1LLVM-OBJCOPY(1) LLVM LLVM-OBJCOPY(1)
2
6 llvm-objcopy - object copying and editing tool
7
9 llvm-objcopy [options] input [output]
10
12 llvm-objcopy is a tool to copy and manipulate objects. In basic usage,
13 it makes a semantic copy of the input to the output. If any options are
14 specified, the output may be modified along the way, e.g. by removing
15 sections.
16 If no output file is specified, the input file is modified in-place. If
18 "-" is specified for the input file, the input is read from the pro‐
19 gram's standard input stream. If "-" is specified for the output file,
20 the output is written to the standard output stream of the program.
21 If the input is an archive, any requested operations will be applied to
23 each archive member individually.
24 The tool is still in active development, but in most scenarios it works
26 as a drop-in replacement for GNU's objcopy.
27
29 The following options are either agnostic of the file format, or apply
30 to multiple file formats.
31 --add-gnu-debuglink <debug-file>
33 Add a .gnu_debuglink section for <debug-file> to the output.
34 --disable-deterministic-archives, -U
36 Use real values for UIDs, GIDs and timestamps when updating ar‐
37 chive member headers.
38 --discard-all, -x
40 Remove most local symbols from the output. Different file for‐
41 mats may limit this to a subset of the local symbols. For exam‐
42 ple, file and section symbols in ELF objects will not be dis‐
43 carded.
44 --enable-deterministic-archives, -D
46 Enable deterministic mode when copying archives, i.e. use 0 for
47 archive member header UIDs, GIDs and timestamp fields. On by de‐
48 fault.
49 --help, -h
51 Print a summary of command line options.
52 --only-section <section>, -j
54 Remove all sections from the output, except for sections named
55 <section>. Can be specified multiple times to keep multiple
56 sections.
57 --regex
59 If specified, symbol and section names specified by other
60 switches are treated as extended POSIX regular expression pat‐
61 terns.
62 --remove-section <section>, -R
64 Remove the specified section from the output. Can be specified
65 multiple times to remove multiple sections simultaneously.
66 --strip-all-gnu
68 Remove all symbols, debug sections and relocations from the out‐
69 put. This option is equivalent to GNU objcopy's --strip-all
70 switch.
71 --strip-all, -S
73 For ELF objects, remove from the output all symbols and non-al‐
74 loc sections not within segments, except for .gnu.warning,
75 .ARM.attribute sections and the section name table.
76 For COFF objects, remove all symbols, debug sections, and relo‐
78 cations from the output.
79 --strip-debug, -g
81 Remove all debug sections from the output.
82 --strip-symbol <symbol>, -N
84 Remove all symbols named <symbol> from the output. Can be speci‐
85 fied multiple times to remove multiple symbols.
86 --strip-symbols <filename>
88 Remove all symbols whose names appear in the file <filename>,
89 from the output. In the file, each line represents a single sym‐
90 bol name, with leading and trailing whitespace ignored, as is
91 anything following a '#'. Can be specified multiple times to
92 read names from multiple files.
93 --strip-unneeded-symbol <symbol>
95 Remove from the output all symbols named <symbol> that are local
96 or undefined and are not required by any relocation.
97 --strip-unneeded-symbols <filename>
99 Remove all symbols whose names appear in the file <filename>,
100 from the output, if they are local or undefined and are not re‐
101 quired by any relocation. In the file, each line represents a
102 single symbol name, with leading and trailing whitespace ig‐
103 nored, as is anything following a '#'. Can be specified multiple
104 times to read names from multiple files.
105 --strip-unneeded
107 Remove from the output all local or undefined symbols that are
108 not required by relocations.
109 --version, -V
111 Display the version of this program.
112
114 The following options are implemented only for COFF objects. If used
115 with other objects, llvm-objcopy will either emit an error or silently
116 ignore them.
117 --only-keep-debug
119 Remove the contents of non-debug sections from the output, but
120 keep the section headers.
