1 //===- ArchiveReader.cpp - Code to read LLVM bytecode from .a files -------===//
3 // The LLVM Compiler Infrastructure
5 // This file was developed by the LLVM research group and is distributed under
6 // the University of Illinois Open Source License. See LICENSE.TXT for details.
8 //===----------------------------------------------------------------------===//
10 // This file implements the ReadArchiveFile interface, which allows a linker to
11 // read all of the LLVM bytecode files contained in a .a file. This file
12 // understands the standard system .a file format. This can only handle the .a
13 // variant prevalent on Linux systems so far, but may be extended. See
14 // information in this source file for more information:
15 // http://sources.redhat.com/cgi-bin/cvsweb.cgi/src/bfd/archive.c?cvsroot=src
17 //===----------------------------------------------------------------------===//
19 #include "llvm/Bytecode/Reader.h"
20 #include "llvm/Module.h"
21 #include "Config/sys/stat.h"
22 #include "Config/sys/mman.h"
23 #include "Config/fcntl.h"
36 char fmag[2]; // Always equal to '`\n'
40 UserObject, // A user .o/.bc file
41 Unknown, // Unknown file, just ignore it
42 SVR4LongFilename, // a "//" section used for long file names
43 ArchiveSymbolTable, // Symbol table produced by ranlib.
47 // getObjectType - Determine the type of object that this header represents.
48 // This is capable of parsing the variety of special sections used for various
50 static enum ObjectType getObjectType(ar_hdr *H, unsigned Size) {
51 // Check for sections with special names...
52 if (!memcmp(H->name, "__.SYMDEF ", 16))
53 return ArchiveSymbolTable;
54 if (!memcmp(H->name, "// ", 16))
55 return SVR4LongFilename;
57 // Check to see if it looks like an llvm object file...
58 if (Size >= 4 && !memcmp(H+1, "llvm", 4))
65 static inline bool Error(std::string *ErrorStr, const char *Message) {
66 if (ErrorStr) *ErrorStr = Message;
70 static bool ParseLongFilenameSection(unsigned char *Buffer, unsigned Size,
71 std::vector<std::string> &LongFilenames,
73 if (!LongFilenames.empty())
74 return Error(S, "archive file contains multiple long filename entries");
77 // Long filename entries are newline delimited to keep the archive readable.
78 unsigned char *Ptr = (unsigned char*)memchr(Buffer, '\n', Size);
80 return Error(S, "archive long filename entry doesn't end with newline!");
83 if (Ptr == Buffer) break; // Last entry contains just a newline.
85 unsigned char *End = Ptr;
86 if (End[-1] == '/') --End; // Remove trailing / from name
88 LongFilenames.push_back(std::string(Buffer, End));
96 static bool ParseSymbolTableSection(unsigned char *Buffer, unsigned Size,
98 // Currently not supported (succeeds without doing anything)
102 static bool ReadArchiveBuffer(const std::string &ArchiveName,
103 unsigned char *Buffer, unsigned Length,
104 std::vector<Module*> &Objects,
105 std::string *ErrorStr) {
106 if (Length < 8 || memcmp(Buffer, "!<arch>\n", 8))
107 return Error(ErrorStr, "signature incorrect for an archive file!");
108 Buffer += 8; Length -= 8; // Skip the magic string.
110 std::vector<char> LongFilenames;
112 while (Length >= sizeof(ar_hdr)) {
113 ar_hdr *Hdr = (ar_hdr*)Buffer;
114 unsigned Size = atoi(Hdr->size);
115 if (Size+sizeof(ar_hdr) > Length)
116 return Error(ErrorStr, "invalid record length in archive file!");
118 // Get name of archive member.
119 char *startp = Hdr->name;
120 char *endp = strchr (startp, '/');
121 if (startp == endp && isdigit (Hdr->name[1])) {
122 // Long filenames are abbreviated as "/I", where I is a decimal
123 // index into the LongFilenames vector.
124 unsigned Index = atoi (&Hdr->name[1]);
125 assert (LongFilenames.size () > Index
126 && "Long filename for archive member not found");
127 startp = &LongFilenames[Index];
128 endp = strchr (startp, '/');
130 std::string MemberName (startp, endp);
131 std::string FullMemberName = ArchiveName + "(" + MemberName + ")";
133 switch (getObjectType(Hdr, Size)) {
134 case SVR4LongFilename:
135 // If this is a long filename section, read all of the file names into the
136 // LongFilenames vector.
137 LongFilenames.assign (Buffer+sizeof(ar_hdr), Buffer+sizeof(ar_hdr)+Size);
140 Module *M = ParseBytecodeBuffer(Buffer+sizeof(ar_hdr), Size,
141 FullMemberName, ErrorStr);
143 Objects.push_back(M);
146 case ArchiveSymbolTable:
147 if (ParseSymbolTableSection(Buffer+sizeof(ar_hdr), Size, ErrorStr))
151 std::cerr << "ReadArchiveBuffer: WARNING: Skipping unknown file: "
152 << FullMemberName << "\n";
153 break; // Just ignore unknown files.
156 // Round Size up to an even number...
158 Buffer += sizeof(ar_hdr)+Size; // Move to the next entry
159 Length -= sizeof(ar_hdr)+Size;
166 // ReadArchiveFile - Read bytecode files from the specified .a file, returning
167 // true on error, or false on success. This does not support reading files from
170 bool ReadArchiveFile(const std::string &Filename, std::vector<Module*> &Objects,
171 std::string *ErrorStr) {
172 int FD = open(Filename.c_str(), O_RDONLY);
174 return Error(ErrorStr, "Error opening file!");
176 // Stat the file to get its length...
178 if (fstat(FD, &StatBuf) == -1 || StatBuf.st_size == 0)
179 return Error(ErrorStr, "Error stat'ing file!");
181 // mmap in the file all at once...
182 int Length = StatBuf.st_size;
183 unsigned char *Buffer = (unsigned char*)mmap(0, Length, PROT_READ,
185 if (Buffer == (unsigned char*)MAP_FAILED)
186 return Error(ErrorStr, "Error mmapping file!");
188 // Parse the archive files we mmap'ped in
189 bool Result = ReadArchiveBuffer(Filename, Buffer, Length, Objects, ErrorStr);
191 // Unmmap the archive...
192 munmap((char*)Buffer, Length);
194 if (Result) // Free any loaded objects
195 while (!Objects.empty()) {
196 delete Objects.back();
203 } // End llvm namespace