1LLVM-AR(1) LLVM Command Guide LLVM-AR(1)
2
6 llvm-ar - LLVM archiver
7
9 llvm-ar [-]{dmpqrtx}[Rabfikouz] [relpos] [count] <archive> [files...]
10
12 The llvm-ar command is similar to the common Unix utility, "ar". It
13 archives several files together into a single file. The intent for this
14 is to produce archive libraries by LLVM bitcode that can be linked into
15 an LLVM program. However, the archive can contain any kind of file. By
16 default, llvm-ar generates a symbol table that makes linking faster
17 because only the symbol table needs to be consulted, not each
18 individual file member of the archive.
19 The llvm-ar command can be used to read both SVR4 and BSD style archive
21 files. However, it cannot be used to write them. While the llvm-ar
22 command produces files that are almost identical to the format used by
23 other "ar" implementations, it has two significant departures in order
24 to make the archive appropriate for LLVM. The first departure is that
25 llvm-ar only uses BSD4.4 style long path names (stored immediately
26 after the header) and never contains a string table for long names. The
27 second departure is that the symbol table is formated for efficient
28 construction of an in-memory data structure that permits rapid (red-
29 black tree) lookups. Consequently, archives produced with llvm-ar
30 usually won't be readable or editable with any "ar" implementation or
31 useful for linking. Using the "f" modifier to flatten file names will
32 make the archive readable by other "ar" implementations but not for
33 linking because the symbol table format for LLVM is unique. If an SVR4
34 or BSD style archive is used with the "r" (replace) or "q" (quick
35 update) operations, the archive will be reconstructed in LLVM format.
36 This means that the string table will be dropped (in deference to BSD
37 4.4 long names) and an LLVM symbol table will be added (by default).
38 The system symbol table will be retained.
39 Here's where llvm-ar departs from previous "ar" implementations:
41 Symbol Table
43 Since llvm-ar is intended to archive bitcode files, the symbol
44 table won't make much sense to anything but LLVM. Consequently, the
45 symbol table's format has been simplified. It consists simply of a
46 sequence of pairs of a file member index number as an LSB 4byte
47 integer and a null-terminated string.
48 Long Paths
50 Some "ar" implementations (SVR4) use a separate file member to
51 record long path names (> 15 characters). llvm-ar takes the BSD 4.4
52 and Mac OS X approach which is to simply store the full path name
53 immediately preceding the data for the file. The path name is null
54 terminated and may contain the slash (/) character.
55 Compression
57 llvm-ar can compress the members of an archive to save space. The
58 compression used depends on what's available on the platform and
59 what choices the LLVM Compressor utility makes. It generally favors
60 bzip2 but will select between "no compression" or bzip2 depending
61 on what makes sense for the file's content.
62 Directory Recursion
64 Most "ar" implementations do not recurse through directories but
65 simply ignore directories if they are presented to the program in
66 the files option. llvm-ar, however, can recurse through directory
67 structures and add all the files under a directory, if requested.
68 TOC Verbose Output
70 When llvm-ar prints out the verbose table of contents ("tv"
71 option), it precedes the usual output with a character indicating
72 the basic kind of content in the file. A blank means the file is a
73 regular file. A 'Z' means the file is compressed. A 'B' means the
74 file is an LLVM bitcode file. An 'S' means the file is the symbol
75 table.
76
78 The options to llvm-ar are compatible with other "ar" implementations.
79 However, there are a few modifiers (zR) that are not found in other
80 "ar"s. The options to llvm-ar specify a single basic operation to
81 perform on the archive, a variety of modifiers for that operation, the
82 name of the archive file, and an optional list of file names. These
83 options are used to determine how llvm-ar should process the archive
84 file.
85 The Operations and Modifiers are explained in the sections below. The
87 minimal set of options is at least one operator and the name of the
88 archive. Typically archive files end with a ".a" suffix, but this is
89 not required. Following the archive-name comes a list of files that
90 indicate the specific members of the archive to operate on. If the
91 files option is not specified, it generally means either "none" or
92 "all" members, depending on the operation.
93 Operations
95 d Delete files from the archive. No modifiers are applicable to this
96 operation. The files options specify which members should be
97 removed from the archive. It is not an error if a specified file
98 does not appear in the archive. If no files are specified, the
99 archive is not modified.
100 m[abi]
102 Move files from one location in the archive to another. The a, b,
103 and i modifiers apply to this operation. The files will all be
104 moved to the location given by the modifiers. If no modifiers are
105 used, the files will be moved to the end of the archive. If no
106 files are specified, the archive is not modified.