121
123 The following options are implemented only for ELF objects. If used
124 with other objects, llvm-objcopy will either emit an error or silently
125 ignore them.
126 --add-section <section=file>
128 Add a section named <section> with the contents of <file> to the
129 output. The section will be of type SHT_NOTE, if the name starts
130 with ".note". Otherwise, it will have type SHT_PROGBITS. Can be
131 specified multiple times to add multiple sections.
132 --add-symbol <name>=[<section>:]<value>[,<flags>]
134 Add a new symbol called <name> to the output symbol table, in
135 the section named <section>, with value <value>. If <section> is
136 not specified, the symbol is added as an absolute symbol. The
137 <flags> affect the symbol properties. Accepted values are:
138 • global = the symbol will have global binding.
140 • local = the symbol will have local binding.
142 • weak = the symbol will have weak binding.
144 • default = the symbol will have default visibility.
146 • hidden = the symbol will have hidden visibility.
148 • file = the symbol will be an STT_FILE symbol.
150 • section = the symbol will be an STT_SECTION symbol.
152 • object = the symbol will be an STT_OBJECT symbol.
154 • function = the symbol will be an STT_FUNC symbol.
156 • indirect-function = the symbol will be an STT_GNU_IFUNC sym‐
158 bol.
159 Additionally, the following flags are accepted but ignored: de‐
161 bug, constructor, warning, indirect, synthetic, unique-object,
162 before.
163 Can be specified multiple times to add multiple symbols.
165 --allow-broken-links
167 Allow llvm-objcopy to remove sections even if it would leave in‐
168 valid section references. Any invalid sh_link fields will be set
169 to zero.
170 --binary-architecture <arch>, -B
172 Specify the architecture to use, when transforming an architec‐
173 ture-less format (e.g. binary) to another format. Valid options
174 are:
175 • aarch64
177 • arm
179 • i386
181 • i386:x86-64
183 • mips
185 • powerpc:common64
187 • riscv:rv32
189 • riscv:rv64
191 • sparc
193 • sparcel
195 • x86-64
197 --build-id-link-dir <dir>
199 Set the directory used by --build-id-link-input and
200 --build-id-link-output.
201 --build-id-link-input <suffix>
203 Hard-link the input to <dir>/xx/xxx<suffix>, where <dir> is the
204 directory specified by --build-id-link-dir. The path used is de‐
205 rived from the hex build ID.
206 --build-id-link-output <suffix>
208 Hard-link the output to <dir>/xx/xxx<suffix>, where <dir> is the
209 directory specified by --build-id-link-dir. The path used is de‐
210 rived from the hex build ID.
211 --change-start <incr>, --adjust-start
213 Add <incr> to the program's start address. Can be specified mul‐
214 tiple times, in which case the values will be applied cumula‐
215 tively.
216 --compress-debug-sections [<style>]
218 Compress DWARF debug sections in the output, using the specified
219 style. Supported styles are zlib-gnu and zlib. Defaults to zlib
220 if no style is specified.
221 --decompress-debug-sections
223 Decompress any compressed DWARF debug sections in the output.
224 --discard-locals, -X
226 Remove local symbols starting with ".L" from the output.
227 --dump-section <section>=<file>
229 Dump the contents of section <section> into the file <file>. Can
230 be specified multiple times to dump multiple sections to differ‐
231 ent files. <file> is unrelated to the input and output files
232 provided to llvm-objcopy and as such the normal copying and
233 editing operations will still be performed. No operations are
234 performed on the sections prior to dumping them.
235 --extract-dwo
237 Remove all sections that are not DWARF .dwo sections from the
238 output.
239 --extract-main-partition
241 Extract the main partition from the output.
242 --extract-partition <name>
244 Extract the named partition from the output.
245 --globalize-symbol <symbol>
247 Mark any defined symbols named <symbol> as global symbols in the
248 output. Can be specified multiple times to mark multiple sym‐
249 bols.
