRef getBuffer() const {
88       return StringRef((const char *)Buffer.begin(), Buffer.size());
89     }
90 
getStrtab()91     StringRef getStrtab() const { return Strtab; }
92 
getModuleIdentifier()93     StringRef getModuleIdentifier() const { return ModuleIdentifier; }
94 
95     /// Read the bitcode module and prepare for lazy deserialization of function
96     /// bodies. If ShouldLazyLoadMetadata is true, lazily load metadata as well.
97     /// If IsImporting is true, this module is being parsed for ThinLTO
98     /// importing into another module.
99     Expected<std::unique_ptr<Module>> getLazyModule(LLVMContext &Context,
100                                                     bool ShouldLazyLoadMetadata,
101                                                     bool IsImporting);
102 
103     /// Read the entire bitcode module and return it.
104     Expected<std::unique_ptr<Module>> parseModule(LLVMContext &Context);
105 
106     /// Returns information about the module to be used for LTO: whether to
107     /// compile with ThinLTO, and whether it has a summary.
108     Expected<BitcodeLTOInfo> getLTOInfo();
109 
110     /// Parse the specified bitcode buffer, returning the module summary index.
111     Expected<std::unique_ptr<ModuleSummaryIndex>> getSummary();
112 
113     /// Parse the specified bitcode buffer and merge its module summary index
114     /// into CombinedIndex.
115     Error readSummary(ModuleSummaryIndex &CombinedIndex, StringRef ModulePath,
116                       uint64_t ModuleId);
117   };
118 
119   struct BitcodeFileContents {
120     std::vector<BitcodeModule> Mods;
121     StringRef Symtab, StrtabForSymtab;
122   };
123 
124   /// Returns the contents of a bitcode file. This includes the raw contents of
125   /// the symbol table embedded in the bitcode file. Clients which require a
126   /// symbol table should prefer to use irsymtab::read instead of this function
127   /// because it creates a reader for the irsymtab and handles upgrading bitcode
128   /// files without a symbol table or with an old symbol table.
129   Expected<BitcodeFileContents> getBitcodeFileContents(MemoryBufferRef Buffer);
130 
131   /// Returns a list of modules in the specified bitcode buffer.
132   Expected<std::vector<BitcodeModule>>
133   getBitcodeModuleList(MemoryBufferRef Buffer);
134 
135   /// Read the header of the specified bitcode buffer and prepare for lazy
136   /// deserialization of function bodies. If ShouldLazyLoadMetadata is true,
137   /// lazily load metadata as well. If IsImporting is true, this module is
138   /// being parsed for ThinLTO importing into another module.
139   Expected<std::unique_ptr<Module>>
140   getLazyBitcodeModule(MemoryBufferRef Buffer, LLVMContext &Context,
141                        bool ShouldLazyLoadMetadata = false,
142                        bool IsImporting = false);
143 
144   /// Like getLazyBitcodeModule, except that the module takes ownership of
145   /// the memory buffer if successful. If successful, this moves Buffer. On
146   /// error, this *does not* move Buffer. If IsImporting is true, this module is
147   /// being parsed for ThinLTO importing into another module.
148   Expected<std::unique_ptr<Module>> getOwningLazyBitcodeModule(
149       std::unique_ptr<MemoryBuffer> &&Buffer, LLVMContext &Context,
150       bool ShouldLazyLoadMetadata = false, bool IsImporting = false);
151 
152   /// Read the header of the specified bitcode buffer and extract just the
153   /// triple information. If successful, this returns a string. On error, this
154   /// returns "".
155   Expected<std::string> getBitcodeTargetTriple(MemoryBufferRef Buffer);
156 
157   /// Return true if \p Buffer contains a bitcode file with ObjC code (category
158   /// or class) in it.
159   Expected<bool> isBitcodeContainingObjCCategory(MemoryBufferRef Buffer);
160 
161   /// Read the header of the specified bitcode buffer and extract just the
162   /// producer string information. If successful, this returns a string. On
163   /// error, this returns "".
164   Expected<std::string> getBitcodeProducerString(MemoryBufferRef Buffer);
165 
166   /// Read the specified bitcode file, returning the module.
167   Expected<std::unique_ptr<Module>> parseBitcodeFile(MemoryBufferRef Buffer,
168                                                      LLVMContext &Context);
169 
170   /// Returns LTO information for the specified bitcode file.
171   Expected<BitcodeLTOInfo> getBitcodeLTOInfo(MemoryBufferRef Buffer);
172 
173   /// Parse the specified bitcode buffer, returning the module summary index.
174   Expected<std::unique_ptr<ModuleSummaryIndex>>
175   getModuleSummaryIndex(MemoryBufferRef Buffer);
176 
177   /// Parse the specified bitcode buffer and merge the index into CombinedIndex.