107 p[k]
109 Print files to the standard output. The k modifier applies to this
110 operation. This operation simply prints the files indicated to the
111 standard output. If no files are specified, the entire archive is
112 printed. Printing bitcode files is ill-advised as they might
113 confuse your terminal settings. The p operation never modifies the
114 archive.
115 q[Rfz]
117 Quickly append files to the end of the archive. The R, f, and z
118 modifiers apply to this operation. This operation quickly adds the
119 files to the archive without checking for duplicates that should be
120 removed first. If no files are specified, the archive is not
121 modified. Because of the way that llvm-ar constructs the archive
122 file, its dubious whether the q operation is any faster than the r
123 operation.
124 r[Rabfuz]
126 Replace or insert file members. The R, a, b, f, u, and z modifiers
127 apply to this operation. This operation will replace existing files
128 or insert them at the end of the archive if they do not exist. If
129 no files are specified, the archive is not modified.
130 t[v]
132 Print the table of contents. Without any modifiers, this operation
133 just prints the names of the members to the standard output. With
134 the v modifier, llvm-ar also prints out the file type (B=bitcode,
135 Z=compressed, S=symbol table, blank=regular file), the permission
136 mode, the owner and group, the size, and the date. If any files are
137 specified, the listing is only for those files. If no files are
138 specified, the table of contents for the whole archive is printed.
139 x[oP]
141 Extract archive members back to files. The o modifier applies to
142 this operation. This operation retrieves the indicated files from
143 the archive and writes them back to the operating system's file
144 system. If no files are specified, the entire archive is extract.
145 Modifiers (operation specific)
147 The modifiers below are specific to certain operations. See the
148 Operations section (above) to determine which modifiers are applicable
149 to which operations.
150 [a] When inserting or moving member files, this option specifies the
152 destination of the new files as being "a"fter the relpos member. If
153 relpos is not found, the files are placed at the end of the
154 archive.
155 [b] When inserting or moving member files, this option specifies the
157 destination of the new files as being "b"efore the relpos member.
158 If relpos is not found, the files are placed at the end of the
159 archive. This modifier is identical to the the i modifier.
160 [f] Normally, llvm-ar stores the full path name to a file as presented
162 to it on the command line. With this option, truncated (15
163 characters max) names are used. This ensures name compatibility
164 with older versions of "ar" but may also thwart correct extraction
165 of the files (duplicates may overwrite). If used with the R option,
166 the directory recursion will be performed but the file names will
167 all be "f"lattened to simple file names.
168 [i] A synonym for the b option.
170 [k] Normally, llvm-ar will not print the contents of bitcode files when
172 the p operation is used. This modifier defeats the default and
173 allows the bitcode members to be printed.
174 [N] This option is ignored by llvm-ar but provided for compatibility.
176 [o] When extracting files, this option will cause llvm-ar to preserve
178 the original modification times of the files it writes.
179 [P] use full path names when matching
181 [R] This modifier instructions the r option to recursively process
183 directories. Without R, directories are ignored and only those
184 files that refer to files will be added to the archive. When R is
185 used, any directories specified with files will be scanned
186 (recursively) to find files to be added to the archive. Any file
187 whose name begins with a dot will not be added.
188 [u] When replacing existing files in the archive, only replace those
190 files that have a time stamp than the time stamp of the member in
191 the archive.
192 [z] When inserting or replacing any file in the archive, compress the
194 file first. This modifier is safe to use when (previously)
195 compressed bitcode files are added to the archive; the compressed
196 bitcode files will not be doubly compressed.
197 Modifiers (generic)
199 The modifiers below may be applied to any operation.
200 [c] For all operations, llvm-ar will always create the archive if it
202 doesn't exist. Normally, llvm-ar will print a warning message
203 indicating that the archive is being created. Using this modifier
204 turns off that warning.
205 [s] This modifier requests that an archive index (or symbol table) be
207 added to the archive. This is the default mode of operation. The
208 symbol table will contain all the externally visible functions and
209 global variables defined by all the bitcode files in the archive.
210 Using this modifier is more efficient that using llvm-ranlib which
211 also creates the symbol table.
212 [S] This modifier is the opposite of the s modifier. It instructs llvm-
214 ar to not build the symbol table. If both s and S are used, the
215 last modifier to occur in the options will prevail.
216 [v] This modifier instructs llvm-ar to be verbose about what it is
218 doing. Each editing operation taken against the archive will
219 produce a line of output saying what is being done.
220
222 The llvm-ar utility is intended to provide a superset of the IEEE Std
223 1003.2 (POSIX.2) functionality for "ar". llvm-ar can read both SVR4 and
224 BSD4.4 (or Mac OS X) archives. If the "f" modifier is given to the "x"
225 or "r" operations then llvm-ar will write SVR4 compatible archives.