250 --globalize-symbols <filename>
252 Read a list of names from the file <filename> and mark defined
253 symbols with those names as global in the output. In the file,
254 each line represents a single symbol, with leading and trailing
255 whitespace ignored, as is anything following a '#'. Can be spec‐
256 ified multiple times to read names from multiple files.
257 --input-target <format>, -I
259 Read the input as the specified format. See SUPPORTED FORMATS
260 for a list of valid <format> values. If unspecified, llvm-obj‐
261 copy will attempt to determine the format automatically.
262 --keep-file-symbols
264 Keep symbols of type STT_FILE, even if they would otherwise be
265 stripped.
266 --keep-global-symbol <symbol>
268 Make all symbols local in the output, except for symbols with
269 the name <symbol>. Can be specified multiple times to ignore
270 multiple symbols.
271 --keep-global-symbols <filename>
273 Make all symbols local in the output, except for symbols named
274 in the file <filename>. In the file, each line represents a sin‐
275 gle symbol, with leading and trailing whitespace ignored, as is
276 anything following a '#'. Can be specified multiple times to
277 read names from multiple files.
278 --keep-section <section>
280 When removing sections from the output, do not remove sections
281 named <section>. Can be specified multiple times to keep multi‐
282 ple sections.
283 --keep-symbol <symbol>, -K
285 When removing symbols from the output, do not remove symbols
286 named <symbol>. Can be specified multiple times to keep multiple
287 symbols.
288 --keep-symbols <filename>
290 When removing symbols from the output do not remove symbols
291 named in the file <filename>. In the file, each line represents
292 a single symbol, with leading and trailing whitespace ignored,
293 as is anything following a '#'. Can be specified multiple times
294 to read names from multiple files.
295 --localize-hidden
297 Make all symbols with hidden or internal visibility local in the
298 output.
299 --localize-symbol <symbol>, -L
301 Mark any defined non-common symbol named <symbol> as a local
302 symbol in the output. Can be specified multiple times to mark
303 multiple symbols as local.
304 --localize-symbols <filename>
306 Read a list of names from the file <filename> and mark defined
307 non-common symbols with those names as local in the output. In
308 the file, each line represents a single symbol, with leading and
309 trailing whitespace ignored, as is anything following a '#'. Can
310 be specified multiple times to read names from multiple files.
311 --output-target <format>, -O
313 Write the output as the specified format. See SUPPORTED FORMATS
314 for a list of valid <format> values. If unspecified, the output
315 format is assumed to be the same as the input file's format.
316 --prefix-alloc-sections <prefix>
318 Add <prefix> to the front of the names of all allocatable sec‐
319 tions in the output.
320 --prefix-symbols <prefix>
322 Add <prefix> to the front of every symbol name in the output.
323 --preserve-dates, -p
325 Preserve access and modification timestamps in the output.
326 --redefine-sym <old>=<new>
328 Rename symbols called <old> to <new> in the output. Can be spec‐
329 ified multiple times to rename multiple symbols.
330 --redefine-syms <filename>
332 Rename symbols in the output as described in the file <file‐
333 name>. In the file, each line represents a single symbol to re‐
334 name, with the old name and new name separated by an equals
335 sign. Leading and trailing whitespace is ignored, as is anything
336 following a '#'. Can be specified multiple times to read names
337 from multiple files.
338 --rename-section <old>=<new>[,<flag>,...]
340 Rename sections called <old> to <new> in the output, and apply
341 any specified <flag> values. See --set-section-flags for a list
342 of supported flags. Can be specified multiple times to rename
343 multiple sections.
344 --set-section-flags <section>=<flag>[,<flag>,...]
346 Set section properties in the output of section <section> based
347 on the specified <flag> values. Can be specified multiple times
348 to update multiple sections.
349 Following is a list of supported flags and their effects:
351 • alloc = add the SHF_ALLOC flag.
353 • load = if the section has SHT_NOBITS type, mark it as a
355 SHT_PROGBITS section.
356 • readonly = if this flag is not specified, add the SHF_WRITE
358 flag.