178   Error readModuleSummaryIndex(MemoryBufferRef Buffer,
179                                ModuleSummaryIndex &CombinedIndex,
180                                uint64_t ModuleId);
181 
182   /// Parse the module summary index out of an IR file and return the module
183   /// summary index object if found, or an empty summary if not. If Path refers
184   /// to an empty file and IgnoreEmptyThinLTOIndexFile is true, then
185   /// this function will return nullptr.
186   Expected<std::unique_ptr<ModuleSummaryIndex>>
187   getModuleSummaryIndexForFile(StringRef Path,
188                                bool IgnoreEmptyThinLTOIndexFile = false);
189 
190   /// isBitcodeWrapper - Return true if the given bytes are the magic bytes
191   /// for an LLVM IR bitcode wrapper.
isBitcodeWrapper(const unsigned char * BufPtr,const unsigned char * BufEnd)192   inline bool isBitcodeWrapper(const unsigned char *BufPtr,
193                                const unsigned char *BufEnd) {
194     // See if you can find the hidden message in the magic bytes :-).
195     // (Hint: it's a little-endian encoding.)
196     return BufPtr != BufEnd &&
197            BufPtr[0] == 0xDE &&
198            BufPtr[1] == 0xC0 &&
199            BufPtr[2] == 0x17 &&
200            BufPtr[3] == 0x0B;
201   }
202 
203   /// isRawBitcode - Return true if the given bytes are the magic bytes for
204   /// raw LLVM IR bitcode (without a wrapper).
isRawBitcode(const unsigned char * BufPtr,const unsigned char * BufEnd)205   inline bool isRawBitcode(const unsigned char *BufPtr,
206                            const unsigned char *BufEnd) {
207     // These bytes sort of have a hidden message, but it's not in
208     // little-endian this time, and it's a little redundant.
209     return BufPtr != BufEnd &&
210            BufPtr[0] == 'B' &&
211            BufPtr[1] == 'C' &&
212            BufPtr[2] == 0xc0 &&
213            BufPtr[3] == 0xde;
214   }
215 
216   /// isBitcode - Return true if the given bytes are the magic bytes for
217   /// LLVM IR bitcode, either with or without a wrapper.
isBitcode(const unsigned char * BufPtr,const unsigned char * BufEnd)218   inline bool isBitcode(const unsigned char *BufPtr,
219                         const unsigned char *BufEnd) {
220     return isBitcodeWrapper(BufPtr, BufEnd) ||
221            isRawBitcode(BufPtr, BufEnd);
222   }
223 
224   /// SkipBitcodeWrapperHeader - Some systems wrap bc files with a special
225   /// header for padding or other reasons.  The format of this header is:
226   ///
227   /// struct bc_header {
228   ///   uint32_t Magic;         // 0x0B17C0DE
229   ///   uint32_t Version;       // Version, currently always 0.
230   ///   uint32_t BitcodeOffset; // Offset to traditional bitcode file.
231   ///   uint32_t BitcodeSize;   // Size of traditional bitcode file.
232   ///   ... potentially other gunk ...
233   /// };
234   ///
235   /// This function is called when we find a file with a matching magic number.
236   /// In this case, skip down to the subsection of the file that is actually a
237   /// BC file.
238   /// If 'VerifyBufferSize' is true, check that the buffer is large enough to
239   /// contain the whole bitcode file.
SkipBitcodeWrapperHeader(const unsigned char * & BufPtr,const unsigned char * & BufEnd,bool VerifyBufferSize)240   inline bool SkipBitcodeWrapperHeader(const unsigned char *&BufPtr,
241                                        const unsigned char *&BufEnd,
242                                        bool VerifyBufferSize) {
243     // Must contain the offset and size field!
244     if (unsigned(BufEnd - BufPtr) < BWH_SizeField + 4)
245       return true;
246 
247     unsigned Offset = support::endian::read32le(&BufPtr[BWH_OffsetField]);
248     unsigned Size = support::endian::read32le(&BufPtr[BWH_SizeField]);
249     uint64_t BitcodeOffsetEnd = (uint64_t)Offset + (uint64_t)Size;
250 
251     // Verify that Offset+Size fits in the file.
252     if (VerifyBufferSize && BitcodeOffsetEnd > uint64_t(BufEnd-BufPtr))
253       return true;
254     BufPtr += Offset;
255     BufEnd = BufPtr+Size;
256     return false;
257   }
258 
259   const std::error_category &BitcodeErrorCategory();
260   enum class BitcodeError { CorruptedBitcode = 1 };
make_error_code(BitcodeError E)261   inline std::error_code make_error_code(BitcodeError E) {
262     return std::error_code(static_cast<int>(E), BitcodeErrorCategory());
263   }
264 
265 } // end namespace llvm
266 
267 namespace std {
268 
269 template <> struct is_error_code_enum<llvm::BitcodeError> : std::true_type {};
270 
271 } // end namespace std
272 
273 #endif // LLVM_BITCODE_BITCODEREADER_H
274