226 Without this modifier, llvm-ar will write BSD4.4 compatible archives
227 that have long names immediately after the header and indicated using
228 the "#1/ddd" notation for the name in the header.
229
231 The file format for LLVM Archive files is similar to that of BSD 4.4 or
232 Mac OSX archive files. In fact, except for the symbol table, the "ar"
233 commands on those operating systems should be able to read LLVM archive
234 files. The details of the file format follow.
235 Each archive begins with the archive magic number which is the eight
237 printable characters "!<arch>\n" where \n represents the newline
238 character (0x0A). Following the magic number, the file is composed of
239 even length members that begin with an archive header and end with a \n
240 padding character if necessary (to make the length even). Each file
241 member is composed of a header (defined below), an optional newline-
242 terminated "long file name" and the contents of the file.
243 The fields of the header are described in the items below. All fields
245 of the header contain only ASCII characters, are left justified and are
246 right padded with space characters.
247 name - char[16]
249 This field of the header provides the name of the archive member.
250 If the name is longer than 15 characters or contains a slash (/)
251 character, then this field contains "#1/nnn" where "nnn" provides
252 the length of the name and the "#1/" is literal. In this case, the
253 actual name of the file is provided in the "nnn" bytes immediately
254 following the header. If the name is 15 characters or less, it is
255 contained directly in this field and terminated with a slash (/)
256 character.
257 date - char[12]
259 This field provides the date of modification of the file in the
260 form of a decimal encoded number that provides the number of
261 seconds since the epoch (since 00:00:00 Jan 1, 1970) per Posix
262 specifications.
263 uid - char[6]
265 This field provides the user id of the file encoded as a decimal
266 ASCII string. This field might not make much sense on non-Unix
267 systems. On Unix, it is the same value as the st_uid field of the
268 stat structure returned by the stat(2) operating system call.
269 gid - char[6]
271 This field provides the group id of the file encoded as a decimal
272 ASCII string. This field might not make much sense on non-Unix
273 systems. On Unix, it is the same value as the st_gid field of the
274 stat structure returned by the stat(2) operating system call.
275 mode - char[8]
277 This field provides the access mode of the file encoded as an octal
278 ASCII string. This field might not make much sense on non-Unix
279 systems. On Unix, it is the same value as the st_mode field of the
280 stat structure returned by the stat(2) operating system call.
281 size - char[10]
283 This field provides the size of the file, in bytes, encoded as a
284 decimal ASCII string. If the size field is negative (starts with a
285 minus sign, 0x02D), then the archive member is stored in compressed
286 form. The first byte of the archive member's data indicates the
287 compression type used. A value of 0 (0x30) indicates that no
288 compression was used. A value of 2 (0x32) indicates that bzip2
289 compression was used.
290 fmag - char[2]
292 This field is the archive file member magic number. Its content is
293 always the two characters back tick (0x60) and newline (0x0A). This
294 provides some measure utility in identifying archive files that
295 have been corrupted.
296 The LLVM symbol table has the special name "#_LLVM_SYM_TAB_#". It is
298 presumed that no regular archive member file will want this name. The
299 LLVM symbol table is simply composed of a sequence of triplets: byte
300 offset, length of symbol, and the symbol itself. Symbols are not null
301 or newline terminated. Here are the details on each of these items:
302 offset - vbr encoded 32-bit integer
304 The offset item provides the offset into the archive file where the
305 bitcode member is stored that is associated with the symbol. The
306 offset value is 0 based at the start of the first "normal" file
307 member. To derive the actual file offset of the member, you must
308 add the number of bytes occupied by the file signature (8 bytes)
309 and the symbol tables. The value of this item is encoded using
310 variable bit rate encoding to reduce the size of the symbol table.
311 Variable bit rate encoding uses the high bit (0x80) of each byte to
312 indicate if there are more bytes to follow. The remaining 7 bits in
313 each byte carry bits from the value. The final byte does not have
314 the high bit set.
315 length - vbr encoded 32-bit integer
317 The length item provides the length of the symbol that follows.
318 Like this offset item, the length is variable bit rate encoded.
319 symbol - character array
321 The symbol item provides the text of the symbol that is associated
322 with the offset. The symbol is not terminated by any character. Its
323 length is provided by the length field. Note that is allowed (but
324 unwise) to use non-printing characters (even 0x00) in the symbol.
325 This allows for multiple encodings of symbol names.
326
328 If llvm-ar succeeds, it will exit with 0. A usage error, results in an
329 exit code of 1. A hard (file system typically) error results in an exit
330 code of 2. Miscellaneous or unknown errors result in an exit code of 3.
331
333 llvm-ranlib, ar(1)
334
336 Maintained by the LLVM Team (<http://llvm.org>).
337CVS 2010-05-07 LLVM-AR(1)