359 • code = add the SHF_EXECINSTR flag.
361 • merge = add the SHF_MERGE flag.
363 • strings = add the SHF_STRINGS flag.
365 • contents = if the section has SHT_NOBITS type, mark it as a
367 SHT_PROGBITS section.
368 The following flags are also accepted, but are ignored for GNU
370 compatibility: noload, debug, data, rom, share.
371 --set-start-addr <addr>
373 Set the start address of the output to <addr>. Overrides any
374 previously specified --change-start or --adjust-start options.
375 --split-dwo <dwo-file>
377 Equivalent to running llvm-objcopy with --extract-dwo and
378 <dwo-file> as the output file and no other options, and then
379 with --strip-dwo on the input file.
380 --strip-dwo
382 Remove all DWARF .dwo sections from the output.
383 --strip-non-alloc
385 Remove from the output all non-allocatable sections that are not
386 within segments.
387 --strip-sections
389 Remove from the output all section headers and all section data
390 not within segments. Note that many tools will not be able to
391 use an object without section headers.
392 --target <format>, -F
394 Equivalent to --input-target and --output-target for the speci‐
395 fied format. See SUPPORTED FORMATS for a list of valid <format>
396 values.
397 --weaken-symbol <symbol>, -W
399 Mark any global symbol named <symbol> as a weak symbol in the
400 output. Can be specified multiple times to mark multiple symbols
401 as weak.
402 --weaken-symbols <filename>
404 Read a list of names from the file <filename> and mark global
405 symbols with those names as weak in the output. In the file,
406 each line represents a single symbol, with leading and trailing
407 whitespace ignored, as is anything following a '#'. Can be spec‐
408 ified multiple times to read names from multiple files.
409 --weaken
411 Mark all defined global symbols as weak in the output.
412
414 The following values are currently supported by llvm-objcopy for the
415 --input-target, --output-target, and --target options. For GNU objcopy
416 compatibility, the values are all bfdnames.
417 • binary
419 • ihex
421 • elf32-i386
423 • elf32-x86-64
425 • elf64-x86-64
427 • elf32-iamcu
429 • elf32-littlearm
431 • elf64-aarch64
433 • elf64-littleaarch64
435 • elf32-littleriscv
437 • elf64-littleriscv
439 • elf32-powerpc
441 • elf32-powerpcle
443 • elf64-powerpc
445 • elf64-powerpcle
447 • elf32-bigmips
449 • elf32-ntradbigmips
451 • elf32-ntradlittlemips
453 • elf32-tradbigmips
455 • elf32-tradlittlemips
457 • elf64-tradbigmips
459 • elf64-tradlittlemips
461 • elf32-sparc
463 • elf32-sparcel
465 Additionally, all targets except binary and ihex can have -freebsd as a
467 suffix.
468
470 If binary is used as the value for --input-target, the input file will
471 be embedded as a data section in an ELF relocatable object, with sym‐
472 bols _binary_<file_name>_start, _binary_<file_name>_end, and _bi‐
473 nary_<file_name>_size representing the start, end and size of the data,
474 where <file_name> is the path of the input file as specified on the
475 command line with non-alphanumeric characters converted to _.
476 If binary is used as the value for --output-target, the output file
478 will be a raw binary file, containing the memory image of the input
479 file. Symbols and relocation information will be discarded. The image
480 will start at the address of the first loadable section in the output.
481
483 llvm-objcopy exits with a non-zero exit code if there is an error.
484 Otherwise, it exits with code 0.
485
487 To report bugs, please visit <http://llvm.org/bugs/>.
488 There is a known issue with --input-target and --target causing only
490 binary and ihex formats to have any effect. Other values will be ig‐
491 nored and llvm-objcopy will attempt to guess the input format.
492
494 llvm-strip(1)
495
497 Maintained by the LLVM Team (https://llvm.org/).
498
500 2003-2021, LLVM Project
5019 2021-07-22 LLVM-OBJCOPY(1)